[House Hearing, 109 Congress]
[From the U.S. Government Printing Office]




 
                   EIGHTH IN A SERIES OF SUBCOMMITTEE
                       HEARINGS ON PROTECTING AND
                     STRENGTHENING SOCIAL SECURITY

=======================================================================

                                HEARING

                               before the

                    SUBCOMMITTEE ON SOCIAL SECURITY


                      COMMITTEE ON WAYS AND MEANS
                     U.S. HOUSE OF REPRESENTATIVES

                       ONE HUNDRED NINTH CONGRESS

                             FIRST SESSION

                               __________

                             JUNE 23, 2005

                               __________

                           Serial No. 109-27

                               __________

         Printed for the use of the Committee on Ways and Means



                    U.S. GOVERNMENT PRINTING OFFICE
26-386                      WASHINGTON : 2006
_____________________________________________________________________________
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov  Phone: toll free (866) 512-1800; (202) 512ï¿½091800  
Fax: (202) 512ï¿½092250 Mail: Stop SSOP, Washington, DC 20402ï¿½090001

                      COMMITTEE ON WAYS AND MEANS

                   BILL THOMAS, California, Chairman

E. CLAY SHAW, JR., Florida           CHARLES B. RANGEL, New York
NANCY L. JOHNSON, Connecticut        FORTNEY PETE STARK, California
WALLY HERGER, California             SANDER M. LEVIN, Michigan
JIM MCCRERY, Louisiana               BENJAMIN L. CARDIN, Maryland
DAVE CAMP, Michigan                  JIM MCDERMOTT, Washington
JIM RAMSTAD, Minnesota               JOHN LEWIS, Georgia
JIM NUSSLE, Iowa                     RICHARD E. NEAL, Massachusetts
SAM JOHNSON, Texas                   MICHAEL R. MCNULTY, New York
PHIL ENGLISH, Pennsylvania           WILLIAM J. JEFFERSON, Louisiana
J.D. HAYWORTH, Arizona               JOHN S. TANNER, Tennessee
JERRY WELLER, Illinois               XAVIER BECERRA, California
KENNY C. HULSHOF, Missouri           LLOYD DOGGETT, Texas
RON LEWIS, Kentucky                  EARL POMEROY, North Dakota
MARK FOLEY, Florida                  STEPHANIE TUBBS JONES, Ohio
KEVIN BRADY, Texas                   MIKE THOMPSON, California
THOMAS M. REYNOLDS, New York         JOHN B. LARSON, Connecticut
PAUL RYAN, Wisconsin                 RAHM EMANUEL, Illinois
ERIC CANTOR, Virginia
JOHN LINDER, Georgia
BOB BEAUPREZ, Colorado
MELISSA A. HART, Pennsylvania
CHRIS CHOCOLA, Indiana
DEVIN NUNES, California

                    Allison H. Giles, Chief of Staff

                  Janice Mays, Minority Chief Counsel

                                 ______

                    SUBCOMMITTEE ON SOCIAL SECURITY

                    JIM MCCRERY, Louisiana, Chairman

E. CLAY SHAW JR., Florida            SANDER M. LEVIN, Michigan
SAM JOHNSON, Texas                   EARL POMEROY, North Dakota
J.D. HAYWORTH, Arizona               XAVIER BECERRA, California
KENNY C. HULSHOF, Missouri           STEPHANIE TUBBS JONES, Ohio
RON LEWIS, Kentucky                  RICHARD E. NEAL, Massachusetts
KEVIN BRADY, Texas
PAUL RYAN, Wisconsin

Pursuant to clause 2(e)(4) of Rule XI of the Rules of the House, public 
hearing records of the Committee on Ways and Means are also published 
in electronic form. The printed hearing record remains the official 
version. Because electronic submissions are used to prepare both 
printed and electronic versions of the hearing record, the process of 
converting between various electronic formats may introduce 
unintentional errors or omissions. Such occurrences are inherent in the 
current publication process and should diminish as the process is 
further refined.


                            C O N T E N T S

                              ----------                              Page
Advisory of June 16, 2005 announcing the hearing.................     2

                               WITNESSES

Social Security Administration, Hon. James B. Lockhart, Deputy 
  Commissioner...................................................     5
U.S. government Accountability Office, Barbara D. Bovbjerg, 
  Director, Education, Workforce, and Income Security............    10
                                 ------                                
American Enterprise Institute, Alex J. Pollock...................    52
Congressional Research Service, Patrick J. Purcell...............    42
Federal Retirement Thrift Investment Board, Francis X. Cavanaugh.    67
National Academy of Social Insurance, Virginia P. Reno...........    74
National Women's Law Center, Joan Entmacher......................    58
Wharton School, University of Pennsylvania, Olivia S. Mitchell, 
  Ph.D...........................................................    35

                       SUBMISSIONS FOR THE RECORD

Clark, Robert, North Carolina State University, Raleigh, NC, 
  statement......................................................    89
Copeland, Craig, Employee Benefit Research Institute, statement..    91
Cyr, Paul, Greene, ME, statement.................................    93
Jones, Ike, America's Community Bankers, statement...............    94
Lancon, Renee, West Hills, CA, statement.........................    96
Mitchell, Olivia, Wharton School, Univ. of Pennsylvania, 
  Philadelphia, PA, statement....................................    99
Sewell, Robin, Littleton, MA, statement..........................   104


                   EIGHTH IN A SERIES OF SUBCOMMITTEE
                       HEARINGS ON PROTECTING AND
                     STRENGTHENING SOCIAL SECURITY

                              ----------                              


                        THURSDAY, JUNE 23, 2005

             U.S. House of Representatives,
                       Committee on Ways and Means,
                            Subcommittee on Social Security
                                                    Washington, DC.

    The Subcommittee met, pursuant to notice, at 10:00 a.m., in 
room B-318, Rayburn House Office Building, Hon. Jim McCrery 
(Chairman of the Subcommittee) presiding.
    [The advisory announcing the hearing follows:]

ADVISORY FROM THE COMMITTEE ON WAYS AND MEANS

                    SUBCOMMITTEE ON SOCIAL SECURITY

                                                CONTACT: (202) 225-9263
FOR IMMEDIATE RELEASE
June 16, 2005
No. SS-8

                McCrery Announces Eighth in a Series of

                        Subcommittee Hearings on

              Protecting and Strengthening Social Security

    Congressman Jim McCrery (R-LA), Chairman, Subcommittee on Social 
Security of the Committee on Ways and Means, today announced that the 
Subcommittee will hold the eighth in a series of Subcommittee hearings 
on protecting and strengthening Social Security. The hearing will 
examine options for the administration of personal retirement accounts. 
The hearing will take place on Thursday, June 23, 2005, in room B-318 
Rayburn House Office Building, beginning at 10:00 a.m.
      
    In view of the limited time available to hear witnesses, oral 
testimony at this hearing will be from invited witnesses only. However, 
any individual or organization not scheduled for an oral appearance may 
submit a written statement for consideration by the Subcommittee and 
for inclusion in the printed record of the hearing.
      

BACKGROUND:

      
    The past two Administrations, the 1994-1996 Social Security 
Advisory Council, and the 2001 President's Commission to Strengthen 
Social Security, have laid out proposals to establish personal accounts 
that are either integrated with Social Security benefits or in addition 
to Social Security benefits. Personal accounts have been proposed as a 
means to enhance individuals' retirement income, as Social Security's 
Trustees have warned that current-law promised Social Security benefits 
cannot be paid in full in the future absent action to address the 
program's long-term insolvency.
      
    An important aspect of the development of personal accounts 
involves system design, including account management and recordkeeping, 
investment options, and methods to pay benefits to workers at 
retirement. An equally important aspect is that administrative expenses 
must be kept low to preserve workers' account balances.
      
    Numerous nonpartisan studies, including studies by the U.S. 
Government Accountability Office and the Social Security 
Administration, indicate that system design issues are of vital 
importance for the successful widespread use of personal retirement 
accounts. While the design possibilities are many and varied, the 
experience derived from the management of other large-scale retirement 
systems, such as the Federal Thrift Savings Plan, can provide valuable 
insights.
      
    In announcing the hearing, Chairman McCrery stated, ``As we 
consider how personal retirement accounts would strengthen retirement 
security, we must acknowledge that the proper design of a personal 
account system is not a mere technical detail. Rather, we must 
carefully consider key questions on implementation, administration, and 
public education to ensure workers receive the quality service they 
deserve, along with low expenses that preserve account balances.''

FOCUS OF THE HEARING:

      
    The hearing will focus on options for designing a system of 
personal retirement accounts to ensure that the accounts are managed 
efficiently and accurately, with low administrative fees to preserve 
account balances. Options for paying out account balances at retirement 
will also be examined.
      

DETAILS FOR SUBMISSION OF WRITTEN COMMENTS:

      
    Please Note: Any person(s) and/or organization(s) wishing to submit 
for the hearing record must follow the appropriate link on the hearing 
page of the Committee website and complete the informational forms. 
From the Committee homepage, http://waysandmeans.house.gov, select 
``109th Congress'' from the menu entitled, ``Hearing Archives'' (http:/
/waysandmeans.house.gov/Hearings.asp?congress=17). Select the hearing 
for which you would like to submit, and click on the link entitled, 
``Click here to provide a submission for the record.'' Once you have 
followed the online instructions, completing all informational forms 
and clicking ``submit'' on the final page, an email will be sent to the 
address which you supply confirming your interest in providing a 
submission for the record. You MUST REPLY to the email and ATTACH your 
submission as a Word or WordPerfect document, in compliance with the 
formatting requirements listed below, by close of business Thursday, 
July 7, 2005. Finally, please note that due to the change in House mail 
policy, the U.S. Capitol Police will refuse sealed-package deliveries 
to all House Office Buildings. For questions, or if you encounter 
technical problems, please call (202) 225-1721.
      

FORMATTING REQUIREMENTS:

      
    The Committee relies on electronic submissions for printing the 
official hearing record. As always, submissions will be included in the 
record according to the discretion of the Committee. The Committee will 
not alter the content of your submission, but we reserve the right to 
format it according to our guidelines. Any submission provided to the 
Committee by a witness, any supplementary materials submitted for the 
printed record, and any written comments in response to a request for 
written comments must conform to the guidelines listed below. Any 
submission or supplementary item not in compliance with these 
guidelines will not be printed, but will be maintained in the Committee 
files for review and use by the Committee.
      
    1. All submissions and supplementary materials must be provided in 
Word or WordPerfect format and MUST NOT exceed a total of 10 pages, 
including attachments. Witnesses and submitters are advised that the 
Committee relies on electronic submissions for printing the official 
hearing record.
      
    2. Copies of whole documents submitted as exhibit material will not 
be accepted for printing. Instead, exhibit material should be 
referenced and quoted or paraphrased. All exhibit material not meeting 
these specifications will be maintained in the Committee files for 
review and use by the Committee.
      
    3. All submissions must include a list of all clients, persons, 
and/or organizations on whose behalf the witness appears. A 
supplemental sheet must accompany each submission listing the name, 
company, address, telephone and fax numbers of each witness.
      
    Note: All Committee advisories and news releases are available on 
the World Wide Web at http://waysandmeans.house.gov.
      
    The Committee seeks to make its facilities accessible to persons 
with disabilities. If you are in need of special accommodations, please 
call 202-225-1721 or 202-226-3411 TTD/TTY in advance of the event (four 
business days notice is requested). Questions with regard to special 
accommodation needs in general (including availability of Committee 
materials in alternative formats) may be directed to the Committee as 
noted above.

                                 

    Chairman MCCRERY. The hearing will come to order. Good 
morning. Welcome, everyone, to our eighth Subcommittee hearing 
on protecting and strengthening Social Security. Today, we will 
look at system design issues for personal accounts, including 
account management and recordkeeping, annuities, administrative 
fees, and issues related to converting account balances into a 
monthly income. Yesterday, I and several of my colleagues here 
on the Subcommittee, Mr. Shaw, Mr. Johnson, and Mr. Ryan, took 
the first step on stopping the raid on Social Security. Our 
Chairman, Bill Thomas, called this a common sense approach, and 
I think the American people would agree. It is simple common 
sense that the Social Security surplus needs to be preserved 
for retirement income, and H.R. 3304, the ``Growing Real 
Ownership for Workers (GROW) Accounts Act,'' will do just that.
    The GROW accounts would be fully inheritable, voluntary, 
and initially invested in safe, marketable Treasury bonds. 
Naturally, the success of any type of personal account system 
involves the careful consideration of system design issues. 
America's workers and their families deserve a voluntary 
personal account system that will give them excellent service, 
offer prudent and diversified investment options, protect their 
investments with good stewardship, and preserve their account 
balances through low administrative fees. I welcome our 
distinguished panels today, and I look forward to hearing your 
views. Mr. Levin?
    Mr. LEVIN. Thank you, Mr. Chairman. I am glad this hearing 
has been called so we can delve into issues like offsets and 
clawbacks, so that we can understand fully the administrative 
complexities, and some of the testimony will very much focus on 
that, and I think bring out some complexities that aren't 
understood. We will also have a discussion of administrative 
costs, and we need to do that. Indeed, the timing of this 
hearing could not be more salient because of the developments 
in the last 24 hours. It is very clear, if it wasn't before, 
that privatization of Social Security is basically what this 
argument is all about.
    The proposals of the last 24 hours make it very clear that 
privatization is at the heart of the Republican approach, the 
replacement of Social Security with private accounts, but 
privatization at any cost. What has been suggested in recent 
hours, we would continue the use of Federal Insurance 
Contributions Act (FICA) taxes other than for Social Security, 
one way or another, solvency would be made worse. The fiscal 
irresponsibility of this Congress and this President could 
continue because Social Security funds could continue to be 
used for other purposes, and uncertainty would really increase 
for everyone. Instead of a guaranteed benefit in a Social 
Security system for people who are retired or who are going to 
retire, for people who become disabled, for survivors. Instead 
of the certainty of a guaranteed benefit, we would end up with 
uncertainty for everybody, for those who would be in the Social 
Security system and for those who would have these private 
accounts. So, I think it is useful now to delve into these 
issues of what private accounts would mean in terms of 
complexity, in terms of cost, in terms of offsets. So, let's 
get with it.
    Chairman MCCRERY. Thank you, Mr. Levin. Our first panel 
this morning is a familiar one to us on the Committee on Ways 
and Means, the Honorable James B. Lockhart, Deputy Commissioner 
of the Social Security Administration (SSA) and Ms. Barbara 
Bovbjerg, Director of Education, Workforce and Income Security 
with the U.S. government Accountability Office (GAO). Welcome 
back, it is nice to have you. We welcome your testimony this 
morning. Your entire written testimony will be included in the 
record. If you could summarize that in about five minutes each, 
we would appreciate it. Mr. Lockhart.

     STATEMENT OF THE HONORABLE JAMES B. LOCKHART, DEPUTY 
          COMMISSIONER, SOCIAL SECURITY ADMINISTRATION

    Mr. LOCKHART. Thank you, Chairman McCrery, and Members of 
the Subcommittee. Thank you for holding this series of hearings 
on protecting and strengthening Social Security, and for 
inviting me to discuss the administration of voluntary personal 
accounts. The idea of creating a system of personal accounts as 
part of reforming Social Security has received much attention, 
including proposals developed by President Clinton and 
President Bush's Social Security Commissions. There are a 
number of models for administering a system of personal 
accounts. The most recent personal accounts proposals envision 
a voluntary program with a centrally administered system 
modeled on the Thrift Savings Plan (TSP), and Social Security's 
existing annual wage reporting system. This is not a 
decentralized small business 401(k) model.
    Assuming a centralized approach, Social Security 
independent actuaries estimate an ultimate cost of 30 basis 
points of assets under management for personal accounts. This 
cost seems reasonable, and in my opinion, may fall as the 
amount of assets under management grows. There will be 
significant challenges, but I believe that over a reasonable 
implementation period of perhaps 3 years, we could produce an 
efficient, equitable, and accurate system of personal 
retirement accounts. Social Security personal accounts could 
quickly achieve very large economies of scale. Under the 
President's personal accounts framework, there would be 120 
million personal accounts, with assets totaling $602 billion by 
the end of 2015. As a centralized system of personal accounts 
is developed, its business processes should reflect a coherent 
set of underlying designed principles.
    As I detail in my written testimony, there are seven basic 
processes: Education, Enrollment, Contribution, Investment, 
Recordkeeping, Compliance, and Payout. There are proven, low-
cost models for each of these processes. Many proposals suggest 
creating a new government agency with an independent board 
similar to the TSP that would have primary responsibility for 
administering the plan, with strong support from the SSA.
    The TSP is a very good, low-cost model for personal 
accounts. It has 3.4 million participants, and $157 billion of 
assets under management. It offers five investment 
alternatives, including a Treasury bond option, and four very 
low-cost index funds which are designed to replicate the 
returns of broad, corporate bond and stock market indexes. It 
is also adding a life cycle fund this summer. The TSP's costs 
are six basis points, and the investment manager receives very, 
very little of that. Some people have suggested that the TSP is 
not an appropriate model because there are about six-and-a-half 
million employers in the United States, while there are only 
about 130 government agencies providing payroll information to 
the TSP. While this is true, I believe our proven and low-cost 
wage reporting process could provide a single interface between 
employers and self-employed individuals and the program's 
central record keeper.
    Social Security processes 240 million W-2s annually for 
approximately 149 million workers. Over the last 6 years, 
electronic filing has grown from just 7 percent to 65 percent. 
In 3 years, we are targeting to hit 82 percent. Social Security 
already makes software available that allows small employers to 
report electronically. Social Security receives W-2s beginning 
in January for the previous year. Due to the lag in reporting, 
some plans suggest there could be a holding fund, where the 
contributions would be deposited until individual accounts 
could be credited, with not only the amount taken out from 
payroll taxes, but also the interest on that amount.
    A strengthened Social Security program that includes 
personal accounts that provide individual Americans with 
ownership and more personal control over their retirement 
income is feasible. Using the model of the TSP, along with 
systems already in place, costs could be minimized. As the 
assets grow, more choices and flexibilities could be added at a 
reasonable cost. As President Bush said in his State of the 
Union Address, and I quote, ``The goal here is greater security 
in retirement, so, we will set careful guidelines for personal 
accounts.'' Commissioner Barnhart and I are committed to 
strengthening and protecting Social Security, and to making 
sure the SSA is ready to assist the Administration and Congress 
in doing so. As the Social Security trustees, President Bush, 
and many others have said, we need to take action soon to save 
Social Security for future generations. Thank you, and I will 
be happy to address any questions.
    [The prepared statement of Mr. Lockhart follows:]

  Statement of The Honorable James B. Lockhart, Deputy Commissioner, 
                     Social Security Administration

    I would like to thank Chairman McCrery and the Members of the 
Subcommittee for holding this series of hearings on protecting and 
strengthening Social Security and for inviting me to discuss the 
administration of voluntary personal retirement accounts.
    My remarks today are based on the work that has been done by many 
groups as well as over 30 years of off and on personal experience with 
corporate pensions. My experiences range from serving on the pension 
committee of one of the largest corporations in the country, to 
starting a 401(k) plan for a small business, to serving as the 
Executive Director of the Pension Benefit Guaranty Corporation.
    Over the past decade, the idea of creating a system of personal 
accounts as part of reforming Social Security has received much 
attention, including proposals developed by the 1994-96 Social Security 
Advisory Council and the 2001 President's Commission to Strengthen 
Social Security. As a result, SSA has looked at the issues involved in 
administering such a program. We have also studied the Federal Thrift 
Savings Plan (TSP) and met with their staff and their systems 
contractor to discuss pertinent issues and draw on their experience. I 
have also seen SSA successfully implement other large projects such as 
our new electronic disability system and the rollout of the new 
application for extra help with prescription drug costs under Part D of 
Medicare.
    A number of models for administering a system of personal accounts 
have been discussed. Over the last several years, most personal account 
proposals envision a voluntary program with a centrally administered 
system modeled on the TSP.
    Assuming this approach, Social Security's independent actuaries 
reduce their estimate of real investment returns of 4.9% by an ultimate 
annual cost of 0.3% (30 basis points) of assets under management for 
personal accounts. A basis point is 1/100th of one percent, so 30 basis 
points would equal $3 per $1,000 of assets. This cost seems reasonable 
and, in my opinion, may fall as the system matures and gets larger.
    There would be significant challenges, but I believe that over a 
reasonable implementation period of perhaps three years, we could 
produce an efficient, equitable, and accurate system of personal 
retirement accounts.
    I want to note that I have not been talking about a small business 
corporate 401(k) plan model, which some critics of personal accounts 
do. 401(k) plans are expensive for small businesses to administer 
because of their small scale and because of the regulatory burden. Nor 
am I using a large employer 401(k) model that has lots of choices, 
although some do have annual fees less than 30 basis points. To the 
contrary, I am focusing on a centralized model using the Social 
Security Administration's existing annual wage reporting (Form W-2) 
system and the proven TSP model. I believe that this approach is the 
key to successful and timely implementation of a personal account 
system.
    Social Security personal accounts could quickly become the largest 
defined contribution plan in the world and achieve very large economies 
of scale. Assuming a two-thirds participation rate, the Social Security 
actuaries project that under the President's personal account 
framework, which phases in personal accounts from 2009 to 2011, there 
would be about 120 million personal accounts with assets totaling $602 
billion (in constant 2004 dollars) at the end of 2015.
    As a centralized system of personal accounts is developed, its 
business processes should reflect a coherent set of underlying design 
principles. The goal should be to enable the pieces of the new system 
to fit together in a seamless manner, keep administrative costs to a 
minimum, speed implementation, and boost public acceptance. There are 7 
basic processes in a personal account system, which are:

      Education--The process of providing plan information at 
various points in time and to distinct categories of individuals, 
enabling them to make informed decisions.

    Initially, the general public would need to be educated about the 
plan's structure, operation, and participation benefits so that workers 
could decide whether or not to enroll. Subsequently, those who enroll 
would need more detailed information about investment opportunities and 
the status of their accounts. Finally, those about to retire and 
beneficiaries of deceased participants would need information that 
outlines options available for accessing account assets.

      In addition to the initial, one-time education of the 
general public about the account plan, a continuous educational program 
would need to be in place for the over 4 million new workers a year who 
would need to make an enrollment decision.
      Enrollment--The process of obtaining a worker's consent 
and supporting information to create a personal account. The supporting 
information includes the person's identifying information, investment 
fund selections, beneficiary data, and contact information.

    A plan could enroll workers via the Internet and machine readable 
paper forms. Enrollment could be done by ``opting-in'' or ``opting-
out''. The latter means that a worker would have to fill out a form 
stating that he or she did not want to volunteer for personal accounts. 
In the corporate 401(k) world, opting-out has successfully raised 
enrollment rates. No matter which enrollment option was chosen, the 
administrator would likely need to mail out confirming ``welcome'' 
packages to enrollees that acknowledge receipt of the applications and 
provide more detailed educational materials about the plan.
    This process entails a large, initial start-up process to establish 
accounts for current workers and then a continuing process to create 
accounts for new workers.

      Contribution--The process of collecting, verifying, and 
crediting wage information and money from approximately 6.5 million 
employers and 15 million self-employed workers.

    The most efficient method for collecting the contributions would be 
through the current payroll deduction process, as it would demand the 
least change for employers. However, other options exist that would 
entail more frequent employer reporting, electronic reporting, direct 
employer reporting to the account administrator, or use of other 
reporting/collection avenues, such as the State workforce agencies.

      Investment--The process of quickly and accurately 
investing contributions in funds chosen by the participants.

    This process involves setting the plan's investment policies, 
establishing available funds for participant investment, selecting 
investment managers, making timely fund purchases, and updating account 
balances based on fund performance.

      Recordkeeping--The process of maintaining account 
information and providing service to participants.

    The process would need to provide participants with periodic 
account statements and the ability to update account records as 
personal situations change; for example, as addresses, marital status, 
or beneficiary selections change. It would also need to allow 
participants to modify how their contributions are invested and to 
reallocate their assets between funds.
    In addition, the plan would need to provide the means to answer 
account or plan-related questions from participants, beneficiaries, 
employers, and the general public. This would entail a large internet/
website operation as well as a large teleservice component.

      Compliance--The process for monitoring the program to 
ensure that workers, employers, beneficiaries, and fiduciaries comply 
with statutes and regulations.

    A plan would need to develop procedures to monitor transactions and 
audit financial records, and make corrections to account records where 
errors are detected. In addition, the plan would need to have an 
appellate process that could hear worker or employer requests for 
reconsideration of the corrective actions.

      Payout--The process for dispersing account assets to 
participants during retirement or to beneficiaries of workers who die 
prior to retirement.

    Many proposals envision that workers would annuitize to remain 
above a poverty level-related threshold. Options for releasing the 
remaining assets above the threshold include paying them out as lump 
sums or as phased withdrawals.
    Included in this process would be making various types of annuities 
available to retirees and their spouses, and possibly the 
administration of the entire annuitization program. Moreover, the 
process would need to have procedures in place to locate beneficiaries 
and deal with abandoned accounts.
    There are proven, low cost models for each of these processes. As 
others point out, there are also proven, expensive models. In designing 
a personal account system, Congress should take care to choose options 
that follow the low cost model. Over time, more flexibility and options 
could be added as needed.
    Many proposals suggest a new independent government agency with an 
independent board similar to the TSP that would have primary 
responsibility for most of these activities, with the strong support of 
the Social Security Administration. In particular, SSA could play a key 
role in the front-end education, enrollment and contribution phases and 
potentially in the payout phase. The education process in particular 
could involve many government agencies building on the Financial 
Literacy and Education Council as well as business and not-for-profits 
ongoing financial education efforts.
    The TSP is a very good, low cost role model for personal accounts. 
It has 3.4 million participants and $157 billion of assets under 
management. It offers 5 investment alternatives including a Treasury 
bond option and four very low cost indexed funds, which are invested to 
replicate the returns of broad market indexes. These four TSP funds are 
very comprehensive and include a corporate bond fund, a Standard and 
Poor's 500 fund, a U.S. smaller companies fund covering U.S. stocks not 
in the S&P 500, and an international equity fund.
    In addition, by September 2005, TSP plans to add lifecycle funds 
using a combination of investments in each of the 5 funds in their 
system. These lifecycle funds gradually and automatically move assets 
into less volatile investments as the participant gets closer to 
retirement. For instance, the TSP lifecycle fund for younger workers 
will have 85 percent invested in equities and by retirement age the 
equity percentage would be reduced to 20 percent. The idea, which 
President Bush has endorsed, is to lessen investment volatility as one 
reaches retirement age. Life cycle funds are especially appealing to 
persons who do not wish to make a fund selection or actively manage 
their accounts. At payout or retirement, TSP offers lump sums, monthly 
payouts, or annuities.
    The TSP does all of that with only 90 people plus approximately 400 
contract employees and a net cost of $95 million in FY 2005, or 6 basis 
points of the assets under management. Almost all of this cost is 
administrative fees. The fees of Barclays Global Investors, the TSP 
investment managers, represent a very small portion of those 6 basis 
points.
    The TSP recordkeeping system would be an excellent model for the 
administration of personal accounts even though it has more 
capabilities than would be needed initially for a Social Security 
personal account system. TSP and its contractor have told us that their 
computer system could be adapted to Social Security personal accounts.
    Some people have suggested that the TSP is not an appropriate model 
because there are about 6.5 million employers in the United States 
while there are only about 130 government agencies providing payroll 
information to the TSP. While this is true, I believe the existing SSA 
wage reporting process could provide a similar single interface between 
eligible employees and self-employed individuals and the program's 
central record keeper.
    Using the existing wage reporting system would provide a low cost 
and efficient way to collect contribution information. Social Security 
processes 240 million W-2s annually for approximately 149 million 
workers. We have a major push under way to increase electronic filing. 
Over the last six years, electronic filing has grown from 7 percent to 
65 percent and we are targeting 82 percent by 2008, and 95 percent by 
2012. By law, employers with over 250 employees must report 
electronically.
    Social Security begins receiving W-2s in January for the previous 
year. We begin processing immediately with 82 percent processed by 
April 30th and 99 percent by September. Because of the lag in 
reporting, the President's Commission on Strengthening Social Security 
suggested that there could be a ``holding'' fund where contributions 
would be deposited until individual accounts could be reconciled. This 
reconciliation would occur once the individual's W-2 information was 
processed and the personal account would receive the amount contributed 
plus interest.
    As you look at the design of a personal account system, you may 
want to consider basic principles that would facilitate a simple, 
efficient process that minimizes administrative burdens and participant 
costs. These include:

      Utilize existing and proven processes. As previously 
discussed, the TSP and Social Security's annual wage reporting system 
provide an excellent foundation for a timely, low cost and effective 
implementation of a Social Security personal account program.
      Minimize worker and employer burden. The public would 
most likely prefer a system that causes little additional work for 
workers and employers. Significant reporting responsibilities for 
employers could raise business costs, and may adversely impact 
employment.
      Minimize the use of paper processes. Building a new 
system would allow the unique opportunity to develop processes based on 
new and innovative methods where cost-effective processes do not 
already exist. Developing electronic-based means for collecting, 
storing, and releasing information would be consistent with e-
Government concepts, could potentially reduce administrative costs, 
speed processes, and allow smoother interface with other administrative 
systems.
      Limit investor-initiated changes. Limiting investors 
initially to a few investment allocation changes per year would reduce 
administrative costs.
      Limit account reporting. Providing investors with 
quarterly account statements would help to keep reporting costs low and 
investor inquiries to a minimum. Reducing investor non-electronic 
inquires about account statements would also help to reduce investor 
support costs.
      Prevent pre-retirement account access. By preventing 
access to accounts before retirement, administrative costs could be 
kept to a minimum, the process would be simpler, and the governing 
rules would be more understandable to participants.
      Limit distributional alternatives. Unlike the TSP, which 
has multiple annuity and payout options, the personal account program 
should limit distributional options in order to minimize administrative 
costs. As the first retirees under a personal account plan would retire 
10 years from enactment, there would be the opportunity to develop low 
cost, flexible payout alternatives.

    In conclusion, while specific issues and costs related to the 
administration of personal accounts would vary with the specifics of 
the plan, a strengthened Social Security program that includes personal 
accounts that provide individual Americans with ownership and more 
personal control over their retirement income is feasible. Using the 
model of the Federal TSP along with systems already in place, costs 
could be kept to an acceptable minimum. And as the assets under 
management and account sizes grow, allowing additional economies of 
scale, more choices and flexibilities can be added at a reasonable 
cost.
    As President Bush said in his 2005 state of the Union Address, 
``The goal here is greater security in retirement, so we will set 
careful guidelines for personal account.'' He also said, ``We'll make 
sure there are good options to protect your investments--'' That is 
very doable and I would add that a properly designed personal account 
plan could ensure a better deal for younger workers.
    Commissioner Barnhart and I are committed to strengthening and 
protecting Social Security and to making sure that SSA is ready to 
assist the Administration and Congress in doing so. Given sufficient 
time and resources, SSA could successfully implement and administer our 
share of a personal account program. I will be happy to address any 
questions you may have.


                                 

    Chairman MCCRERY. Thank you, Mr. Lockhart. Ms. Bovbjerg.

    STATEMENT OF BARBARA D. BOVBJERG, DIRECTOR, EDUCATION, 
 WORKFORCE AND INCOME SECURITY, U.S. GOVERNMENT ACCOUNTABILITY 
                             OFFICE

    Ms. BOVBJERG. Thank you, Mr. Chairman, Members of the 
Subcommittee. Thank you for continuing to invite me to discuss 
Social Security reform issues before you; I really appreciate 
it. Today, you have asked me to address issues of individual 
account design. There are many options and issues to consider 
when designing a system of individual accounts, and the choices 
that we make could affect not only participation in the 
accounts, but also, the benefits that might ultimately be 
received. I would like to structure my comments today around 
the three phases associated with retirement savings vehicles 
generally: The contribution phase, the accumulation phase, and 
the distribution phase. My remarks are drawn from a body of 
work that we have done, much of it for this Subcommittee, over 
the last several years. Let me begin with the contribution 
phase.
    Determining how contributions to an individual account will 
be made involves choices as to size, whether the account is a 
substitute or a supplement, and whether participation should be 
voluntary or mandatory. Deciding whether accounts should be 
voluntary or mandatory is one of the most important design 
considerations. While offering the choice of whether to 
participate may be desirable, voluntary accounts require 
additional design considerations that mandatory accounts do 
not. For example, voluntary accounts likely will require 
incentives to induce participation, and these can be 
substantial as well as difficult to estimate accurately. 
Administration can be more complex for voluntary accounts, 
especially if participants are permitted to opt in and out of 
the accounts periodically. Although any changes to Social 
Security, any changes at all, must be well explained to the 
public, voluntary accounts make public education campaigns even 
more important.
    Let me turn now to the accumulation phase. Critical 
decisions will be needed as to what investment choices will be 
offered and how the funds will be managed. A wide range of 
investment choices would offer individuals the ability to 
customize their investments to their own financial needs and 
preferences, but raises the risks that individuals could invest 
unwisely, and find accounts insufficient to finance their 
retirement. Wider choices also mean higher administration 
costs. Limiting investment choice would help to minimize risk, 
of course, but would limit the possible returns. Similarly, a 
centrally managed system would take advantage of economies of 
scale, as Mr. Lockhart mentioned, while a more decentralized 
approach would be more flexible, but also more costly. 
Essentially, the challenge is finding the right balance between 
individual choice, and risks and costs to the individual and to 
the government.
    The last piece I will discuss is the distribution phase. 
Individual accounts could use three basic ways to pay 
retirement benefits: annuitization, timed withdrawals, and lump 
sum payments. Individuals could be allowed to choose from the 
three approaches, but offering such a range of choice could 
risk individuals outliving their retirement resources, and 
would represent relatively high administrative costs. 
Alternatively, mandatory annuitization could help ensure 
retirement income for the participants' lifetimes and would 
minimize adverse selection, which, in turn, would keep annuity 
prices lower, but would retain control over payouts from the 
individual.
    Most importantly, policy makers must also consider whether 
to allow participants to borrow against their accounts at 
relatively low interest rates, as is permitted with 401(k) 
plans. Our work has shown that this option is important to 
inducing participation in voluntary savings plans, but results 
in lower account balances at retirement. While having access to 
one's own savings may likely be considered a basic aspect of 
account ownership, this consideration would have to be balanced 
against the potential diminution of retirement income that such 
a policy would risk. In conclusion, the wide range of possible 
options complicates the design of an individual account system. 
Our work shows that providing flexibility and choice generally 
increases system costs and complexity, and these will be 
important tradeoffs to consider if Social Security reform 
includes such restructuring.
    Although bigger picture policy concerns necessarily 
dominate the Social Security debate at this stage, if 
individual accounts are to be features of the reformed program, 
design and administrative specifics much be addressed well in 
advance to changes of the law. This is not only for ease of 
implementation, but also to assure that time is allotted for 
the necessary public education. Even if an existing account 
structure, such as the TSP, were to be used for Social 
Security, the sheer number and diversity of Social Security 
participants and employers would complicate implementation and 
would require careful planning. That concludes my statement, 
Mr. Chairman. I am happy to answer any questions.
    [The prepared statement of Ms. Bovbjerg follows:]

 Statement of Barbara Bovbjerg, Director of Education, Workforce, and 
         Income Security, U.S. Government Accountability Office

    Mr. Chairman and Members of the Subcommittee:
    I am pleased to be here today to discuss options for designing a 
system of individual accounts within the Social Security program. 
Social Security forms the foundation for our retirement income system 
and, in so doing, provides critical benefits to millions of Americans. 
However, the Social Security program is facing significant future 
financial challenges as a result of profound demographic changes. A 
wide variety of proposals to reform the program are currently being 
discussed, including restructuring the program to incorporate 
individual accounts. When designing a system with individual accounts, 
there are many options and issues to consider, such as whether the 
accounts should be voluntary or mandatory, the amount of choice 
individuals have over their investments, and how and when the funds are 
withdrawn from the accounts. The choices that have to be made will 
affect not only participation in the accounts, but also the amount of 
savings accumulated in the accounts and the benefit received from the 
account.
    Today I will discuss options for the design of individual accounts 
specifically corresponding to the phases of a pension or similar 
retirement savings vehicle: the contribution phase, the accumulation 
phase, and the distribution phase. GAO has conducted several studies 
related to the design, implementation, and administration of individual 
accounts. My statement is largely based on that work. \1\
---------------------------------------------------------------------------
    \1\ See the list of related GAO products at the end of this 
statement.
---------------------------------------------------------------------------
    In summary, the creation of an individual account system faces key 
design decisions in each of the phases that comprise the dynamics of a 
retirement savings vehicle. For example, regarding contributions, the 
size of the contribution and whether the accounts will be mandatory or 
voluntary must be decided. This decision will be shaped to some degree 
by the implicit relationship of the accounts to the current Social 
Security program. In the accumulation phase, individual account design 
must negotiate a number of trade-offs in setting, for example, the 
amount of choice in investment options and the level of customer 
service provided. Finally, individual accounts, like current defined 
contribution (DC) plans and individual retirement accounts (IRAs), must 
distribute accumulated account balances to individuals. A system of 
individual accounts covering over 156 million workers would constitute 
a fundamental change to Social Security and would be significantly 
larger than any existing retirement investment program. Affected 
individuals need to know about and understand the features of such a 
new system to make informed life decisions about work, savings, and 
retirement.

BACKGROUND
    According to the Social Security Trustees' 2005 intermediate, or 
best-estimate, assumptions, Social Security's cash surplus begins to 
decline in 2009, and in 2017 cash flow is expected to turn negative. In 
addition, all of the accumulated Treasury obligations held by the trust 
funds are expected to be exhausted by 2041. Social Security's long-term 
financing shortfall stems primarily from the fact that people are 
living longer and having fewer children. As a result, the number of 
workers paying into the system for each beneficiary has been falling 
and is projected to decline from 3.3 today to about 2 by 2040.
    A common feature of many Social Security reform proposals is the 
creation of a system of individual accounts. Individual accounts would 
generally not by themselves achieve solvency for the Social Security 
system. Achieving solvency requires more revenue, lower benefits, or 
both. Many proposals that incorporate a system of individual accounts 
into the current program would reduce benefits under the current system 
and make up for those reductions to some degree with income from the 
individual accounts. Individual accounts also try to increase revenues, 
in effect, by providing the potential for higher rates of return on 
account investments than the trust funds would earn under the current 
system, but this exposes workers to a greater degree of risk.
    Three key distinctions help to identify the differences between 
Social Security's current structure and one that would create 
individual accounts.
    Insurance verus savings. Social Security is a form of insurance, 
while individual accounts would be a form of savings. As social 
insurance, Social Security protects workers and their dependents 
against a variety of risks such as the inability to earn income due to 
death, disability, or old age. In contrast, a savings account provides 
income only from individuals' contributions and any interest on them; 
in effect, individuals insure themselves under a savings approach.
    Defined benefit.Social Security provides a defined benefit (DB) 
pension while individual accounts would provide a defined contribution 
(DC) pension. Defined benefit pensions typically determine benefit 
amounts using a formula that takes into account individuals' earnings 
and years of earnings. The provider assumes the financial and insurance 
risk associated with funding those promised benefit levels. Defined 
contribution pensions, such as 401(k) plans, determine benefit amounts 
based on the contributions made to the accounts and any earnings on 
those contributions. As a result, the individual bears the financial 
and insurance risks under a defined contribution plan until retirement. 
\2\
---------------------------------------------------------------------------
    \2\ At retirement, individuals have the option of purchasing an 
annuity with their defined contribution accounts, which then transfers 
the financial and insurance risk to the annuity provider. Before 
retirement, individuals may also have the option of purchasing deferred 
annuities.
---------------------------------------------------------------------------
    Pay-as-you-go verus full funding. Social Security is financed 
largely on a pay-as-you-go basis, while individual accounts would be 
fully funded. In a pay-as-you-go system, contributions that workers 
make in a given year fund the payments to beneficiaries in that same 
year, and the system's trust funds are kept to a relatively small 
contingency reserve. \3\ In contrast, in a fully funded system, 
contributions for a given year are put aside to pay for future 
benefits. The investment earnings on these funds contribute 
considerable revenues and reduce the size of contributions that would 
otherwise be required to pay for the benefits. Defined contribution 
pensions and individual retirement savings accounts are fully funded by 
definition. Both mandatory and voluntary individual account plans would 
reflect all of these distinctions.
---------------------------------------------------------------------------
    \3\ Social Security is now temporarily deviating from pure pay-as-
you-go financing by building up substantial trust fund reserves. Social 
Security is collecting more in revenues than it pays in benefits each 
year partly because the baby boom generation makes the size of the 
workforce larger relative to the beneficiary population. In 2017, 
shortly after the baby boomers start to retire, the benefit payments 
are expected to exceed revenues, and the trust fund reserves and the 
interest they earn will help pay the baby boomers' retirement benefits. 
For more detail about this temporary trust fund buildup and how it 
interacts with the federal budget, see GAO, Social Security Reform: 
Demographic Trends Underlie Long-Term Financing Shortage, GAO/T-HEHS-
98-43 (Washington, D.C.: Nov. 20, 1997).
---------------------------------------------------------------------------
    In addition to these key distinctions, options for the design of 
individual accounts can be grouped in three categories corresponding to 
the different phases of a retirement savings vehicle:

      contribution phase: who should contribute, how much, and 
with what funds;
      accumulation phase: how are funds invested to make them 
grow; and
      distribution phase: how much of a benefit is received, 
when is it received, and in what form is it received.

    As we have reported previously with respect to Social Security 
reform as a whole, as policy makers decide whether and how to create a 
system of individual accounts, they must balance a range of difficult 
concerns. These concerns include broad macroeconomic issues, such as 
how to finance the accounts and how the accounts would affect the 
economy and program solvency, as well as program benefit issues, such 
as how to balance opportunities for improved individual investment 
returns with the need to maintain an adequate income for those who rely 
on Social Security the most. No less important is the need to consider 
how readily individual accounts could be implemented, administered, and 
explained to the public. An essential challenge would be to help people 
understand the relationship between their individual accounts and 
traditional Social Security benefits, thereby avoiding any gap in 
expectations about current or future benefits. Individuals would also 
need to be informed enough to make prudent investment decisions, which 
would require investor education, especially if individual accounts 
were mandatory. This would be especially important for individuals who 
are unfamiliar with making investment choices.

DESIGN CONSIDERATIONS IN THE CONTRIBUTION PHASE
    Determining how contributions to an individual account will be made 
requires choices about the role these contributions play vis-`-vis the 
current Social Security system. These choices include determining the 
size and role of contributions, management of contributions, whether 
the account is a substitute or a supplement, and whether participation 
in the accounts should be voluntary or mandatory.

Size and Role of Contributions
    An individual account plan can provide for contributions in a 
variety of ways. For example, a plan might set contributions at a fixed 
rate, such as 2 percent of pay, or allow a range of rates with, 
possibly, a certain dollar limit. Some proposals provide for greater 
average contribution rates for lower earners than for higher earners. 
Individual accounts could be designed to include some progressive 
features, which could mirror the redistributive effects of the current 
Social Security program. For example, contribution rates may go down 
gradually as earnings rise, or alternatively, all workers might pay a 
fixed percentage but have a dollar cap on contribution amounts.
    Ultimately the size of the individual account contribution rate 
determines the relative role of the DC aspect of the account versus the 
DB portion of the Social Security program. As a result, depending on 
their design, individual accounts will have a varying effect on the 
adequacy of benefits for certain subgroups of beneficiaries. For 
instance, disabled beneficiaries leave the workforce sooner than 
retired workers. With fewer years to make contributions (and accrue 
interest), disabled beneficiaries will likely have smaller account 
balances. At the same time, reform provisions that disfavor subgroups 
of earners can be offset by other provisions that favor them. As a 
result, any evaluations of reform proposals should not focus solely on 
individual account proposals but should consider both the DC and DB 
aspect of a proposal's provisions as a whole.

Management of Contributions
    In managing individual accounts, contributions might be collected 
and deposited by the government in a centralized process or by 
employers or account providers in a decentralized process. Under a 
centralized process, which would build on the current payroll reporting 
and tax collection system, a federal agency, such as the Social 
Security Administration, would assume record-keeping responsibilities. 
Alternatively, a new centralized government clearinghouse could assume 
responsibility for centralized record keeping, similar to the structure 
for the federal Thrift Savings Plan. A decentralized structure could 
build on the system that has grown up around employer-sponsored 401(k) 
plans or individually managed IRAs. Under 401(k) plans, individual 
records are maintained by either the employer or a separate entity 
hired to manage the plan, or both. Under an IRA, the record-keeping 
responsibility rests with the individual investor and the financial 
institution where the funds are invested.

Substitute versus Supplementary Contributions
    Individual accounts can either supplement current Social Security 
contributionsor substitute for all or part of them.With supplemental 
accounts, sometimes referred to as add-ons, the individual account and 
contributions to it have no effect on existing Social Security 
benefits. The supplemental account approach effectively leaves the 
entire current 12.4 percent payroll tax contribution available to 
finance the program while dedicating additional revenues for individual 
accounts. With substitute accounts, or carve-outs, the existing Social 
Security benefit is reduced (or offset) in some way to account for 
contributions that have been diverted from the program. \4\ The obvious 
effect is that less revenue is available to finance the current benefit 
structure, which creates a problem of transition costs. Absent any 
other reforms, these transition costs increase in proportion to the 
individual account contribution rate. This means that either benefits 
must be reduced or additional resources must be devoted to the defined 
benefit portion of the Social Security program in the near term. The 
trade-off to incurring transition costs is that the expected higher 
rate of return on the individual accounts may permit somewhat higher 
benefits to be paid, although with increased risk.
---------------------------------------------------------------------------
    \4\ In GAO's work to date, we have used the term ``add-on'' 
accounts to refer to accounts that would have no effect on Social 
Security benefits, would supplement those benefits, and would draw 
contributions from new revenue streams. In contrast, we have used the 
term ``carve-out'' accounts to refer to accounts that would result in 
some reduction or offset to Social Security benefits because 
contributions to those accounts would draw on existing Social Security 
revenues. Others have used these terms in different manners. For 
example, some have used ``add-ons'' in connection with new individual 
accounts funded from new revenue sources that result in a reduction or 
offset to some or all Social Security benefits. In the final analysis, 
there are two key dimensions: first, whether individual accounts are 
funded from existing or new revenue sources; second, whether individual 
accounts result in some reduction or offset to Social Security 
benefits.
---------------------------------------------------------------------------
Voluntary Contributions Require Additional Considerations
    Another important design feature to consider with respect to the 
contribution phase is whether the individual account is voluntary or 
mandatory. As we have previously reported, voluntary individual 
accounts require additional design considerations that mandatory 
accounts do not. \5\ For instance a voluntary account could offer 
participants the ability to opt in and opt out of the account 
periodically; most U.S. proposals for voluntary accounts have not 
explicitly considered whether people would face a onetime or a periodic 
decision to participate. Individuals may consider the extent of such 
flexibility in deciding whether to participate in the accounts. 
Moreover, the need to track individuals' participation decisions 
requires additional administrative tasks and complexity. Educational 
efforts would be needed to inform individuals if their participation in 
an individual account would be advantageous or not, especially if the 
account substitutes for existing Social Security benefits.
---------------------------------------------------------------------------
    \5\ See GAO, Social Security Reform: Information on Using a 
Voluntary Approach to Individual Accounts, GAO-03-309, (Washington, 
D.C.: March 10, 2003).
---------------------------------------------------------------------------
    Voluntary individual account plans may also require incentives to 
induce participation, while mandatory plans do not. In addition to 
increasing participation, incentives generally add to the value of the 
accounts and, therefore, ultimately to retirement income. Government 
contributions and tax advantages are just a few of the potential 
incentives for voluntary individual accounts. The costs of incentives 
can be difficult to estimate and can be substantial. Further, in 
certain circumstances, the net effect of voluntary individual account 
incentives may not result in improving overall retirement income. For 
example, if the voluntary account was also supplementary, then it might 
be difficult to determine whether a voluntary account adds to total 
retirement income, as it might merely substitute for other forms of 
saving. On the other hand, if the individual accounts truly add to 
total retirement income, they allow workers the opportunity and choice 
to build up additional savings to meet both income and health care cost 
needs in retirement.
    Voluntary individual account plans can also affect the total system 
costs to the government, providers, employers, or participants, 
depending on design. In some cases, offering choice involves additional 
administrative, incentive, and educational costs. In particular, 
tracking individuals' participation decisions would require 
administrative processes that do not arise in mandatory plans. 
Moreover, the uncertainty of participation rates in turn creates 
uncertainty for a variety of costs associated with voluntary individual 
account plans. For instance if individuals accurately perceive any 
built-in incentive in the benefit offsets, given their personal 
circumstances, and make their participation decision accordingly, then 
adverse selection could result. This occurs when certain groups of 
individuals (for example, those with longer life expectancies) are more 
(or less) likely to participate than others and when such participation 
patterns result in a net cost to the government.

DESIGN CONSIDERATIONS IN THE ACCUMULATION PHASE
    A system of individual accounts would provide workers with 
opportunities to assert greater control over their retirement savings. 
Therefore, when designing a system, critical decisions would need to be 
made about who will manage and invest funds and what investment choices 
will be offered. These decisions, in part, would determine the cost and 
complexity of the system and the degree of public education needed. 
Moreover, offering the level of customer service found in the private 
sector, such as frequent deposits and accessibility of account 
information, would add costs and administrative complexity to a system.

Options for Investment Management
    Alternatives for designing the investment structure of a system of 
individual accounts range from offering the individual a limited number 
of preselected funds, such as those offered by the federal Thrift 
Savings Plan (TSP), to offering a broad array of private market 
choices, such as those available through IRAs. Options for managing 
these investment choices could vary from a centralized, government-
managed system to a decentralized, privately managed system. A 
centralized system would take advantage of economies of scale, which is 
to say that the more accounts managed by a single entity, the lower the 
cost for each; thus such an approach could have lower administrative 
costs than a decentralized system. This is especially important when 
considering that a number of individuals may initially have small 
account balances. Depending on how administrative costs are assessed, 
administrative costs may eat into the accumulated savings of all 
accounts but could have a greater impact on smaller accounts.

Tradeoffs Between Investment Choices
    There are trade-offs associated with the range of investment 
choices offered. When individuals have more investment choices, they 
have more opportunity to tailor their financial situation to their own 
tastes and preferences and assert greater control over their personal 
property. However, with a greater variety of choices comes the 
possibility that individuals will not choose a diversified portfolio or 
will simply make a bad selection, thus lessening their retirement 
income from the individual account. As the range and variety of 
investment choices grow, so does the range of possible outcomes for 
individual account returns. This means that a number of individual 
accounts could perform very well, while others will not perform well at 
all. This results in increased risk to the government that individuals 
with inadequate income will turn to the government for support through 
other programs. In addition, a wider range of investment choices can 
also lead to higher administrative costs, which, if not offset by 
significantly higher returns, could undermine retirement income for 
individuals. Limiting investment choice would help to minimize risk and 
administrative costs, but doing so could also limit the possible return 
on investments. Moreover, limiting choices raises concerns about the 
role of government in selecting the investment vehicles and the 
possibility of political influence over these selections. Essentially, 
the challenge becomes finding the right balance between individual 
choice and the related risks and costs to the individual and the 
government.
    Investment decisions become more complicated as the number of 
choices increase. If individuals do not make an investment choice, 
managers would need to decide how to invest the contributions for those 
individuals. Some have proposed placing these contributions in the 
lowest risk accounts. One such option would be to place these 
contributions in a limited number of funds and then weight individual 
portfolios differently depending on the age of the worker, similar to a 
life-cycle fund, so that workers increasingly assume less risk as they 
neared retirement.
    Public education about the choices available and the risks 
associated with each would be needed under any system. However, the 
need to educate the public about the consequences of using different 
investment strategies would be less under a system with limited choice 
than under a system with a broader range of choice. When the number of 
choices is limited, the degree of risk is more defined and the program 
is less complex. However, as the number of choices increases, the 
public would need a greater level of education to learn about the wider 
variety of investment options, to understand and use the information 
disclosed to them, and to fully appreciate the consequences of 
investment choices.

Customer Service Considerations
    Frequent statements indicating the actual account value, daily or 
periodic valuation of account balances, and the ability to transfer 
funds between investment options are some of the different services 
that could be available with individual accounts. When more services 
and more flexibility are offered, the costs and administrative 
complexity of managing the investments increase. Moreover, if 
individuals consider the individual accounts as their personal 
property, they may expect options and service consistent with those 
often provided by private sector fund managers, such as frequent 
detailed account statements and allowing frequent interfund transfers.

DESIGN CONSIDERATIONS IN THE DISTRIBUTION PHASE
    The final design element centers on how the accumulated earnings in 
individual accounts would be preserved for retirement. Ensuring that 
retirement income is available for the life of the retiree is a 
fundamental goal of Social Security. With respect to the distribution 
phase, individual account systems could use three basic ways to pay 
retirement benefits: annuitization, timed withdrawals, and lump sum 
payments. The appropriateness of additional distribution features such 
as loans or early withdrawals, which are common in 401(k) plans, would 
also need to be considered. While such features would enhance the 
account holder's sense of ownership and control, loans or early 
withdrawals create a risk for leakage of account income that could 
diminish adequacy in retirement. Further, administrative aspects of the 
distribution must be considered. These include any guarantees that may 
be offered as well as the tax treatment of the distributions.

Annuities
    Under a system of annuities, retirees would receive monthly 
payments for an agreed-upon length of time, and the size of those 
payments would depend on the total value of the individual accounts. 
Under individual account proposals, annuities would be obtained either 
through government agencies or the private market. Further, such 
annuitization could be mandatory, voluntary, or some hybrid of both. 
For example, some individual account proposals have suggested mandatory 
annuitization up to an amount necessary to avoid poverty, and then any 
remaining account monies could be distributed at the account holder's 
discretion.
    Mandatory annuitization could help ensure that the accounts 
provided retirement income for the entire remaining lifetimes of 
participants. Mandatory annuitization of accounts could also minimize 
adverse selection. Adverse selection occurs, for example, when only 
healthy people buy annuities and on average live longer than nonbuyers, 
driving up the cost of annuities. According to one study, annuity 
prices in a voluntary environment can be as much as 14 percent higher 
than they would be if every retiree were required to purchase an 
annuity. However, mandatory annuitization also effectively transfers 
income from the shorter-lived to those that are longer-lived.
    Additional design considerations for annuities include the type of 
annuities that could be offered. For example, monthly income can be a 
fixed amount per month (fixed annuity); a steadily increasing amount 
based on an index, such as the Consumer Price Index (indexed annuity); 
or a variable amount based on returns from investing the premium 
(variable annuity). Under a single-life annuity, the annuitant receives 
a guaranteed stream of payments that end with the annuitant's death. 
Under a joint and survivor annuity, the payments continue to be made, 
sometimes at a reduced rate, to a second annuitant, such as a spouse, 
on the death of the primary annuitant. For a term-certain annuity, 
payments are not contingent on the annuitant's life; instead, they are 
guaranteed for a specified period of time, such as 5 or 10 years. With 
a variable annuity, the annuitant assumes some of the risk from the 
investment returns on the annuity.
    The current Social Security retirement benefit provides a fixed 
lifetime annuity that increases with inflation. In addition, Social 
Security provides auxiliary benefits to workers' eligible spouses, 
children, and survivors without reducing the size of the worker's own 
annuity. While annuity providers could potentially replicate some of 
the features of Social Security benefits, some important features would 
not likely be replicated. Adding components such as inflation indexing 
or a joint and survivor annuity will require the primary annuitant to 
accept less monthly income than under a single-life annuity. 
Furthermore, individuals with small account balances at retirement 
could have difficulty purchasing annuities in the private sector 
insurance market. Insurers may find provision of annuities to be 
inefficient and costly for individuals with small accounts because of 
the relatively high cost of issuing monthly checks and other 
administrative costs.

Timed Withdrawals and Lumps Sums
    Other options for the payout of accounts include timed withdrawals 
(also referred to as self-annuitization) and lump sum payments. In a 
timed withdrawal, retirees specify a withdrawal schedule with the 
investment manager or record keeper. Each month, they receive their 
predetermined amount, while the balance of the individual account 
remains invested. Under a lump sum payment option, individuals may 
liquidate their accounts through a single payment at retirement and 
choose to spend or save their money according to their needs or 
desires. Both timed withdrawals and lump sums give the individual the 
most immediate control of their account. Such options also underscore 
that increased personal choice comes with increased personal 
responsibility if the retirement income is to be preserved for the long 
term.

Guarantees
    A unique distribution phase design feature of some proposals 
involves a guarantee of a certain benefit level at retirement. This 
guarantee could be provided in tandem with other benefit structure 
changes such that the worker would be guaranteed a minimum benefit. One 
such approach would guarantee the current Social Security defined 
benefit. If the individual account provided less than the current 
benefit, then the system would ensure that benefits were provided to 
fill the gap. Such an arrangement might be desirable from a benefit 
adequacy perspective but would require safeguards against the 
government becoming an insurer of excessive risk taking by individuals. 
This risk taking could occur if individuals assumed unwarranted 
investment risk knowing that the government would still guarantee a 
minimum benefit or rate of return.

Preretirement Access
    While the above design features consider design options in the 
distribution phase at retirement, individual account design may also 
consider whether to allow preretirement access. For example, most 
401(k) pension plans allow participants to borrow against their pension 
accounts at relatively low interest rates. In past work we have shown 
that preretirement access improves participation in 401(k) pension 
plans and might also be an incentive for participation in a system of 
voluntary individual accounts. \6\ However, those plan participants who 
borrow from their accounts risk having substantially lower pension 
balances at retirement and, on average, may be less economically secure 
than nonborrowers. While some may argue that individuals should be 
allowed the freedom to access income through borrowing from their 
accounts before retirement, the added complexity and potential 
diminution of retirement income need to be given serious consideration.
---------------------------------------------------------------------------
    \6\ Participants in plans that allow borrowing contribute, on 
average, 35 percent more to their pension accounts than participants in 
plans that do not allow borrowing. See GAO, 401(k) Pension Plans: Loan 
Provisions Enhance Participation but May Affect Income Security for 
Some, GAO/HEHS-98-5, (Washington, D.C.: Oct. 1, 1997).
---------------------------------------------------------------------------
Tax Treatment
    Any payout option, whether pre--or postretirement, would need to 
consider the tax treatment of the individual account distribution. 
Benefits from individual accounts could be taxed in a variety of ways. 
For example, individual account benefits could be taxed like current 
Social Security benefits. Persons who currently receive Social Security 
benefits and have income over a certain amount may have to pay taxes on 
their benefits. \7\ Generally, the higher one's total income, the 
greater the taxable part of one's benefits. Typically, up to 50 percent 
of one's benefits will be taxable. However, up to 85 percent can be 
taxable if, for example, a person filed a federal tax return and one-
half of his or her benefit and all other income exceeds $34,000. 
Alternatively, individual accounts could be taxed similarly to ordinary 
income. Individual accounts could also be treated like pension payments 
(such as DC pensions like 401k plans) or annuity payments from a 
qualified employer retirement plan, which may either be fully or 
partially taxable, depending on the type of retirement plan.
---------------------------------------------------------------------------
    \7\ Individual income tax filers pay this tax if their adjusted 
gross income plus tax-exempt interest income plus one-half their Social 
Security benefits exceeds $25,000. A married couple filing jointly will 
pay the tax if this income exceeds $32,000. These levels are not 
adjusted for inflation, so the percentage of beneficiaries paying tax 
on Social security benefits is expected to rise in the future.
---------------------------------------------------------------------------
CONCLUSIONS
    Clearly, the wide range of possible options complicates the design 
of an individual account system. In general, our work shows that the 
features that provide additional flexibility and choice may increase 
system costs. Such features would include making participation 
voluntary, rather than mandatory, and expanding the number of 
investment options. \8\ Other key decisions also have cost 
implications. For example, the contribution phase, the accumulation 
phase, and the distribution phase could each be administered in a 
centralized or decentralized manner, and at various levels by the 
government or by private contractors. In general, costs of individual 
accounts will rise with increasing decentralization.
---------------------------------------------------------------------------
    \8\ See GAO, Social Security Reform: Information on Using a 
Voluntary Approach to Individual Accounts, GAO-03-309 (Washington, 
D.C.: Mar. 10, 2003), and GAO, Social Security Reform: Implementation 
Issues for Individual Accounts, GAO/HEHS-99-122 (Washington, D.C.: June 
18, 1999).
---------------------------------------------------------------------------
    No matter what sort of features individual accounts include, any 
related administrative, management, and data systems must be developed 
and tested before the individual accounts are made available to 
American workers. If reforms are implemented with haste and key 
administrative functions are neglected, the ensuing problems have the 
potential to undermine an otherwise well-designed accounts system. The 
federal Thrift Savings Plan has been suggested as a model for providing 
a limited amount of options that reduce risk and administrative costs 
while still providing some degree of choice. While using this existing 
model could mitigate administrative issues, a system of accounts that 
spans the entire national workforce and millions of employers would be 
significantly larger and more complex than the TSP.
    The choice to include individual accounts as part of broader reform 
could fundamentally alter the defined benefit aspect of current Social 
Security benefits. Under its current structure, Social Security 
redistributes benefits to lower-income workers. Mirroring the 
redistributive effects of the current Social Security program, 
individual accounts could be designed to include some progressive 
features. However, it is important to distinguish between progressivity 
and benefit adequacy. Greater progressivity is not the same thing as 
greater adequacy and may result in less equity. As a result, any 
evaluation of a Social Security reform proposal that includes 
individual accounts should consider not only the overall costs to the 
system but also, very importantly, the impact on individuals and 
families. Administering the accounts and educating the public about a 
system of individual accounts requires difficult choices and trade-
offs; and these choices will determine the degree and speed of public 
acceptance. Ultimately, what matters most is that we maintain a strong 
retirement security system for the millions of American workers and 
their families.
    Mr. Chairman and Members of the Subcommittee, this concludes my 
prepared statement. I would be happy to respond to any questions you or 
the other Members of the Subcommittee may have.

GAO Contacts and Staff Acknowledgments
    For further information regarding this testimony, please contact 
Barbara D. Bovbjerg, Director, Education, Workforce, and Income 
Security Issues, on (202) 512-7215. Blake Ainsworth, Alicia Cackley, 
Charlie Jeszeck, Michael Collins, and Charles Ford also contributed to 
this statement.

RELATED PRODUCTS
    Social Security Reform: Answers to Key Questions. GAO-05-193SP. 
Washington, D.C.: May 2005.
    Options for Social Security Reform. GAO-05-649R. Washington, D.C.: 
May6, 2005.
    Social Security Reform: Early Action Would Be Prudent. GAO-05-397T. 
Washington, D.C.: Mar.9, 2005.
    Social Security: Distribution of Benefits and Taxes Relative to 
Earnings Level. GAO-04-747. Washington, D.C.: June 15, 2004.
    Social Security Reform: Analysis of a Trust Fund Exhaustion 
Scenario. GAO-03-907. Washington, D.C.: July29, 2003.
    Social Security and Minorities: Earnings, Disability Incidence, and 
Mortality Are Key Factors That Influence Taxes Paid and Benefits 
Received. GAO-03-387. Washington, D.C.: Apr. 23, 2003.
    Social Security Reform: Analysis of Reform Models Developed by the 
President's Commission to Strengthen Social Security. GAO-03-310. 
Washington, D.C.: Jan. 15, 2003.
    Social Security Reform: Information on Using a Voluntary Approach 
to Individual Accounts. GAO-03-309. Washington, D.C.: Mar. 10, 2003.
    Social Security: Program's Role in Helping Ensure Income Adequacy. 
GAO-02-62. Washington, D.C.: Nov.30,2001.
    Social Security Reform: Potential Effects on SSA's Disability 
Programs and Beneficiaries. GAO-01-35. Washington, D.C.: Jan.24,2001.
    Social Security Reform: Information on the Archer-Shaw Proposal. 
GAO/AIMD/HEHS-00-56. Washington, D.C.: Jan. 18, 2000.
    Social Security: Evaluating Reform Proposals. GAO/AIMD/HEHS-00-29. 
Washington, D.C.: Nov.4,1999.
    Social Security: Issues in Comparing Rates of Return with Market 
Investments. GAO/HEHS-99-110. Washington, D.C.: Aug.5,1999.
    Social Security Reform: Implementation Issues for Individual 
Accounts. GAO/HEHS-99-122. Washington, D.C.: June 18, 1999.
    Social Security: Criteria for Evaluating Social Security Reform 
Proposals. GAO/T-HEHS-99-94. Washington, D.C.: Mar.25,1999.
    Social Security: Different Approaches for Addressing Program 
Solvency. GAO/HEHS-98-33. Washington, D.C.: July22,1998.
    Social Security: Restoring Long-Term Solvency Will Require 
Difficult Choices. GAO/T-HEHS-98-95. Washington, D.C.: Feb.10,1998.

                                 

    Chairman MCCRERY. Thank you, Ms. Bovbjerg. Mr. Lockhart, 
first I want to commend you and Commissioner Barnhart on your 
efficient and results-oriented management of Social Security 
programs. Because of the aging of the population and the new 
responsibilities you have under the Medicare Modernization Act 
(P.L. 108-73), the President requested a 7.5-percent increase 
in your appropriation for administration in 2006, and I am 
pleased that the Committee on Appropriations was able to grant 
almost all of that request, and the bill will be on the floor 
today. So, congratulations.
    Mr. LOCKHART. Thank you.
    Chairman MCCRERY. Would you comment on the support that the 
SSA could give to an independent board or other entity that 
Congress would establish to administer personal accounts? Does 
the SSA already do a lot of the work that would be done 
ordinarily by such a board? If so, could that be dovetailed 
with their work to reduce the administrative expenses?
    Mr. LOCKHART. Most definitely we could be very helpful in 
this process. Certainly, as I mentioned in my testimony, in the 
collection phase, using our present annual wage reporting 
system would reduce the cost very significantly. We also have, 
as you know, 65,000 people, a very big tele-service operation, 
and field offices all over the country, which are very 
experienced in education. As you mentioned, right now we are in 
the process of educating the American people about the Medicare 
Drug Benefit Extra Help Program. So, we have a lot of 
experience, and we could use our own people or use the 
additional people that might be hired for this. I think we 
would be very helpful in the education phase, the enrollment 
phase, as I said, the contribution phase, and potentially even 
in the payout phase.
    Chairman MCCRERY. Thank you. In the next panel, we are 
going to hear from a witness who says that due to a large 
number of small employers, a TSP-type model of personal account 
administration, as proposed under the President's Commission to 
Strengthen Social Security, and several other proposals, would 
have much higher administrative fees than is currently being 
estimated by Social Security's actuaries. Do you agree or 
disagree with that assessment?
    Mr. LOCKHART. I really disagree. I have close to 30 years 
experience in the world of pensions; I have sat on big company 
pensions Committees, and boards; I have actually started a 
401(k) for a small business, so, I have a lot of experience in 
this area. My sense is the 401(k) small business model is not 
the one we would follow here. The costs of that model are 
reasonably high, and a lot of it has to do with the burden of 
regulations and other things that have to be encompassed. The 
model that I think would make sense would not be that model, 
but a much more centralized model, using again, the annual wage 
reporting system and something like the TSP. In fact, the TSP 
already has systems in place that probably could be scaled up 
to the size of the Social Security program.
    Chairman MCCRERY. You touched on the effort to educate 
seniors with respect to the new drug program under Medicare. 
Obviously, if we go to a system of personal accounts, people 
are going to need an education in their investment choices, in 
the structure of their accounts, participation, benefits, so, 
they can decide whether or not to enroll. Then once they are 
enrolled, they need information on the investment opportunities 
that will be presented to them. What does the SSA already have 
in place that could be built upon in administering a system of 
education to inform the public?
    Mr. LOCKHART. Well, certainly, Social Security has a Web 
site that is one of the most visited in government, and I think 
that would be a real foundation for any educational effort. We 
have a very large and effective communications group, 
publications we put out all the time, and we have, again, the 
field office structure, if we decide to go that way. I would 
think, to a large extent, we would try to use, not only Social 
Security, but other government agencies through the Financial 
Literacy Council that has been created. We would also hope to 
get nonprofits involved--there are several of them that have 
been created over the last few years to improve financial 
education. It can be a country-wide effort. I think the bottom 
line would be that it would really be great for the country if 
we could get more education about financial literacy and 
investment choices.
    Chairman MCCRERY. Thank you. Ms. Bovbjerg, you touched on 
this in your testimony, I am going to see if you want to 
expound on it a little bit. You talked about the balance 
between offering more investment choices and the administrative 
costs associated with offering more choices. Would you like to 
expound on that just a little bit? Have you, in your own 
research, come upon a balance that you think would be about 
right for investment choices versus administrative expenses?
    Ms. BOVBJERG. The balance you choose really depends on the 
purpose of the program, and the proposals for individual 
accounts, both here and in other countries, have varied widely 
as to what ultimately they are inclined to achieve. Our main 
message is that choice costs money. Choice also requires more 
education. You may want to have more choice for a variety of 
reasons. If you have more investment choices, you provide a 
greater range of potential returns. So, while people are under 
greater risk of having lowered returns than they expect, there 
are also people who are going to get higher returns. If you 
limit that choice, the opportunity narrows from both ends.
    I did want to comment that I think that SSA is absolutely 
well-positioned to be the center of a public education campaign 
on retirement and Social Security, whatever happens with the 
Social Security program. I think it is crucial that the SSA 
work with the U.S. Department of Labor, the U.S. Department of 
Education, and as Mr. Lockhart says, all the other agencies 
that are involved in financial literacy and saver education. 
We, as a government, have not done a good job of this in the 
past. For everyone did not know the Social Secuirty retirement 
age was rising. People to this day--the Subcommittee may not 
believe this--do not know that their pension will be offset if 
they work in non-covered employment. It is very difficult to 
get the word out, even on smaller changes that we make to the 
program. More comprehensive change would require a lot of 
attention and preparation.
    Chairman MCCRERY. Thank you. Mr. Levin.
    Mr. LEVIN. Thank you. Welcome to both of you. Mr. Lockhart, 
I just wanted to comment. You, in your testimony, say you are 
committed to strengthening and protecting Social Security; that 
is the lingo that is used. I think a more straightforward way 
is to acknowledge that private accounts would not strengthen 
and protect Social Security, but over time would replace them, 
replace Social Security benefits. That is now even more clear 
with developments of recent hours, it is privatization 
replacing Social Security. Have you ever run a TSP program?
    Mr. LOCKHART. No, I have not run a TSP program. I ran the 
Pension Guaraty Corporation, but not the TSP.
    Mr. LEVIN. Have you read Mr. Cavanaugh's testimony?
    Mr. LOCKHART. Yes, I have read his testimony many times.
    Mr. LEVIN. Are you going to be here--you won't be here, I 
suppose, when he gives his testimony.
    Mr. LOCKHART. No, I have another meeting.
    Mr. LEVIN. He has run the TSP, as you know, and he has very 
different ideas than your gloss about how relatively easy it 
would be. So, it would be useful if you would take his 
testimony and give us your response. Will you do that?
    Mr. LOCKHART. Well, first of all, we will certainly give 
you a detailed response, but I can give you some----
    Mr. LEVIN. I would like you to go through it--I don't know 
if you have read it today, but give us a written response. Can 
you do that?
    Mr. LOCKHART. I would be happy to. I can also say that we 
have met with the TSP many times, including its chairman, the 
Executive Director, and their systems contractors. Mr. 
Cavanaugh was there in the startup phase, and now this is 20 
years into the program. So----
    Mr. LEVIN. Startup, he was there until 1994, right? He was 
here from 1986 to 1994.
    Mr. LOCKHART. Right.
    [The information follows:]

    Mr. Cavanaugh states that ``the Administration's estimate of 30 
basis points is optimistically low. . . .'' To be clear from the 
outset, the Administration makes no estimates of personal account 
administrative costs. Rather, it relies upon estimates and projections 
from the independent, non-partisan Office of the Chief Actuary at the 
Social Security Administration (SSA).
    Mr. Cavanaugh argues that similar low-cost estimates for the 
administration of personal accounts have been based on large-employer 
401(k) plans or the TSP which are unrealistic for consideration as PA 
models. Instead, he contends the small business corporate 401(k) model 
is a better predictor of administrative costs and burdens. He 
illustrates his point by citing an annual $3,000 cost to employers for 
maintaining a 401(k) plan--a cost he states that many small employers 
can not afford. Mr. Cavanaugh also discusses other issues he believes 
would be barriers to implementing a PA system, including the difficulty 
of making timely investments and problems communicating with large 
numbers of system participants.
    Despite Mr. Cavanaugh's claims, personal accounts need not be 
modeled after small business 401(k) plans. Under existing 401(k) rules, 
employers not only decide investment options available to employees 
under their plans, but also assume costs for ERISA compliance, setup 
expenses, payroll administration, investment selection, and funds 
reporting. Despite differences pointed out by Mr. Cavanaugh, personal 
accounts could follow the Thrift Savings Plan (TSP) model, where a 
central administrative agency would relieve the employers of these 
administrative burdens.
    A centrally administered system could collect account contributions 
through the existing payroll deduction and wage reporting processes. 
The calculation of contributions, transfers of money, and crediting of 
personal accounts would occur between Treasury, SSA, and the PA 
administrator. Employers would have no new administrative 
responsibilities under this scenario because they would continue to 
deposit and report payroll taxes as they do today.
    Mr. Cavanaugh's numbers are highly inflated because he uses the 
decentralized small business 401(k) model. Many of the cost studies 
that use a centralized, simple PA model have mature cost similar to 
Social Security's actuaries. The Employee Benefit Research Institute 
study 1998, which he cites as backing his conclusions, also had a low 
cost scenario of 10 basis points.
    Although Mr. Cavanaugh is correct that the current wage reporting 
system identifies individual employees and self-employed workers only 
once a year, this would not preclude a PA system from providing 
participants with earnings on their contributions throughout the 
reporting period. Contributions could be deposited in a ``holding'' 
fund where they would accrue interest until the reporting and 
reconciliation process is complete each year. The contributions plus 
interest earned would then be credited to the individual accounts and 
invested in the funds of choice.
    Mr. Cavanaugh exaggerates the inability of small businesses to 
report wages. He states that 72 percent of employer reports are on 
paper and implies employers with 68 million employees would have 
difficulty reporting. Social Security has a major effort underway to 
increase electronic W-2 reporting, and we now receive nearly 61 percent 
of W-2s electronically. Only 21 percent or 48 million W-2s are filed in 
paper form and many of these are printed from small business computer 
accounting systems. We have a successful, rapidly growing and award 
winning W-2 online system that is designed for businesses with under 20 
employees, it is helping to reduce the cost of annual wage reporting. 
We would expect a major increase in electronic reporting by the time 
PAs were implemented. Electronic reporting would allow more timely wage 
reporting for PA account purposes.
    Communications is another issue raised by Mr. Cavanaugh that lends 
itself to central administration. It is worth nothing that TSP receives 
over one-half of its inquiries on loan-related issues. The President's 
proposal would not allow such loans and would likely generate fewer 
inquiries on a participant percentage basis. A focused educational 
campaign developed by a PA central administrator could reach virtually 
the entire U.S. population through the use of the Internet, television, 
radio, and print materials. In addition, a PA administrator could seek 
voluntary assistance in delivering the educational information from 
employers, unions, professional associations, financial educational 
associations and other government agencies.
    The 30-basis-point cost estimate cited by Mr. Cavanaugh was in the 
final report of the President's Commission to Strengthen Social 
Security and was provided by the Office of the Chief Actuary of SSA. 
\1\ This figure applies to a mature system of personal accounts, but is 
still significantly higher than the TSP's 6 basis point cost. The 
estimated cost reflects realistic assumptions for a system having a 
much broader participant universe and more limited administrative 
features than the TSP. A centralized system would realize large cost-
savings as the result of aggregating the contributions of virtually the 
entire U.S. workforce into a single, low-cost, limited-option system of 
personal accounts, regardless of the size of individual employers. Due 
to the potential economies of scale, in my opinion even the ultimate 
30-basis-point estimate for the cost of personal accounts may prove to 
be high in the long run.
---------------------------------------------------------------------------
    \1\ President's Commission to Strengthen Social Security. 2001. 
Strengthening Social Security and Creating Personal Wealth for All 
Americans, p. 97.
---------------------------------------------------------------------------
    Setting up a system to accommodate millions of participating 
workers would entail certain initial costs. However, administrative 
costs would decline as a percentage of assets under management as the 
system matured, startup expenses diminished, and assets accumulated. As 
an example, the administrative costs of the TSP declined from roughly 
30 basis points in 1988 to about 6 basis points in 2004, while adding 
numerous new capabilities.
    The differential between TSP and PA administrative costs would be 
less than predicted by Mr. Cavanaugh. While the TSP has the benefit a 
single employer as payroll agent, the existing SSA wage reporting 
process could provide a single interface between PA participants and 
the program's central record-keeper. Therefore, if the existing wage 
reporting process were to be used, relatively little additional up 
front administrative cost would be incurred under a PA plan.
    In addition, most PA proposals call for fewer participant services 
and options than the TSP simplifying administrative tasks and reducing 
costs. Annual contributions would be fixed by wage level, whereas TSP 
participants can alter what they contribute. PA account statements 
could be issued annually or quarterly and changes to investment 
allocations could be limited, rather than allowed daily as under the 
TSP. In addition, a system of PAs would not provide for loans or early 
withdrawal of account assets, both of which are currently available in 
the TSP and are major cost drivers to that program. All of these 
factors would reduce PA administrative costs relative to a 401(k) model 
and keep them closer to those of the TSP. As with the TSP, more 
capabilities could be added over time.
    The current centralized administration of Social Security provides 
direct evidence that Federal administrative costs for PAs could be kept 
quite low. SSA provides efficient and effective service to the public 
and employers while administering the complex Social Security and 
Supplemental Security Income programs at very low cost. The Agency 
manages staffing, facilities, automated systems, and business processes 
with the capacity to service virtually the entire U.S. workforce of 159 
million people. As example of Social Security's efficiency, the annual 
``Social Security Statement'' is mailed to 144 million people at a per 
statement cost of 38". The undeliverable mail returns are less than 4 
percent, much lower than the 25 percent that Mr. Cavanaugh stated in 
his testimony.
    Based on his questionable small business model, Mr. Cavanaugh 
arrives at a questionable recommendation, for direct investment in 
corporate stocks and bonds by the trust fund. If done in any 
significant size, as a CBO policy brief \2\stated, ``Government 
ownership of stocks could affect corporate decisionmaking, interfere 
with the nation's competitive market system, and impede the efficient 
operation of financial markets--potentially limiting growth''. Despite 
Mr. Cavanaugh's contention, direct investment by the trust fund in the 
equities markets would have precisely the same transition issues as 
would personal accounts. There is no ``multi-trillion dollar transition 
cost'' that distinguishes individually owned investment from 
government-controlled investment.
---------------------------------------------------------------------------
    \2\ Congressional Budget Office, Acquiring Financial Assets to Fund 
Future Entitlements, June 16, 2003, p. 1.
---------------------------------------------------------------------------
    In conclusion, Mr. Cavanaugh raises certain administrative cost 
issues that are important factors in the discussion of creating 
personal accounts. However, when examined closely, none of those issues 
appear to be obstacles for the implementation of PAs. Looking beyond 
premium-service small business 401(k) plans offered by some plan 
managers, the low administrative costs of the TSP and basic-service 
401(k) plans provide better cost models for designing personal 
accounts.
    Mr. Cavanaugh's assertion that ``the Administration's plan for 
universal PAs is not feasible--'' is inconsistent with the experience 
of the TSP, large company 401(k) plans, the experiences of other 
nations, and the findings of non-partisan analysts ranging from the 
Social Security Administration Office of the Actuary, to the 
Congressional Budget Office. We believe that this inconsistency is 
based largely on Mr. Cavanaugh's adoption of a number of assumptions 
that do not reflect the personal account proposals that have been put 
forward by the Administration or by many Members of Congress.

                                 

    Mr. LEVIN. All right. Ms. Bovbjerg, let me just ask you: It 
has been suggested that some of the payroll taxes that are 
placed in the form of Treasury bonds would be given to private 
accounts holders. If that were to happen, would the cash still 
be in the Treasury?
    Ms. BOVBJERG. It would depend on how the bill is written. I 
have not seen bill language, but I think it would be important 
to know where the cash goes, what the interactions are between 
the rest of the government, the trust fund and individual 
accounts, how that is scored from a budgetary perspective----
    Mr. LEVIN. If bonds are given to individual holders, the 
cash remains in the Treasury, doesn't it?
    Ms. BOVBJERG. I am not sure how to answer that question 
because I have not seen the language in the bill.
    Mr. LEVIN. How would it not be in the Treasury if simply 
you give bonds to private accountholders?
    Ms. BOVBJERG. Well, I am not sure that that is the only 
thing that is envisioned. I haven't seen it.
    Mr. LEVIN. Say that is envisioned.
    Ms. BOVBJERG. If that is the only thing, if you just give 
bonds, it is only a promise to pay later, you are correct.
    Mr. LEVIN. That money could continue to be used for any 
purpose this Congress decided, right?
    Ms. BOVBJERG. Correct.
    Mr. LEVIN. Okay. Now let me ask you about annuitization; 
mandatory versus voluntary. If annuitization is mandatory, is 
there anything that a holder of an annuity policy can pass on 
to heirs?
    Ms. BOVBJERG. No. Unless it is a joint and survivor 
annuity.
    Mr. LEVIN. Unless it is joint and survivor. If there isn't 
that provision, there is nothing to pass onto heirs, right?
    Ms. BOVBJERG. Jim reminds me that there are different kinds 
of annuities, but I was thinking of a lifetime annuity, and for 
a lifetime annuity, that is correct.
    Mr. LEVIN. All right. My time is up. Thank you.
    Chairman MCCRERY. Mr. Shaw.
    Mr. SHAW. Mr. Levin asked if the cash coming in as surplus 
goes to the Treasury and bonds are issued, where is the money? 
Well, obviously at that point it would be in the Treasury. If 
the money or the surplus was invested in the private sector, 
then the money would be in the private sector or the bond or 
stock, whatever it would be, it would be in the individual 
accounts. Or third, the surplus could just simply be put into a 
vault and not draw any interest at all, which would just be 
damn stupid. Now, is there anything else that can be done with 
the money that you can think of, either one of you, other than 
just let it go? The Social Security law was written in a way 
that requires, at this particular time, the money that is not 
spent be placed in a government-backed security. That is what 
the law says; is that not correct?
    Mr. LOCKHART. That is correct.
    Mr. SHAW. So, you don't have the privilege right now to go 
out and invest in the private sector or to make it like a 
thrift account, you have to put it into the Treasury and you 
have to receive government-secured--government bonds for that 
particular cash.
    Mr. LOCKHART. That is correct, Congressman.
    Mr. SHAW. That cash does draw interest?
    Mr. LOCKHART. The cash draws interest, yes.
    Mr. SHAW. Now, if the Congress should pass a bill similar 
to the one that Mr. Levin is referring to, that in no way stops 
the Treasury bills from going back into the trust fund as they 
do today. In addition, an additional bond in a negotiable 
security in the name of the worker is placed into the Social 
Security so that you actually have two sets of bonds. Now, does 
that at all affect the liquidity of the trust fund? Remember, 
we are not disturbing the existing model at all, and there is 
two bonds being issued on the same money, one as it is today, 
and then an additional bond with someone's name on it. Would 
that not, in fact, increase the solvency of the Social Security 
Trust Fund rather than decrease it? Is that not correct?
    Mr. LOCKHART. Again, as Barbara said, I haven't read the 
legislation, so, it is a little difficult----
    Mr. SHAW. Well, nobody has.
    Mr. LEVIN. Including you.
    Mr. SHAW. It is being written right now, but we have put 
the details----
    Mr. LOCKHART. Well, certainly, from a Social Security Trust 
Fund standpoint, if there is a bond that is issued to the trust 
fund, there would be no change. Then, going forward, I don't 
know the details if there is an offset or other issues--that 
would all need to be looked at.
    Mr. SHAW. The more bonds that are being held by the Social 
Security Trust Fund, the more solvency you will acquire.
    Mr. LOCKHART. That is right, but as you all know, when we 
come to draw down on those bonds, we have got to go to the U.S. 
Treasury. We need to get the money, and they are going to have 
to----
    Mr. SHAW. I understand that. That is a problem that we are 
going to have to be dealing with. I think the question is, and 
I think the problem that we are trying to address with this new 
bill--and it is not really fair to be talking too much about 
it, other than the fact that Sandy opened the door to my 
questioning in this particular regard. The question is, how do 
you account to the taxpayer for the surplus that is now being 
spent by the Federal Government? The answer on the bill that 
will soon be filed simply says, by putting the taxpayer's name 
on a negotiable bond that is held in the trust fund, but 
nevertheless, that that person's name is on it. This is a 
question of we are not--it is not going to be contribute that 
much to solvency, in fact, the preliminary figures we are 
receiving is it only adds about 2 years to solvency. So, it 
doesn't address the issue that the President wants to address, 
but it does address one issue, which I think is very important, 
and that puts ownership into Social Security, which is missing 
the way Social Security is written now. Thank you, Mr. 
Chairman. I yield back.
    Chairman MCCRERY. Thank you, Mr. Shaw. Mr. Becerra.
    Mr. BECERRA. Thank you, Mr. Chairman. Thank you for the 
testimony, thanks for being here today. Ms. Bovbjerg, I think 
we are going to be inviting you quite a bit, so, thank you for 
always willing to be here. Let me make sure I have a correct 
impression of where we are with the Social Security Trust Fund 
surpluses. We are still running significant surpluses in Social 
Security's account because we are collecting far more than we 
are having to pay out on a yearly basis, and that is as a 
result of the changes that were made in the early eighties, 
because at that point we were approaching a crisis, and this is 
what President Reagan and Congress decided to do to try to make 
sure we didn't have to face that problem, and also because of 
the demographics of the baby boomers, we decided to go ahead 
and try to resolve the problem into the future as well. We are 
collecting, what is it this year that we are collecting?
    Mr. LOCKHART. Well, an extra $69 billion in taxes in 2004, 
and then an additional $89 billion in interest.
    Mr. BECERRA. Eighty-nine billion in interest. It is more 
than we are collecting in the surplus--in taxes--because we 
have got so much that has been collecting over time that all 
those Treasury certificates earning this interest, that totals 
$89 billion just this year alone.
    Mr. LOCKHART. Right.
    Mr. BECERRA. So, the $89 billion plus the $69 billion gives 
us about $160 billion this year alone that Social Security will 
have received or earned above what it needs to pay out benefits 
to those who are currently required or are survivors of 
American workers who died or are disabled and who qualify for 
Social Security benefits. That is going to continue on. We 
right now have an overall collective surplus in the Social 
Security account of over a trillion dollars, close to $2 
trillion, isn't it?
    Mr. LOCKHART. Yes, $1.7 trillion.
    Mr. BECERRA. That's $1.7 trillion today of surplus money. 
That is going to continue growing until about what, 2017? To 
about $5.6 trillion. My understanding is it will reach over $6 
trillion by 2027 in surplus. That is the money that we will 
then be able to turn to to make sure we can cover any shortfall 
between what we collect in workers contributions, workers taxes 
for Social Security, and what we need, and then start paying 
out in retirees, survivor, and disability benefits, correct?
    Mr. LOCKHART. Well, that is correct. You said money, and I 
am not sure there is money there. The money will be there when 
we turn in the bonds and get it from the U.S. Treasury--and 
Treasury has to go raise the money, and that will be in the 
form of borrowing new money or increasing taxes.
    Mr. BECERRA. So, let me make sure about something, Mr. 
Lockhart; are saying that you don't believe that those Treasury 
certificates will be redeemable?
    Mr. LOCKHART. No. I believe they will be redeemed, I am 
just telling you how they have to be redeemed.
    Mr. BECERRA. So, are you telling me that they are not as 
good as money?
    Mr. LOCKHART. I am telling you that they are as good as the 
credit of the United States, and the credit of the United 
States is the power to tax the American people.
    Mr. BECERRA. So, unless you are telling me that you believe 
that a President or Congress in the year 2027 or so, or 2017, 
is going to change his or her philosophy and say we don't wish 
to repay those Treasury certificates that American workers 
helped create, it is as good as money.
    Mr. LOCKHART. It is as good as money, yes, at that point.
    Mr. BECERRA. I think that is important because, if indeed 
the American public today is contributing this money for Social 
Security, their Social Security, their parents' Social Security 
or grandparents' Social Security, who are today retired, and 
for their kids Social Security, who will be working in 20 years 
and not retiring for 50 years, I think they want to know that 
these Treasury certificates will be there into the future.
    Mr. LOCKHART. Right. Well, the money has been spent as you 
know, so, it will be new money in the future that will have to 
be raised to redeem the bonds.
    Mr. BECERRA. You raised a point. Money has been spent, and 
there has to be future money raised. You do believe that the 
money will be raised to pay these Treasury certificates that 
the U.S. Government has put its full faith and credit behind.
    Mr. LOCKHART. Certainly.
    Mr. BECERRA. Okay. In terms of money being spent, when 
President Bush took office in 2001, he had an operating surplus 
where we didn't have to touch the Social Security surplus 
moneys in order to pay for all the expenses of government, 
correct?
    Mr. LOCKHART. Well, the budget that he inherited already 
had the deficit----
    Mr. BECERRA. In 2001, if I recall--Mr. Lockhart, let me 
make sure about something. Correct me if I am wrong. In 2001, 
we experienced the largest budget surplus this country has ever 
witnessed, correct? Which meant we didn't have to touch Social 
Security surplus moneys. The President, in his State of the 
Union Address in 2001 said, we have enough money that we can do 
tax cuts--for mostly to wealthy folks--and we won't have to 
touch a cent of the Social Security surplus. Yet, every year we 
have touched and spent the Social Security surplus. So, it 
seems to me that what we need to do is figure out ways to have 
an operating budget, the one we had before, where we didn't 
touch Social Security surpluses, versus figuring out how to 
take the money out of the Social Security Trust Fund and use it 
for private accounts. If we are trying to protect surplus 
money, you don't need it put into personal accounts, you just 
don't have to spend it on other activities.
    Mr. SHAW. Will the gentleman yield to me on that point?
    Mr. BECERRA. Certainly.
    Mr. SHAW. This Committee had hearings on that very subject, 
and what we did, in fact, we used the surplus to pay down the 
general obligation debt of the United States. So, what we did 
was to take the surplus cash coming in from Social Security, as 
well as from taxes, and paid down general obligation debt.
    Mr. BECERRA. Mr. Shaw, I think you are right, that may have 
happened a little bit in 2001, but quickly thereafter, we now 
have seen nothing but massive deficits. So, today, none of the 
money that is in the Social Security surplus that is being used 
by the Bush Administration is being used to pay down debt. In 
fact, we are seeing debt increase more than we have ever seen 
in the history of this country. It is even with the use of all 
the Social Security surplus moneys that we still see debt and 
deficits, annual deficits growing at massive rates. Thank you, 
though. I appreciate your responses.
    Chairman MCCRERY. Thank you, Mr. Becerra. Mr. Hayworth.
    Mr. HAYWORTH. Thank you, Mr. Chairman. Hearings, of course, 
are very useful to draw on both the experience and expertise of 
the witnesses, but they also offer incite a recent 
understanding of history and the perception that, of course, 
departs from policy and becomes purely political. I welcome the 
assessment of my friend from California because, carefully 
ignored in his formulation of question was the sudden and 
brutal attack on the American people of 9/11, and all the 
commensurate economic difficulties, not to mention national 
trauma and challenges that this Nation confronts in the wake of 
that. Certainly, we are all entitled to offer instant 
revisionism of history, or to ignore part of the challenges 
that led to the current economic situation. One thing that I 
thought was interesting, in recent weeks, and in terms of the 
way the agenda is set, and certainly in a free press, it can be 
buried back in the news sections, and maybe some who join us 
today, Mr. Chairman, failed to notice, but I believe I heard in 
the litany from my friend from California the refrain of tax 
cuts, mostly to the wealthy, which we hear time and again.
    Of course, what we discovered recently is that revenues to 
the government actually increased. Yes, we have challenges that 
abound, but certainly the crux of the argument simply comes 
down to this: As we hear concerns about trying to deal with the 
coming demographic challenges that Social Security will 
confront, we do have a choice--I guess we could go back to what 
was done in 1982, and certainly at least one Member on the 
other side of the aisle to his credit has offered one plan 
which calls for dramatic payroll tax increases, and that is an 
option that we are free to pursue. Or perhaps we ought to 
look--and this, again, is just more a question of philosophy 
than policy, although it has great consequence for policy--we 
can raise taxes, and indeed there are those that believe in 
that command control, but there is a consequence to that, that 
may not result in the great economic expansion that we have 
seen historically when marginal tax rates are reduced, it fires 
the engines of economic opportunity and actually increases 
revenues to the government. So, there are different choices 
here.
    Again, we welcome the chance for a hearing and the chance 
to see legislation put forth, not as some of my former 
colleagues in journalism; rather than who, what, when, where, 
and why, offer their own analysis and say this is an exit 
strategy. I thought my friend from Florida made it very clear, 
the ideas put on the table yesterday and the legislation being 
formulated there is an entrance strategy to try and confront 
this very real challenge. Of course, to the extent that policy 
is predicated on politics, there is that holding pattern--I am 
using a diplomatic turn, others might more accurately define it 
as obstruction and no, no, no, no, no, no, no--which some may 
feel strategically politically offers an advantage in the 2006 
elections, doesn't do much for our kids and what transpires in 
coming years, but it is interesting to get that insight. I see 
my time is about to expire. Mr. Chairman, I ask your 
indulgence. Just one question to Ms. Bovbjerg here. Given the 
experience of the Federal TSP and private sector defined 
contribution employer pension plans, how many investment 
choices are workers usually provided?
    Ms. BOVBJERG. It varies tremendously, it really does. The 
TSP, I know, has currently five different funds, and I think 
one of the things we have talked about in various work we have 
done is a way to look at the range that you can provide that at 
the same time offers some structure, some protections for 
people who might otherwise invest unwisely.
    Mr. HAYWORTH. You offered a cautionary note in your 
testimony that I heard briefly. At what point do you think the 
administrative and personal costs associated with increasing 
the number of investment options outweighs the benefits of 
having those options?
    Ms. BOVBJERG. I am not really the person to judge that; 
that is really a policy decision for the Congress. You must 
consider whether it is more important to allow people to earn 
higher returns but with assuming greater risk and how much you 
think you want to spend in administrative costs. These are 
really the trade offs; the more choice, the more it will cost 
to manage it.
    Mr. HAYWORTH. That is one of the challenges. Mr. Lockhart, 
your assessment of----
    Mr. LOCKHART. Well, in my experience, many 401(k) plans 
offer a lot more options than the five that the TSP does. Some 
of the studies I have seen show that if you offer too many 
choices it confuses people, and they don't actually even sign 
up for a 401(k). So, you have to be careful about offering too 
many choices. The TSP range is pretty extensive--it really 
covers the total U.S. stock market and bond market, and also 
the international stock market. So, with those five funds, you 
really capture a large part of the investment world.
    Mr. HAYWORTH. Thank you, both. Thank you, Mr. Chairman.
    Chairman MCCRERY. Thank you, Mr. Hayworth. Mr. Hulshof.
    Mr. HULSHOF. Thank you, Mr. Chairman. As a quick aside, I 
would say to my friends from Michigan and California, you might 
want to rescue your colleague from New York, because I think 
Mr. Ryan is actually making some headway with Mr. Rangel.
    Mr. LEVIN. I am not worried.
    Mr. HULSHOF. I don't know, they have been pretty animated.
    Mr. RANGEL. Paul Ryan makes a lot of sense because he is 
the only one that is discussing Social Security with me on the 
Republican side.
    Mr. RYAN. You are getting me in trouble again.
    Mr. HULSHOF. Mr. Chairman, if you don't have any pride in 
authorship, a Ryan-Rangel bill sounds pretty good.
    Mr. LEVIN. I don't think you would like it.
    Mr. HULSHOF. I do want to make a good point, and Mr. 
Becerra is a good friend and a colleague, and we have had a 
number of hearings, and you have consistently made the point 
about excess payroll taxes and what has been done with them. 
You and I have bantered back and forth about what happened in 
the 19nineties, and then of course we did have a period of 
surpluses, operating budget surpluses to use your term, and so, 
that is why I think possibly what has been discussed in the 
last 24 hours or 48 hours, as far as taking those excess 
payroll taxes, those surplus funds that are coming in, as you 
pointed out, all the way through 2017 and allowing those to be 
used in a personal account for those who choose them--again, we 
are talking about possible voluntary accounts. Ms. Bovbjerg, 
let me just go directly to your written statement. On page six, 
you aptly point out that voluntary individual accounts would 
require additional considerations that mandatory accounts 
don't. What I wanted to do was, and then you go on to state--
and again, for the record, since I know that this is being seen 
by others, Members of Congress do pay into Social Security--I 
know there is a misnomer out there that we don't. We do, but we 
do have the TSP.
    What I wanted to ask each of you, or either of you is, Ms. 
Bovbjerg, you suggest that workers might be offered the 
opportunity to opt in and out of participation periodically, in 
other words, it is not just a one-time decision that I am never 
going to participate, but you might see, for instance, a co-
worker who chooses to opt in that might see their portfolio, 
this nest egg begin to grow, and they might decide well, I am 
missing the boat here. I know Members of Congress have the 
option to opt out of Thrift Savings; they don't have to take 
Thrift Savings, but then they are allowed to participate as 
they wish in certain additional amounts of contributions. What 
would be some of the advantages or disadvantages of allowing 
those who, making that decision to participate, allowing them 
to enroll or disenroll periodically? What additional challenges 
would we have if we were to allow that periodic opting in and 
opting out?
    Ms. BOVBJERG. It is partly a recordkeeping issue, 
particularly if people are in and out, depending on how they 
are feeling about the market. It is one thing to run--the TSP 
is relatively small compared to what we would be talking about, 
to run something like that and keep the records, it is a little 
different thing when you are with two million people that they 
are keeping records for. The SSA does really well at keeping 
records for about 270 million members, but you would then have 
to track whether you are accepting contributions, what is 
happening with their investment. It would just complicate 
things. It is not that it is impossible by any means, it just 
makes it a little bit more complex. Your education effort is a 
little different, it is not: okay, I have to decide by next 
week if I am going to do this or not, and that is it. If it is 
in and out, you are going to have a more ongoing educational 
effort.
    Mr. HULSHOF. Again, Mr. Lockhart, I want to get to you 
before my time expires. For instance, again, the TSP, for folks 
who don't understand the voluntary option that we have, there 
are certain periods of time called open season, for instance. 
We know that that is coming, it has been an educational 
process, but we know that during this, say, 30-day calendar 
period of time, we could choose to change--in fact, though, it 
is even to the point where technology would allow us, I think 
daily, should we choose to, to change the percentage of these 
five different types of funds--and again, I understand that we 
are talking about a finite number and then a larger number, but 
Mr. Lockhart, I think in your testimony you indicated that even 
with the complexity and the additional numbers, millions of 
people that might choose this--again, it is voluntary--we are 
talking about a cost of, administrative cost of only about 30 
basis points; is that right?
    Mr. LOCKHART. Yes. That is the ultimate cost that the 
actuaries forecast. That is on a relatively simple system: if 
you add more choice over time, it may go up somewhat in cost. 
Certainly the idea of letting people come into the system over 
time as they become more knowledgeable is probably a reasonably 
good idea. Going in and out, as I think Barbara mentioned, 
could be a paperwork problem.
    Mr. HULSHOF. Just as a final point, Mr. Chairman, again, 
for those in this education process, 30 basis points is 30 
cents for every $100 in the account; is that, in essence, 30 
basis points?
    Mr. LOCKHART. That is correct.
    Mr. HULSHOF. That would be the cost? Thank you, Mr. 
Chairman.
    Chairman MCCRERY. Thank you, Mr. Hulshof. Mr. Rangel. I am 
going to recognize our distinguished Ranking Member of the full 
Committee who has joined us. Welcome. You may use your 5 
minutes, Mr. Rangel, to pose questions to the panel, or if you 
would like to give us all a preview of the Ryan-Rangel bill, we 
would like to hear it. You are recognized.
    Mr. RANGEL. I really appreciate this courtesy. I do enjoy 
talking with Mr. Ryan because we really talk about our 
differences, and I think that if we did more of that, the 
Committee would better understand each other and we could take 
the political questions out of it. I want to deal with Social 
Security to get a clear understanding as to the new roads that 
we follow. We had presidential bills, then we had concepts, 
then we had solvency, then we had personal accounts and private 
accounts. As you understand this process, the Social Security 
system in its present form is going to have a fiscal problem in 
the future; is that correct?
    Mr. LOCKHART. That is correct, it is unsustainable.
    Mr. RANGEL. The President had the courage to take this 
complicated political issue, to bring it to the Congress and 
say that we have to fix it.
    Mr. LOCKHART. Yes, he did, and he brought it to the 
American people as well. I think they have started to get the 
message that we really do need to do something about Social 
Security, and do it soon.
    Mr. RANGEL. I really think you are going beyond your 
training in your subjective thoughts, and I think it would 
depend on what community you lived in, but I will leave that 
alone. Having said that, we are now dealing with the surplus 
that we now have, part of that being put into a private 
accounts, this new idea that we have come up with.
    Mr. LOCKHART. Yes, that is what I understand.
    Mr. RANGEL. Could you tell me how that deals with the 
question of solvency, which you believe that Americans are 
concerned about?
    Mr. LOCKHART. Well, certainly President Bush and the SSA 
has set as a goal achieving sustainable solvency, permanent 
solvency. I think the Senate in mid March voted----
    Mr. RANGEL. I know that, sir. I am just talking about the 
concept of the surplus that we get, beyond what we are paying 
the people. Assuming that when the Baby Boomers come, we won't 
have that surplus, right?
    Mr. LOCKHART. That is correct.
    Mr. RANGEL. That is what makes the crisis.
    Mr. LOCKHART. In three short years, the annual cash flow 
surplus begins to decline.
    Mr. RANGEL. All I am asking is, how does taking money out 
of a non-existing surplus, what happens when the surplus goes 
and you are trying to put a different type of bond in the 
private account than you have in the so-called trust fund, how 
does this concept--assuming that I bought the Ryan-Rangel 
concept, how do I explain that I am taking away the problem of 
solvency?
    Mr. LOCKHART. Well, I am not sure I can comment on the 
legislation that was discussed yesterday, but I can talk about 
personal accounts in general if that would help.
    Mr. RANGEL. No, please don't do that because every week 
they come up with a new concept, and as soon as I understand it 
they have another one. You are the expert, I want you to 
comment the best you can, because Mr. Ryan understands and he 
has left, and I only have you.
    Chairman MCCRERY. Would the gentleman yield?
    Mr. RANGEL. Yes.
    Chairman MCCRERY. I am also to be an author of the 
legislation that is not yet fully drafted, but I would be glad 
to discuss it with the gentleman. The gentleman has not yet 
characterized it correctly, and Mr. Lockhart is in no position 
to comment on the effect on solvency on a proposition that he 
hasn't seen.
    Mr. RANGEL. Well, I haven't seen it either.
    Chairman MCCRERY. I will tell the gentleman that we expect 
the Social Security actuaries to score the bill, the bill we 
are about to introduce, as increasing solvency by 2 years.
    Mr. LEVIN. Would the gentleman yield? If Mr. Rangel would 
yield?
    Mr. RANGEL. This is just such a good answer----
    Chairman MCCRERY. It is the gentleman from New York's time.
    Mr. RANGEL. If the Chairman would explain, this surplus is 
going to go the way that we enjoy now; is that correct? When 
the Baby Boomers come?
    Chairman MCCRERY. Yes, sir.
    Mr. RANGEL. You announced--someone is saying that out of 
the surplus, we are going to take the money to fund the private 
accounts, right?
    Chairman MCCRERY. We are going to put the equivalent amount 
of money of the surplus in personal accounts, in marketable 
securities, which will be in the name of individuals, yes, sir.
    Mr. RANGEL. My question, Mr. Chairman, is what surplus?
    Chairman MCCRERY. The surplus of revenues over outgo of the 
Social Security system. Cash.
    Mr. RANGEL. Is it not true that the money that goes into 
the surplus is going into the General Fund?
    Chairman MCCRERY. Yes, sir.
    Mr. RANGEL. If you take that surplus and put it into a 
private account, don't you leave a gap out there for general 
funds?
    Chairman MCCRERY. It depends on how you do that. In our 
legislation that is not the way we intend to do it, and I will 
be glad to discuss it fully with the gentleman so he 
understands it after the hearing. We will have witnesses on the 
next panel that can expound on various alternatives for doing 
what the gentleman has described. I look forward to questioning 
them when they get to the panel.
    Mr. RANGEL. I thank you.
    Mr. LEVIN. Will the gentleman yield briefly?
    Mr. RANGEL. I hope the Chairman would allow me, yes.
    Mr. LEVIN. Mr. Shaw indicated there would be two bonds. You 
asked a good question, how can you transfer the surplus from 
one place to another and address solvency. I am anxious to see 
the details, because the only way to do that is to use general 
funds. We will see. I think it was interesting, Mr. Rangel, 
that Mr. Shaw gave you an answer we were going to have two 
bonds. Let's see the details. The fact remains that FICA taxes 
are being used for private accounts. That is the basic fact. 
Thank you, Mr. Rangel, for yielding.
    Mr. RANGEL. Thank you.
    Chairman MCCRERY. Under our legislation, that will not be 
the basic fact. Mr. Lewis.
    Mr. LEWIS. Thank you, Mr. Chairman. Mr. Lockhart, right 
now, who owns the Social Security Trust Fund? Is there a 
personal account? Can a Social Security recipient say, this 
money, this Social Security money, is mine, and I own it, it is 
owed to me, and the Congress has to make sure that that money 
is in my personal account?
    Mr. LOCKHART. There are no personal accounts in Social 
Security today. Some people sometimes think there are, but 
there are none. In fact, it is a pay-as-you-go system, so the 
taxes I pay go to pay my parents' Social Security benefits.
    Mr. LEWIS. Absolutely. We rehash this over and over again, 
but since we are talking about personal accounts today, 
retirees do not own their Social Security money.
    Mr. LOCKHART. That is correct, and as you know, by law, 
Congress can change even the benefits under Social Security.
    Mr. LEWIS. Any time. Any Congress. It is not secure. It is 
only secure as long as Congress is willing to make sure that 
that pay-as-you-go system continues.
    Mr. LOCKHART. Correct.
    Mr. LEWIS. So, every time I have a townhall meeting, I have 
people say to me, when are you going to stop spending money out 
of the Social Security Trust Fund and set it aside for the 
Social Security recipients? Well, what we are talking about 
here today is what we are creating, a system now where the 
money will be set aside, the surplus money will be set aside, 
and the recipients will be given a key to the box, to the 
lockbox, and it will not be in the hand of Congress, its key 
will be in the hand of the recipient, of the constituent. They 
have been asking for that for a long time.
    Mr. LOCKHART. I agree with you. I have probably done 50 or 
more townhalls with Members over the last year or so--with 
Democratic and Republican Members, with the AARP and the 
American Federation of Labor and Congress of Industrial 
Organizations (AFL/CIO). That is one of the most common 
questions you get: Why isn't that money staying in Social 
Security?
    Mr. LEWIS. Exactly. So, going back to Mr. Becerra's 
question a minute ago, again, we are rehashing this, but when 
will the Social Security Trust Fund surplus hit the wall and 
start on the road to insolvency?
    Mr. LOCKHART. In 3 years the annual cash flow surplus 
begins to decline. That is when the first Baby Boomers start to 
retire. In 12 years there will not be enough taxes to pay 
benefits. That is when we have to start drawing down the 
interest on the bonds.
    Mr. LEWIS. As Mr. Becerra asked you, will those be paid?
    Mr. LOCKHART. I believe they will, yes, sir.
    Mr. LEWIS. How will they be paid?
    Mr. LOCKHART. There are really only three ways: increasing 
taxes, borrowing the money elsewhere, or reducing government 
spending.
    Mr. LEWIS. What kind of tax increase will our children have 
to bear to pay this? It was a great deal when there were 40 
people paying in for one person on retirement. Now that it is 
down to three for one, and eventually two for one, the pyramid 
is coming to a point here. What kind of taxes are my kids going 
to have to bear?
    Mr. LOCKHART. Well, maybe your grandchildren. Over the 75-
year forecast, the independent actuaries' numbers and the 
trustees' Report shows we will have to increase payroll taxes 
by 46 percent by the end of the period, if you wanted to 
continue to pay scheduled benefits. If you don't want to 
continue to pay scheduled benefits, you would not have to.
    Mr. LEWIS. We are going to have, as Mr. Becerra said, those 
benefits paid. So, it seems to me like my colleagues across the 
aisle, they have a plan, and the plan is this: We are just 
going to tax the daylights out of future generations to meet 
the obligations for Social Security. That sounds like a pretty 
bad plan to me, and a very risky plan. I yield back my time. 
Thank you.
    Chairman MCCRERY. Thank you, Mr. Lewis. Mr. Lockhart, Ms. 
Bovbjerg, thank you very much for joining us. We look forward 
to seeing you again. At this time I call the second panel. We 
have another distinguished panel to give us their views on 
personal accounts and how they might be set up and managed. 
Olivia S. Mitchell, Ph.D., Professor of Insurance and Risk 
Management, Executive Director, Pension Research Council, and 
Director, Boettner Center for Pensions and Retirement Security, 
Wharton School, University of Pennsylvania. Patrick J. Purcell, 
Specialist in Social Legislation, Domestic Social Policy 
Division, Congressional Research Service (CRS). Alex J. 
Pollock, resident fellow of the American Enterprise Institute. 
Joan Entmacher, vice President and Director, Family Economic 
Security, National Women's Law Center (NWLC). Francis X. 
Cavanaugh, former Executive Director and chief executive 
officer, Federal Retirement Thrift Investment Board. Virginia 
Reno, Vice President for Income and Security policy, National 
Academy of Social Insurance (NASI).
    We welcome all of you for our hearing today. Your written 
testimony will be included in the record in its entirety, and 
we would ask you to try to summarize your written testimony in 
about 5 minutes. The little device you see in the center of the 
witness table, and likewise the little box up here has three 
lights. The green light lasts 4 minutes, the amber light lasts 
1 minute, and the red light should not last very long. With 
that, we will begin with Dr. Mitchell.

STATEMENT OF OLIVIA S. MITCHELL, PH.D., PROFESSOR OF INSURANCE 
   AND RISK MANAGEMENT, EXECUTIVE DIRECTOR, PENSION RESEARCH 
    COUNCIL, AND DIRECTOR, BOETTNER CENTER FOR PENSIONS AND 
      RETIREMENT SECURITY, WHARTON SCHOOL, UNIVERSITY OF 
            PENNSYLVANIA, PHILADELPHIA, PENNSYLVANIA

    Dr. MITCHELL. Thank you very much, Mr. Chairman and Members 
of the Subcommittee. It is a pleasure to appear before you 
today. My name is Olivia Mitchell, and I teach insurance and 
risk management at the Wharton School at the University of 
Pennsylvania. The views that I am expressing today are my own. 
As we have heard, the Social Security system is running into 
trouble with a shortage of revenue to pay benefits within just 
a few short years. The 2001 Bipartisan Commission to Strengthen 
Social Security, on which I served, believed in that having two 
separate tiers would be the answer for a reformed Social 
Security program. Social adequacy would be provided by the 
first pillar of the traditional Social Security program, while 
individual equity would be the goal of the personal accounts 
component.
    My testimony today focuses on two aspects of a personal 
retirement account, namely, first, administrative fees and 
charges; and, second, payout issues. What I will hope to 
persuade you of is that voluntary personal retirement accounts 
can and should be formulated so they offer participants some 
investment choice, while still remaining relatively 
inexpensive, they standardize disclosure regarding fees and 
charges so participants can understand what it is they are 
confronting, and require retirees to annuitize a portion of 
their retirement assets so the combined benefit payments of 
Social Security, the traditional piece, and the personal 
retirement accounts will keep them out of poverty.
    Regarding administrative fees and charges, I would like to 
make four points. First of all, measuring pension expenses is a 
tricky business. I believe that a standardized format for 
reporting fees and charges would greatly enhance participants' 
ability to compare products and make informed decisions. 
Second, scale is very important. Large-scale plans, such as my 
university's retirement plan, and the Federal TSP we have heard 
about today, charge participants very low annual fees. Bigger 
is cheaper. Third, private retirement systems are not 
necessarily more expensive than Social Security. Rather, they 
generally offer many more services. I believe privately managed 
competitive fund providers can do better by taking advantage of 
modern technology. Fourth, the Commission on which I served 
proposed that investment choices in the personal accounts 
should be limited to a few indexed funds. We talked about a 
stock index fund, and a bond index fund. One of my personal 
favorites is the Treasury Inflation-Protected Securities (TIPS) 
Fund, which I think has a very crucial role in the retirement 
portfolio, and a life-cycle fund would also be a useful 
addition.
    As you know from speaking to the prior panel, Employee 
Retirement Income Security Act (P.L. 93-406) requires a minimum 
of three funds in the private sector, so, some number between 
three and five seems reasonable to me. The Office of the Chief 
Actuary of Social Security estimated that the personal 
retirement account model we proposed on the Commission would be 
quite inexpensive, costing only about 0.3 percent of assets 
annually to manage.
    The next issue to which I wish to turn is the issue of 
payouts; that is, how older participants in a personal account 
system would access their funds at retirement. The related 
question is what role annuities might play in such a payout 
scheme. Annuities are, of course, financial products that 
protect people from outliving their retirement assets. It is 
very important, I would note, that the money's worth of many of 
these life annuity products is very attractive in the United 
States and abroad. With regard to annuitization, my Commission 
proposed that partial annuitization should be required so that 
the yearly income received from the traditional Social Security 
pillar plus the joint annuity, if the person was married, would 
protect either spouse from falling below poverty in retirement. 
So, assets above what would be needed to achieve this poverty 
protection could be accessed as a lump sum. My written 
testimony discusses how retirees would learn about annuity 
products, who would sell them, and so forth. I will only 
mention a couple of points here. First of all, what annuity 
products would be offered and to whom? I believe Congress 
should set a default payout scheme such as a joint-survivor, 
ideally inflation-indexed, annuity which retirees would 
automatically get unless they opted for something else.
    Second of all, I believe private insurers can offer the 
types of products that retirees want, but there needs to be 
enough oversight in this market to make sure that there is not 
cherry-picking of just the rich retirees, or perhaps just the 
people who are going to die soon, for the annuity market. I 
would not support having the Federal Government sell the 
mandatory annuities under the new system. A couple other points 
bear mention, which we hopefully will hear more about today: 
the government needs to think carefully about tax and transfer 
policy regarding these personal accounts; and, second of all, 
the Federal Government has a crucial role in making sure there 
are enough assets for annuity providers to purchase, so they 
can offer these inflation-indexed lifetime benefits.
    In conclusion, I believe voluntary personal accounts can be 
designed to provide participants with investment choice while 
remaining inexpensive. They can build in incentives for 
competition among fund managers, and they can sensibly require 
retirees to annuitize a portion of their retirement assets. 
Thank you for your interest. I am happy to answer any 
questions.
    [The prepared statement of Dr. Mitchell follows:]

Statement of Olivia S. Mitchell, Ph.D., \1\ International Foundation of 
 Employee Benefit Plans Professor and Professor of Insurance and Risk 
Management, Executive Director, Pension Research Council, and Director, 
 Boettner Center for Pensions and Retirement Security, Wharton School, 
         University of Pennsylvania, Philadelphia, Pennsylvania
---------------------------------------------------------------------------
    \1\ Olivia S. Mitchell is the International Foundation of Employee 
Benefit Plans Professor and Professor of Insurance and Risk Management; 
she is also Executive Director, Pension Research Council and Director, 
Boettner Center for Pensions and Retirement Security, all at The 
Wharton School of the University of Pennsylvania (3620 Locust Walk, St 
3000 SH-DH, Philadelphia, PA 19104; email [email protected]; T 
215-898-0424). The views offered here are solely those of the author 
and do not represent those of any institutions with which she is 
affiliated.
---------------------------------------------------------------------------
    Mr. Chairman and members of the Subcommittee: Thank you for the 
opportunity to appear here today. My name is Olivia S. Mitchell, and I 
am a Professor of Insurance and Risk Management at The Wharton School 
at the University of Pennsylvania.
    As you know, Social Security faces imminent insolvency, with 
payroll tax revenues threatening to fall below benefit payments within 
6 years. The present system also contains many inequities and anomalous 
redistribution patterns, and it offers current workers a surprisingly 
low and very risky return. \2\
---------------------------------------------------------------------------
    \2\ Cogan and Mitchell (2002).
---------------------------------------------------------------------------
    The bipartisan Commission to Strengthen Social Security (CSSS), on 
which I served in 2001, believed that offering two separate tiers under 
a reformed Social Security program, each with its own function, would 
improve the overall program's transparency and equity. Social adequacy 
was to be the principal objective of the traditional defined benefit 
piece, while individual equity was seen as the goal of a personal 
accounts component.
    My testimony before this Subcommittee today focuses on two aspects 
that must evaluated in designing a Personal Retirement Account element 
as part of a reformed Social Security: (1) administrative fees and 
charges, and (2) payout issues. My views derive from the research 
literature on administrative fees and payout issues, particularly 
regarding how Personal Retirement Accounts might be invested and how 
the funds at retirement might be deployed. The views I offer are my own 
and do not represent those of any institutions with which I am 
affiliated.
    My conclusions are that the voluntary Personal Retirement Accounts 
(PRAs) should be formulated so that:

      They offer participants some investment choice while 
still being relatively inexpensive;
      They standardize disclosure regarding fees and charges so 
participants can understand and compare them;
      They require retirees to annuitize part of their 
retirement assets in their Personal Accounts, so that the combined 
benefit payments from Social Security will keep them out of poverty.

Administrative Fees and Charges
    Experience with public and private pension plans the world over 
indicates wide disparity in reported administrative fees and charges 
across systems. Several lessons are worth highlighting:
      Measuring pension expenses requires standardized 
reporting and disclosure standards. Pension systems often structure 
their charges in bewildering ways. For instance, fees can be levied as 
flat commissions, a percent of contributions, or a percent of the 
fund's annual yield. \3\ Such complexity makes it difficult for plan 
participants to compare fund performance. A sensible response, adopted 
by many Latin American pension supervisors, is to require disclosure 
using a standardized table for reporting charges. This has the effect 
of increasing the information available to participants and hence, 
making the market more competitive. A more problematic tactic adopted 
by the UK, for example, is to set a national fee cap. This may limit 
competition and reduce participants' focus on holding down costs.
---------------------------------------------------------------------------
    \3\ Mitchell (1998).
---------------------------------------------------------------------------
      Scale is important in keeping costs down. \4\ Larger 
money managers benefit from scale economies, centralized fund 
administration, and centralized collection of contributions. For 
example, in Australia, retail financial service providers charge three 
times more in pension fees and charges than do institutional managers 
of corporate pensions. While there is little agreement on the minimum 
size of a cost-effective pension, managers of large defined 
contribution plans such as the Federal Thrift Savings Plan which covers 
civil servants and military employees, my University's retirement plan 
(TIAA-CREF and Vanguard), and others, charge pension participants 
annual fees between 0.1-0.4% of assets under management. These fees are 
well below what savers pay in typical Individual Retirement Accounts.
---------------------------------------------------------------------------
    \4\ See Mitchell (1998), Bateman and Mitchell (2004), and 
Whitehouse (2005).
---------------------------------------------------------------------------
      Private retirement systems might seem to be more costly 
than Social Security, but this is a misleading conclusion as they 
generally offer more and different services. Some have suggested that 
the current U.S. Social Security system is one of the lowest-cost 
programs around. Nevertheless, Social Security does not provide the 
wide range of services provided by modern managers of asset-backed 
retirement accounts. For instance, the government program does not 
invest in the capital market, it holds no insurance-type reserves even 
though it offers disability and survivors' insurance, and it takes a 
very long time--more than a year--to post workers' contributions to 
their records. \5\ By contrast, privately managed fund providers would 
and can do better by taking advantage of modern technology.
---------------------------------------------------------------------------
    \5\ Mitchell (1998).

    Taking these and other factors into account, I and other Commission 
members concluded that it would be reasonable to establish personal 
accounts along the lines of the Federal Thrift Saving Plan. 
Accordingly, and for a few years into the system, a central Governing 
Board would be charged with collecting contributions, managing records, 
and selecting private-sector managers who would invest participant 
assets via a competitive bidding process. This Board could either 
handle record-keeping and benefit payments itself, or these functions 
could be outsourced via a competitive process.
    We also proposed that investment choices in the personal accounts 
would be limited but diverse. The options suggested include:

      a Government Securities Investment fund (mainly short-
term U.S. Treasury securities);
      a Fixed Income Index Investment fund (tracking a U.S. 
bond market index);
      a Common Stock Index Investment fund (tracking the 
Standard & Poor's 500 Index of large-company stock);
      a Small Capitalization Stock Index Investment fund 
(tracking the Wilshire 4500 stock index); and
      an International Stock Index Investment fund; and
      a fund that invests in Government Treasury Inflation-
Protected Securities.

    At some later date, plan participants might be permitted to move 
their investments to licensed, supervised, private money managers 
offering an approved set of low-cost investment options. The benefit 
levels that might be expected from alternative investment approaches 
for Personal Retirement Accounts appear in Table 1, along with a 
comparison of current benefits, payable benefits, and scheduled 
benefits.
    The Office of the Chief Actuary at Social Security estimated that 
the proposed CSSS approach would be quite inexpensive, costing only 
about 0.3% of assets annually.

  Table 1: Monthly Social Security Benefits Under Alternative Scenarios
                 Projected to 2052 (CSSS  Model  2  $01)



I. Lifetime low-wage earner\*\
Today's benefit                               $637
Projected Benefit With Personal Account:
      Low yield                               867
      Medium yield                            1,050
      High yield                              1090
Current Program Payable                       713
Scheduled benefit                             986
II. Lifetime medium-wage earner\*\
Today's benefit                               $1,052
Projected Benefit With Personal Account:
      Low yield                               1,204
      Medium yield                            1,525
      High yield                              1,595
Current Program Payable                       1,179
Scheduled benefit                             1,628
III. Lifetime maximum-wage earner\*\
Today's benefit                               1,366
Projected Benefit With Personal Account:
      Low yield                               1,565
      Medium yield                            1,907
      High yield                              1,983
Current Program Payable                       1,557
Scheduled benefit under current law           2,151

\*\ These categories, developed by Social Security actuaries, are
  specified (in $01) such that a lifetime ``low'' earner would have
  averaged approximately $15,900 per year, whereas the medium earner
  averaged $35,300 per annum and the high earner $56,400.

Source: Cogan and Mitchell (2003)

Payout Issues
    When considering how to structure payouts from voluntary Personal 
Accounts under a reformed Social Security system, naturally the 
question arises as to whether and how access to the funds should be 
permitted. CSSS members agreed that pre-retirement access to the money 
should not be allowed to `leak' out before retirement, as early 
consumption would likely increase the chances that the elderly would 
then have to rely on old-age antipoverty programs. Yet, as the 
Commission pointed out, ``a clear appeal of personal retirement 
accounts is that they grant workers ownership over their own assets.'' 
After weighing competing arguments, we concluded that personal accounts 
should be preserved until the nationally-agreed on early retirement 
age, consistent with current Social Security policy which does not 
permit pre-retirement access to old-age benefits.
    By contrast there is more discussion regarding appropriate designs 
for the pension decumulation process under Personal Accounts. This 
refers to the process by which older participants access their 
retirement assets, how they invest their money during retirement, and 
whether annuities--which are financial products designed to cover the 
risk of retirees outliving their assets--should play a central role. 
Regarding post-retirement fund management, my Commission recommended 
several methods of drawdown including phased withdrawals and annuities, 
as well as possibly lump sums.
    To highlight the importance of longevity risk, Table 2 shows that a 
65-year-old U.S. male can anticipate living to age 81, but he has 
almost a 20% chance of living to age 90 or beyond. A woman of the same 
age can expect to live to 85, but she has more than a 30% chance of 
living to age 90 or older (Table 2). In other words, people face 
substantial risk of outliving their life expectancy, implying 
substantial uncertainty regarding how long one must conserve and spend 
retirement assets, combined with a high probability of running out of 
money.

   Table 2: Remaining Life Expectancy and Survival at Age 65 (in 2000)

                     Remaining Life                         Men    Women

Expectancy (years):                                        16.4    19.6
Probility of Surviving to Age:
         70                                                 88%     92%
         75                                                74      82
         80                                                56      69
         85                                                36      51
         90                                                18      31
         95                                                 6      14
        100                                                 1       4


Source: Mitchell and McCarthy (2004)


    A life-long annuity can help protect against this risk, by paying a 
premium to an insurer who then pools a number of people with similar 
longevity expectations. Though some have argued that such insured 
products seem expensive, my research shows that the ``money's worth'' 
(MW) of such life income products is rather substantial. The MW refers 
to the discounted cash flow of the lifetime payments received divided 
by the product premium. For example, Table 3 shows that U.S. purchasers 
of an immediate single-life annuity would expect back 93 cents on the 
dollar from a life annuity; in exchange purchasers have the insurance 
value that they will never outline their lifetime benefit payments. The 
MW ratios are similar in Australia, Italy, and the UK.

       Table 3. Money's Worth of Single Preium Nominal Life Annutities for 65-Year-Olds: An International
                    Comparison(using country Treasury yield curves and annuitant life tables)

                                                  Australia      Canada       Italy          UK           US

Men                                                   0.986        1.014        0.958        0.966        0.927

Women                                                 0.970        1.015        0.965        0.957        0.927



    Source: Derived from Mitchell and McCarthy (2004)

    These issues are complex and potentially politically delicate, 
since some workers will fail to accumulate much in their accounts over 
their worklives; also some retirees might anticipate relatively lower-
than-average life expectancies, making forced annuitization seem 
punitive.
    In balancing the various choices for payout design, the Commission 
concluded that partial annuitization should be mandated so that ``the 
yearly income received from an individual's Social Security benefit 
plus the joint annuity (if married) would protect either spouse from 
falling below the poverty line during retirement'' (CSSS 2001). Any 
funds above those needed to buy the minimum annuity could be accessible 
as a lump sum and/or bequeathed at death. This approach has the dual 
benefit of both protecting the retiree from falling below the poverty 
line while still allowing some access to the funds accumulated in the 
Personal Retirement Account.
    Remaining design issues include how retirees would learn about 
annuity products, who would sell them, and whether the private 
insurance market can do a good job meeting market demand. To date, 
relatively few consumers have purchased payout annuities, making it a 
bit difficult to forecast how the market will develop. Several key 
issues will have to be decided:

      Which annuity products will be offered and to whom?

          Currently private insurers in the U.S. offer a wide and very 
        complex array of annuity products, including immediate versus 
        deferred benefit payments; fixed nominal payouts versus 
        programs with escalating or variable payouts; and term certain 
        versus other payment periods. Also annuities offered through 
        company pensions are mandated to use unisex mortality tables 
        whereas retail annuities do not.
          A logical lesson from the behavioral finance literature is 
        that it would be sensible to establish a ``default'' payout 
        format such as a joint and survivor inflation-linked or 
        escalating life annuity, which retirees would automatically 
        receive unless they specifically opted for something else. As a 
        case in point, retirees in the UK are required to annuitize 
        their pension assets at age 75; in Germany, workers with assets 
        in so-called Reister-pensions may take 20% of their accumulated 
        assets in a lump sum, another 20% in a phased withdrawal 
        format; but at age 85, the retiree must annuitize his balance 
        and the benefit may not be lower than the periodic payment 
        received before that age.
          Of course, since many retirees are not accustomed to thinking 
        about longevity risk, they would require financial education to 
        help them clearly understand the costs and benefits of 
        different ways to manage their Personal Retirement Account 
        assets.

      Which annuity providers will be allowed in the market, 
and how will they be regulated?

          Evidence from other countries adopting personal accounts 
        indicates that private insurers can and do offer the types of 
        products that retirees want. For instance, in Chile, middle and 
        upper income workers generally prefer the annuity payout over a 
        phased withdrawal approach to retirement drawdowns.
          Nevertheless, there will likely have to be some governmental 
        oversight over the annuity market. In Mexico, for instance, all 
        insurers are required to bid on all retirees, and when issuing 
        annuity bids, the companies may learn only a retiring worker's 
        age and sex (but not his identity, his health status, or his 
        account balance). This reduces the chances of ``cherry-
        picking'' rich retirees or those anticipated to die soon.
          Another issue has to do with whether unisex mortality tables 
        would be required for the annuities. Doing so, of course, 
        involves redistribution of wealth away from shorter-lived men 
        and toward longer-lived women, which is already true in the 
        current Social Security System. Requiring joint and survivor 
        benefits as a default would render this issue less important 
        quantitatively.

      What role, if any, would the federal government have?

          As an alternative to building up private annuity markets, 
        some have suggested that the federal government might directly 
        sell the mandatory annuities under the new system. \6\ While 
        this might hold down some costs, it can cause other problems. 
        For example, there could be political interference associated 
        with investing the annuity reserves--amounting to 15% of GDP at 
        maturity--and it raises questions about whether the reserves 
        could truly be saved, or whether they would be `spent' akin to 
        Social Security Trust Fund assets. Further, the government 
        would then have responsibility for mortality and capital market 
        risk, which would likely be incorrectly priced and managed.
---------------------------------------------------------------------------
    \6\ NASI (2005).
---------------------------------------------------------------------------
          One key role for the federal government in this context has 
        to do with tax and transfer policy. For instance, pension and 
        Individual Retirement Account assets are protected in 
        bankruptcy but are divisible in divorce; whether the same 
        treatment would be afforded PRA annuities and assets has yet to 
        be determined. Conversely, annuity flows and lump sums are 
        generally `counted' when retirees apply for SSI and Medicaid 
        benefits; payouts are taxed as income. Whether and how PRA 
        assets and annuities are to be treated for tax and transfer 
        purposes--as well as others (e.g. the estate tax vulnerability 
        of the PRA assets if the worker or spouse dies) will take 
        additional work to get it right.
          Another role for the government is to enhance the range of 
        investments available to insurers providing the products. \7\ 
        Many writers have noted the key role of federal government 
        provision of inflation-indexed bonds sufficient to meet market 
        demand. Expanding their supply would allow private insurers to 
        offer the kinds of indexed annuity products that would give 
        retirees better protection against inflation, which is a source 
        of substantial retirement insecurity.
---------------------------------------------------------------------------
    \7\ Bodie et al. (2002).
---------------------------------------------------------------------------
Conclusions
    My testimony has focused on the role of administrative fees and 
charges in a PRA type approach, and also on payout considerations after 
retirement. I conclude that voluntary Personal Retirement Accounts can 
be designed so as to provide participants with some investment choice 
while still being relatively inexpensive; they can build in incentives 
for competition among fund managers, including disclosure regarding 
fees and charges; and they can sensibly require retirees to annuitize 
part of their retirement assets in their Personal Accounts, so that the 
combined benefit payments will keep them out of poverty.
    Thank you for your interest and I am happy to answer any questions 
you may have about my remarks.

References
    Bateman, H. & O.S. Mitchell. ``New Evidence on Pension Plan Design 
and Administrative Expenses.'' Journal of Pension Finance and 
Economics. 2004: Vol 3(1): 63-76.
    Bodie, Z., B. Hammond, and O.S. Mitchell, eds. Innovations in 
Financing Retirement. Philadelphia, PA: University of Pennsylvania 
Press, 2002.
    Brown, J.R., O.S. Mitchell, J.M. Poterba. ``The Role of Real 
Annuities and Indexed Bonds in an Individual Accounts Retirement 
Program.'' In Risk Aspects of Investment-Based Social Security Reform. 
Eds. J. Campbell and M. Feldstein. 2000: 321-360.
    Brown, J., O.S. Mitchell, J. Poterba, and M. Warshawsky. The Role 
of Annuity Markets in Financing Retirement. MIT Press, 2001.
    Commission to Strengthen Social Security (CSSS), Strengthening 
Social Security and Creating Personal Wealth for all Americans, Final 
Report, Washington, D.C., December 2001.
    Cogan, J.F. & O.S. Mitchell. ``Perspectives from the President's 
Commission on Social Security Reform.'' Journal of Economic 
Perspectives. 17(2). Spring 2003.
    Mitchell, O.S. ``Administrative Costs of Public and Private Pension 
Plans''. In Privatizing Social Security, Ed. M. Feldstein. NBER. 
Chicago: University of Chicago Press, 1998: 403-456.
    Mitchell, Olivia S. & David McCarthy. ``Annuities for an Ageing 
World''. In Developing an Annuities Market in Europe. Eds. E. Fornero & 
E. Luciano. Elgar, 2004: 19-68.
    NASI Uncharted Waters: Final Report. http://www.nasi.org/info-
url_nocat2718/info-url_nocat_show.htm?doc_id=212573
    Whitehouse, E. Testimony Before the Subcommittee on Social Security 
of the House Committee on Ways and Means, Washington, D.C. June 16, 
2005.

                                 

    Chairman MCCRERY. Thank you, Dr. Mitchell. Somehow the 
witnesses got scrambled on the table. I am going to follow my 
list as I introduced the witnesses, if that is okay. You will 
all get to speak. Our next witness is Mr. Purcell with the 
Congressional Research Service.

     STATEMENT OF PATRICK J. PURCELL, SPECIALIST IN SOCIAL 
  LEGISLATION, DOMESTIC SOCIAL POLICY DIVISION, CONGRESSIONAL 
                        RESEARCH SERVICE

    Mr. PURCELL. Mr. Chairman, Congressman Levin, Members of 
the Subcommittee, my name is Patrick Purcell. I am a specialist 
in pension issues with the CRS. Thank you for inviting me to 
speak to you today about the TSP for Federal employees. The 
thrift plan, as you know, is a savings plan for Federal workers 
and members of the uniformed services. It was first authorized 
by the Congress with the ERISA 1986. The thrift plan provides 
Federal employees and members of the uniformed services with a 
tax-deferred savings vehicle similar to those provided by many 
employers in the private sector under section 401(k) of the 
Internal Revenue Code. The thrift plan was designed by Congress 
to be a key part of the retirement benefits for employees who 
are covered by the Federal Employees Retirement System, which 
covers all Federal workers hired since 1984.
    Prior to enactment of the Social Security amendments 1983 
(P.L. 98-21), Federal employees were not covered by Social 
Security. They were instead covered by a separate system, the 
Civil Service Retirement System (CSRS). The Social Security 
System needed additional cash contributions to remain solvent, 
and the 1983 amendments mandated coverage for civilian 
employees hired in 1984 and later. Congress recognized at that 
time that Social Security provided some of the same benefits 
for retirement and disability as the Civil Service Retirement 
System. Moreover, enrolling workers in both plans would have 
required payroll deductions equal to more than 13 percent of 
each employee's pay.
    Consequently, Congress directed the development of a new 
Federal employee retirement system with Social Security as the 
cornerstone, and which would incorporate many features of the 
retirement plans typical among large employers in the private 
sector. The result of this effort was the Federal Employees 
Retirement System (FERS) which consists of three elements: 
Social Security, a traditional pension called the FERS Basic 
Retirement Annuity, and the TSP. The legislative history of the 
TSP indicates that in designing the system, Congress had the 
goals of incorporating Social Security into Federal employee 
retirement, providing a total benefit that was comparable to 
that under the old CSRS, and also keeping costs to the Federal 
Government approximately the same. Congress also for the first 
time allowed employees the opportunity to save for retirement 
on a tax-deferred basis through the TSP.
    In the legislative history of the TSP, two things stand 
out. First, Congress chose then, and has maintained to this 
day, a system in which all of the thrift funds that invest in 
private sector securities are index funds. This was a carefully 
considered choice. As the House Committee report on the 
legislation stated, ``The three funds authorized in the 
legislation are passively managed funds, not subject to 
political manipulation. A great deal of concern was raised 
about the possibility of political manipulation of large pools 
of thrift plan money. This legislation was designed to preclude 
that possibility.'' Likewise, the Senate Committee report on 
the legislation stated, ``Another concern the Committee 
wrestled with was the potential for market manipulation through 
political pressure. The Committee specifically designed the 
plan to avoid this problem. The legislation provides for three 
investment funds that are essentially self-managed.''
    The second item that stands out in the legislative history 
is the strong interest that Congress showed in establishing the 
independence and authority of the Federal Thrift Investment 
Board. The legislation established the Thrift Board as an 
independent government agency, which is required by law to 
operate the plan solely in the interest of plan participants. 
The law charges the Thrift Board with responsibility for 
developing the investment policies of the thrift plan, and 
overseeing the management of the plan. The law authorizes the 
Board to appoint an Executive Director who runs the thrift plan 
on a day-to-day basis. Three members of the Board, including 
the Chairman, are appointed by the President. The President 
chooses a fourth member in consultation with the Speaker of the 
House and the House Minority leader, and a fifth member in 
consultation with the Senate Majority and Minority leaders. 
Members of the Board are subject to Senate confirmation and 
serve four-year terms. All members of the Board must have 
substantial experience in managing financial investments and 
pension plans.
    The Federal Thrift Board receives no appropriations from 
Congress. Administrative expenses are paid through agency 
contributions forfeited by employees who leave Federal service 
before they have investigated and by charges against 
participant accounts. Congress conducts oversight of the TSP 
through the House Committee on government Reform and the Senate 
Committee on Homeland Security and governmental Affairs. The 
TSP is a key component of Federal employees' benefits. It is an 
efficient provider of retirement savings accounts to the 
Federal workforce that has achieved high participation rates 
and low administrative costs. This concludes my statement. I 
would be happy to answer any questions.
    [The prepared statement of Mr. Purcell follows:]

  Statement of Patrick J. Purcell, Specialist in Social Legislation, 
    Domestic Social Policy Division, Congressional Research Service

    Mr. Chairman and members of the subcommittee, my name is Patrick 
Purcell and I am a specialist in pension issues with the Congressional 
Research Service. Thank you for inviting me to speak to you today about 
the Thrift Savings Plan for federal employees.
    The Thrift Savings Plan is a retirement savings plan for federal 
employees and members of the uniformed services. It was authorized by 
Congress in the Federal Employees' Retirement System Act of 1986 (P.L. 
99-335). The Thrift Plan provides federal employees and members of the 
uniformed services with a tax-deferred savings vehicle similar to those 
provided by many employers in the private sector under section 401(k) 
of the Internal Revenue Code. The Thrift Plan was designed by Congress 
to be a key part of the retirement benefits for employees who are 
covered by the Federal Employees' Retirement System (FERS), which 
covers all federal employees hired on or after January 1, 1984.

Origin of the Federal Employees' Retirement System
    Prior to enactment of the Social Security Amendments of 1983 (P.L. 
98-21), federal employees were not covered by Social Security. Federal 
employees were covered instead by the Civil Service Retirement System 
(CSRS). Because the Social Security system needed additional cash 
contributions to remain solvent, the 1983 amendments mandated coverage 
for civilian federal employees hired in 1984 or later.
    Congress recognized, however, that Social Security provided some of 
the same benefits as CSRS. Moreover, enrolling federal workers in both 
plans would have required payroll deductions equal to more than 13% of 
employee pay. Consequently, Congress directed the development of a new 
federal employee retirement system with Social Security as the 
cornerstone and which would incorporate many features of the retirement 
programs typical among large employers in the private sector. The 
result of this effort was the Federal Employees' Retirement System, or 
FERS. FERS consists of three elements: (1) Social Security, (2) a 
traditional pension called the FERS basic retirement annuity, and (3) 
the Thrift Savings Plan.
    The Thrift Plan is administered by an independent government 
agency, the Federal Retirement Thrift Investment Board, which is 
charged in statute with operating the Thrift Plan prudently and solely 
in the interest of the participants and their beneficiaries. \1\ The 
assets of the Thrift Plan are maintained in the Thrift Savings Fund, 
which invests the assets in accordance with participant instructions in 
five investment funds authorized by Congress to be included in the 
plan.
---------------------------------------------------------------------------
    \1\ See 5 U.S.C. Sec. 8472(h).
---------------------------------------------------------------------------
    Federal employees who participate in FERS, or its predecessor, the 
Civil Service Retirement System (``CSRS''), and members of the 
uniformed services are eligible to join the Thrift Plan immediately 
upon being hired. Generally, FERS employees are those employees hired 
on or after January 1, 1984, while CSRS employees are employees hired 
before January 1, 1984, who have not elected to convert to FERS. Each 
group has different rules that govern contribution rates.
    As of March 31, 2005, there were 3.4 million participants in the 
Thrift Plan, with approximately 2.5 million contributing to the plan. 
\2\ Among employees covered by FERS, 86% of those eligible to 
participate in the Thrift Plan do so. Among CSRS employees, about two-
thirds participate. Assets of the plan totaled $154 billion as of March 
31. In terms of both assets and number of participants, the Thrift 
Savings Plan is the largest employer-sponsored retirement savings plan 
in the United States.
---------------------------------------------------------------------------
    \2\ See Table 1 for complete Thrift Savings Plan enrollment 
statistics.
---------------------------------------------------------------------------
    The Thrift Plan is legally a ``defined contribution'' plan. This 
means that it specifies how much an employee may contribute and how 
much the employing agency must contribute to each FERS employee's 
account. The employee owns the account and his or her benefit is equal 
to the account balance, which can be taken as a lump-sum, an annuity, 
or a series of periodic withdrawals.
Contributions
    In 2005, FERS employees can contribute as much as 15 percent of 
basic pay on a tax-deferred basis, up to the $14,000 maximum specified 
in section 402(g) of the Internal Revenue Code. Participants in FERS 
are entitled to receive employer matching contributions on the first 
five percent of pay that they contribute to the Thrift Plan. \3\ 
Participants age 50 and older who are already contributing the maximum 
amount for which they are eligible are allowed to make supplemental 
tax-deferred ``catch-up'' contributions of up to $4,000 in 2005.
---------------------------------------------------------------------------
    \3\ The formula for agency matching contributions is specified in 
law at (5 U.S.C. Sec. 8432(c)).
---------------------------------------------------------------------------
    In 2005, CSRS employees and members of the uniformed services can 
contribute up to ten percent of basic pay on a tax-deferred basis, 
subject to the $14,000 maximum specified in the tax code. Members of 
the uniformed services also may contribute up to 100% of designated3 
special pay, incentive pay, and bonuses to the Thrift Plan. Neither 
CSRS participants nor members of the uniformed services receive 
employer matching contributions because both CSRS and the military 
services provide pension benefits to career employees and career 
military personnel that are substantially larger as a percentage of 
career-average pay than the FERS basic retirement annuity.
    All FERS participants receive from their employing agencies an 
automatic contribution equal to one percent of basic pay. \4\ 
Participants may also transfer funds from a traditional individual 
retirement accounts (IRA) or another eligible employer plan into the 
Thrift Plan.
---------------------------------------------------------------------------
    \4\ Basic pay is defined in statute at (5 U.S.C. Sec. 8401(4)).
---------------------------------------------------------------------------
Investment Options
    As provided for in statute, Thrift Plan participants are offered 
five investment funds. Participants may allocate their contributions 
among any or all of the five investment funds, and they may reallocate 
their account balance among the five investment funds. The four funds 
that invest in private-sector securities are all index funds. These 
funds purchase securities in the same proportion as they are 
represented in an index of stocks or bonds, rather than through the 
decisions of an investment manager. Index funds have lower 
administrative costs than actively-managed funds, and because they 
purchase securities in the same proportion as they are represented in 
an index, there is little or no opportunity for the purchase of 
securities by the fund to be influenced by third parties who might 
benefit from having the fund invest in particular companies or sectors 
of the economy.
    The five funds in the Thrift Plan are:

      the Government Securities Investment Fund, (the ``G 
Fund''). This fund invests exclusively in U.S. Treasury Securities and 
other securities backed by the full faith and credit of the United 
States. Over the period from 1988 through 2004, the ``G'' fund earned 
an average annual rate of return of 6.6%. \5\
---------------------------------------------------------------------------
    \5\ See Table 2 for annual rates of return from 1988 through 2004.
---------------------------------------------------------------------------
      the Fixed Income Investment Fund, (the ``F Fund''). This 
fund invests in a bond index fund that tracks the performance of the 
Shearson Lehman Brothers Aggregate (SLBA) bond index. These securities 
consist of government bonds, corporate bonds, and mortgage-backed 
securities. From 1988 through 2004, the ``F'' fund earned an average 
annual rate of return of 7.7%.
      the Common Stock Index Investment Fund (the ``C Fund''). 
This fund invests in stocks of thecorporations that are represented in 
the Standard and Poor's 500 index in the same proportion as they are 
represented in that index. During the period from 1988 through 2004, 
the ``C'' fund earned an average annual rate of return of 12.0%.
      the Small Capitalization Stock Index Investment Fund (the 
``S Fund''). This fund invests in the stocks of small and medium-sized 
companies incorporated in the United States. Stocks in this fund are 
held in the same proportion as they are represented in the Wilshire 
4500 stock index. The average annual rate of return on the Wilshire 
4500 from 1988 through 2004 was 12.7%.
      the International Stock Index Investment Fund (the ``I 
Fund''). This fund invests in the common stocks of foreign corporations 
represented in the Morgan Stanley Capital Investment EAFE (Europe, 
Australia-Asia, Far East) index. The average annual rate of return on 
the EAFE Index from 1988 through 2004 was 6.1%

    The Thrift Board has contracted with Barclays Global Investors to 
manage the index funds in which the F, C, S, and I Fund assets are 
invested. The contracts for each fund are open to competitive bids by 
qualified investment managers every three to five years.

Participant Vesting
    Thrift Plan participants are immediately vested in all of their own 
contributions and investment earnings on those contributions. \1\ 
Participants also are immediately vested in agency matching 
contributions made to their accounts and attributable earnings. In 
order to be vested in the agency automatic (1%) contributions, a FERS 
employee must have either 2 or 3 years of service as described in 
section 8432(g) of title 5 of the U.S. Code. FERS employees who are not 
vested and who separate from the federal government forfeit all agency 
automatic contributions and attributable earnings. Forfeited funds, 
consisting primarily of monies forfeited pursuant to 8432(g), totaled 
$10,822,000 in 2004 and $7,824,000 in 2003. By law, these funds are 
used to pay accrued administrative expenses of the Thrift Plan. If the 
forfeited funds are not sufficient to meet all administrative expenses, 
earnings on participant investments are then charged for administrative 
costs. In its most recent annual report, the plan reported 
administrative costs of six basis points, or six-hundredths of 1%. 
Thus, the administrative expenses of the Thrift Plan are about 60 cents 
for each $1,000 invested. \7\
---------------------------------------------------------------------------
    \6\ To ``vest'' in a benefit is to gain a legally enforceable right 
to receive it.
    \7\ See Table 3 for the Thrift Savings Plan's assets, income, and 
expenses in 2004 and 2003.
---------------------------------------------------------------------------
Participant Accounts
    The Thrift Plan maintains individual accounts for each participant. 
Participant accounts are credited with the participant's contributions, 
agency automatic and matching contributions, and charged with 
withdrawals. The value of the participant's account reflects the number 
of shares and the daily share prices of the funds in which it is 
invested. Administrative expenses are a component of the share price 
calculation. The benefit to which a participant is entitled is the 
participant's vested account. Thrift Plan participants can receive 
account-balance information and conduct transactions by automated 
telephone service or on the Thrift Plan's web site. \8\
---------------------------------------------------------------------------
    \8\ The URL of the Thrift Savings Plan web site is www.tsp.gov.
---------------------------------------------------------------------------
Participant Loans
    Participants may borrow from their accounts. There are two types of 
Plan loans: general purpose and residential. General purpose loans can 
be obtained for any purpose, with a repayment period from 1 to 5 years. 
Residential loans can be obtained for the purpose of purchasing a 
primary residence, with a repayment period from 1 to 15 years. 
Participant loans may only be taken from participant contributions and 
attributable earnings. The minimum loan amount is $1,000. The interest 
rate for loans is the ``G Fund'' interest rate at the time the loan 
agreement is issued by the Plan's record keeper. The rate is fixed at 
this level for the life of each loan. Interest earned on loans is 
allocated to the participant account upon repayment. Participants whose 
loans are in default have until the end of the following calendar 
quarter to pay the overdue amount. If not repaid by that time, the loan 
plus accrued interest is treated as a taxable distribution to the plan 
participant, which may be subject to the 10% penalty on retirement plan 
distributions made before age 59\1/2\.

Benefit Payments
    After leaving service, participants may elect benefit withdrawals 
in the form of a partial withdrawal or a full withdrawal as a single 
payment, a series of payments, or a life annuity. Participants may 
choose to combine any two, or all three, of the available withdrawal 
options. The Board has contracted with the Metropolitan Life Insurance 
Company to provide annuity products to Thrift Plan participants. The 
contract to issue Thrift Plan annuities is open to competitive bids 
every three to five years.

The Federal Retirement Thrift Investment Board
    The Federal Retirement Thrift Investment Board was established by 
the FERS Act of 1986. \9\ The Board is responsible for developing the 
investment policies of the Thrift Plan and overseeing the management of 
the plan, which is under the day-to-day direction of an Executive 
Director appointed by the Board.
---------------------------------------------------------------------------
    \9\ See 5 U.S.C. Sec. 8472.
---------------------------------------------------------------------------
    Three of the five members of the Board--including the Chairman--are 
appointed by the President. The President chooses a fourth member of 
the Board in consultation with the Speaker of the House and the House 
Minority Leader and a fifth member in consultation with the Majority 
and Minority Leaders of the Senate. Members of the Board serve 4-year 
terms and all nominations are subject to Senate confirmation. The law 
requires that all nominees to the Board must be individuals with 
``substantial experience and expertise in the management of financial 
investments and pension benefit plans.'' \10\
---------------------------------------------------------------------------
    \10\ See 5 U.S.C. Sec. 8472(d).
---------------------------------------------------------------------------
    The authorizing legislation that established the Thrift Board 
defines the Board's authority and responsibilities, and provides for 
substantial independence of the Board from political pressures.

Authority
    The Thrift Board has the authority to:
      Appoint the Executive Director of the Thrift Plan;
      Remove the Executive Director for cause (This requires 4 
votes of the 5-member Board.);
      Establish investment policies for the Thrift Plan;
      Instruct the Director to take whatever actions the Board 
deems appropriate to carry out the policies it establishes;
      Submit to the Congress legislative proposals relating to 
its responsibilities under federal law.

Independence
    Members of the Board are nominated by the President and confirmed 
by the Senate, but once confirmed they cannot be removed from their 4-
year terms without good cause. The selection and nomination process are 
designed to assure that Members of the Board are individuals who are 
supported by the President and Congress. They serve in times of good 
behavior, rather than at the pleasure of the President or Congress, 
assuring that they can carry out the responsibilities of their 
positions without of removal from office. The Federal Retirement Thrift 
Investment Board receives no appropriations from Congress. 
Administrative expenses are paid through agency-automatic contributions 
forfeited by employees who leave federal service before they have 
vested and charges against participant accounts.

Responsibility
    The law requires that the members of the Board shall discharge 
their responsibilities solely in the interest of participants and 
beneficiaries. In practice, this means that the investment policies and 
management practices of the fund are evaluated by the Board exclusively 
in reference to the efficient and prudent management of the Fund's 
assets. This exclusive responsibility serves to further insulate the 
Board from pressures to adopt investment policies or management 
practices that might not be in the long-term interest of preserving and 
increasing the security and investment performance of the Fund's 
assets.

Oversight
    To assure that the Members of the Thrift Board remain aware of the 
interests and concerns of Thrift Plan participants and beneficiaries, 
the authorizing legislation established the Employee Thrift Advisory 
Council. This 14-member council is appointed by the Chairman of the 
Thrift Board and must include representatives of federal employee and 
Postal Service labor organizations, managerial employees, supervisory 
employees, female employees, senior executives, and annuitants.
    All fiduciaries of the plan, including members of the Thrift Board 
are required by law to be bonded. \11\ The Secretary of Labor is 
authorized by law to investigate any suspected breach of duty by a 
fiduciary of the plan. The financial statements of the Thrift Board are 
audited regularly by an independent accounting firm. Congressional 
oversight of the Thrift Plan is performed by the House Committee on 
Government Reform and the Senate Committee on Homeland Security and 
Governmental Affairs.
---------------------------------------------------------------------------
    \11\ A ``fiduciary'' is a person in a position of trust or 
confidence with regard to the property of another. A ``bond'' is form 
of insurance against the potential malfeasance of a plan fiduciary.
---------------------------------------------------------------------------
Conclusion
    The Thrift Savings Plan is an efficient provider of retirement 
savings accounts to the federal workforce. It has achieved high 
participation rates and low administrative costs. The Thrift Plan is a 
key component of federal employees' retirement benefits. This is 
especially true for workers in the middle and upper ranges of the 
federal pay scale who would be unlikely to achieve adequate retirement 
income from just Social Security, the FERS basic annuity, and the 
government's automatic contribution of 1% of pay to the plan. Later 
this year, the Thrift Plan will begin to offer life-cycle funds that 
will allow employees to have their investments re-balanced with a 
greater weight toward corporate and government bonds as they approach 
retirement age, thus protecting their accumulated assets from a sudden 
downturn in the stock market just as they are about to retire.
    This concludes my testimony and I would be happy to answer any 
questions that members of the subcommittee might have.

                                     Table 1. Thrift Savings Fund Statistics
----------------------------------------------------------------------------------------------------------------
      Fund balances, in millions              March 2005             February 2005             January 2005
----------------------------------------------------------------------------------------------------------------
``G'' Fund                                         61,060  40%              60,066  39%              59,760  40%
----------------------------------------------------------------------------------------------------------------
``F'' Fund                                          10,079  7%               10,222  7%               10,279  7%
----------------------------------------------------------------------------------------------------------------
``C'' Fund                                         64,368  41%              65,589  42%              64,163  42%
----------------------------------------------------------------------------------------------------------------
``S'' Fund                                           9,847  6%               10,028  7%                9,681  6%
----------------------------------------------------------------------------------------------------------------
``I'' Fund                                           8,678  6%                8,325  5%                7,451  5%
----------------------------------------------------------------------------------------------------------------
Total                                           $154,032  100%            154,230  100%            151,334  100%
----------------------------------------------------------------------------------------------------------------
Twelve-month returns
----------------------------------------------------------------------------------------------------------------
``G''Fund                                                4.45%                    4.36%                    4.38%
----------------------------------------------------------------------------------------------------------------
``F'' Fund                                               1.17%                    2.36%                    4.07%
----------------------------------------------------------------------------------------------------------------
``C'' Fund                                               6.76%                    6.99%                    6.24%
----------------------------------------------------------------------------------------------------------------
``S'' Fund                                               7.95%                   10.42%                   10.14%
----------------------------------------------------------------------------------------------------------------
``I'' Fund                                              14.96%                   18.64%                   16.22%
----------------------------------------------------------------------------------------------------------------
Participants (thousands)
----------------------------------------------------------------------------------------------------------------
FERS, contributing                                       1,539                    1,543                    1,553
----------------------------------------------------------------------------------------------------------------
FERS, agency 1% only                                       243                      237                      234
----------------------------------------------------------------------------------------------------------------
FERS participation rate                                  86.4%                    86.7%                    86.9%
----------------------------------------------------------------------------------------------------------------
FERS, without agency 1%                                     71                       63                       55
----------------------------------------------------------------------------------------------------------------
Total FERS with contributions                            1,853                    1,843                    1,842
----------------------------------------------------------------------------------------------------------------
CSRS contributing                                          449                      454                      465
----------------------------------------------------------------------------------------------------------------
Uniformed services                                         476                      478                      458
----------------------------------------------------------------------------------------------------------------
Participants, not contributing                             661                      663                      657
----------------------------------------------------------------------------------------------------------------
Total TSP participants                                   3,439                    3,438                    3,422
----------------------------------------------------------------------------------------------------------------
Loans outstanding
----------------------------------------------------------------------------------------------------------------
Number                                                 859,386                  872,240                  883,357
----------------------------------------------------------------------------------------------------------------
Amount (millions of $)                                  $4,908                   $4,969                   $5,033
----------------------------------------------------------------------------------------------------------------
Source : Federal Retirement Thrift Investment Board.



                          Table 2. Annual Rates of Return for Thrift Savings Plan Funds
----------------------------------------------------------------------------------------------------------------
                              Year                                 G Fund   C Fund    F Fund   S Fund    I Fund
----------------------------------------------------------------------------------------------------------------
1988                                                                 8.8%     11.8%     3.6%     20.5%     26.1%
----------------------------------------------------------------------------------------------------------------
1989                                                                 8.8%     31.0%    13.9%     23.9%     10.0%
----------------------------------------------------------------------------------------------------------------
1990                                                                 8.9%     -3.2%     8.0%    -13.6%    -23.6%
----------------------------------------------------------------------------------------------------------------
1991                                                                 8.1%     30.8%    15.7%     43.5%     12.2%
----------------------------------------------------------------------------------------------------------------
1992                                                                 7.2%      7.7%     7.2%     11.9%    -12.2%
----------------------------------------------------------------------------------------------------------------
1993                                                                 6.1%     10.1%     9.5%     14.6%     32.7%
----------------------------------------------------------------------------------------------------------------
1994                                                                 7.2%      1.3%    -3.0%     -2.7%      7.8%
----------------------------------------------------------------------------------------------------------------
1995                                                                 7.0%     37.4%    18.3%     33.5%     11.3%
----------------------------------------------------------------------------------------------------------------
1996                                                                 6.8%     22.8%     3.7%     17.2%      6.1%
----------------------------------------------------------------------------------------------------------------
1997                                                                 6.8%     33.2%     9.6%     25.7%      1.5%
----------------------------------------------------------------------------------------------------------------
1998                                                                 5.7%     28.4%     8.7%      8.6%     20.1%
----------------------------------------------------------------------------------------------------------------
1999                                                                 6.0%     21.0%    -0.8%     35.5%     26.7%
----------------------------------------------------------------------------------------------------------------
2000                                                                 6.4%     -9.1%    11.7%    -15.8%    -14.2%
----------------------------------------------------------------------------------------------------------------
2001                                                                 5.4%    -11.9%     8.6%     -2.2%    -15.4%
----------------------------------------------------------------------------------------------------------------
2002                                                                 5.0%    -22.1%    10.3%    -18.1%    -16.0%
----------------------------------------------------------------------------------------------------------------
2003                                                                 4.1%     28.5%     4.1%     42.9%     37.9%
----------------------------------------------------------------------------------------------------------------
2004                                                                 4.3%     10.8%     4.3%     18.0%     20.0%
----------------------------------------------------------------------------------------------------------------
1988-2004                                                            6.6%     12.0%     7.7%     12.7%      6.1%
----------------------------------------------------------------------------------------------------------------
Source: www.tsp.gov, www.wilshire.com, www.msci.com.
Note: Rates of return for the C, G, and F funds are shown net of TSP expenses.


 Table 3. Financial Statements of the Thrift Savings Fund Statements of
 Net Assets Available for Benefits as of December 31, 2004 and 2003 (In
                               thousands)
------------------------------------------------------------------------
                                                  2004          3003
------------------------------------------------------------------------
ASSETS:
------------------------------------------------------------------------
Investments, at fair value:
------------------------------------------------------------------------
U.S. Government Securities Investment Fund     $56,670,880   $51,121,034
------------------------------------------------------------------------
Barclays U.S. Debt Index Fund                    9,732,943    10,071,287
------------------------------------------------------------------------
Barclays Equity Index Fund                      63,218,611    54,303,506
------------------------------------------------------------------------
Barclays Extended Market Index Fund              9,644,143     5,622,444
------------------------------------------------------------------------
Barclays EAFE Index Fund                         7,021,069     2,211,875
------------------------------------------------------------------------
Participant loans                                5,105,715     5,130,170
------------------------------------------------------------------------
                                               151,393,361   128,460,316
------------------------------------------------------------------------
Total investments
------------------------------------------------------------------------

------------------------------------------------------------------------
Receivables:
------------------------------------------------------------------------
Employer contributions                             166,045       151,497
------------------------------------------------------------------------
Participant contributions                          507,034       446,574
------------------------------------------------------------------------

------------------------------------------------------------------------
Total receivables                                  673,079       598,071
------------------------------------------------------------------------

------------------------------------------------------------------------
Fixed assets, total:                                41,839        39,715
------------------------------------------------------------------------
Other assets                                         5,460        11,236
------------------------------------------------------------------------

------------------------------------------------------------------------
Total assets                                   152,113,739   129,109,338
------------------------------------------------------------------------

------------------------------------------------------------------------
LIABILITIES:
------------------------------------------------------------------------

------------------------------------------------------------------------
Total liabilities                                   99,984       179,216
------------------------------------------------------------------------

------------------------------------------------------------------------
Funds restricted for the purchase of
------------------------------------------------------------------------
Fiduciary Insurance                                 -4,829        -4,978
------------------------------------------------------------------------

------------------------------------------------------------------------
Net Assets Available for Benefits             $152,008,926  $128,925,144
------------------------------------------------------------------------

------------------------------------------------------------------------
ADDITIONS:
------------------------------------------------------------------------

------------------------------------------------------------------------
Investment income (loss):
------------------------------------------------------------------------
U.S. Government Securities Investment Fund      $2,346,104    $2,074,004
------------------------------------------------------------------------
Net appreciation (depreciation) in fair
 value
------------------------------------------------------------------------
of Barclays funds:
------------------------------------------------------------------------
Barclays U.S. Debt Index Fund                      408,397       455,956
------------------------------------------------------------------------
Barclays Equity Index Fund                       6,115,843    11,316,657
------------------------------------------------------------------------
Barclays Extended Market Index Fund              1,249,934       914,990
------------------------------------------------------------------------
Barclays EAFE Index Fund                           870,403       358,102
------------------------------------------------------------------------
Interest income on participant loans               237,684       222,422
------------------------------------------------------------------------
Asset Manager rebates                                1,778         1,616
------------------------------------------------------------------------
Less investment expenses                            -4,503        -3,708
------------------------------------------------------------------------

------------------------------------------------------------------------
Net investment income (loss)                    11,225,640    15,340,039
------------------------------------------------------------------------

------------------------------------------------------------------------
Contributions:
------------------------------------------------------------------------
Participant                                     11,980,077    10,366,123
------------------------------------------------------------------------
Employer                                         4,238,199     3,887,260
------------------------------------------------------------------------
Total contributions                             16,218,276    14,253,383
------------------------------------------------------------------------

------------------------------------------------------------------------
Total additions                                 27,443,916    29,593,422
------------------------------------------------------------------------

------------------------------------------------------------------------
DEDUCTIONS:
------------------------------------------------------------------------

------------------------------------------------------------------------
Benefits paid to participants                    4,110,891     2,774,685
------------------------------------------------------------------------
Administrative expenses                             91,896        75,038
------------------------------------------------------------------------
Participant loans declared taxable                 157,496       130,559
 distributions
------------------------------------------------------------------------

------------------------------------------------------------------------
Total deductions                                 4,360,283     2,980,282
------------------------------------------------------------------------

------------------------------------------------------------------------
Change in funds restricted for the purchase
 of
------------------------------------------------------------------------
Fiduciary Insurance                                    149           375
------------------------------------------------------------------------
Net increase                                    23,083,782    26,613,515
------------------------------------------------------------------------

------------------------------------------------------------------------
NET ASSETS AVAILABLE FOR BENEFITS:
------------------------------------------------------------------------

------------------------------------------------------------------------
Beginning of year                              128,925,144   102,311,629
------------------------------------------------------------------------
End of year                                   $152,008,926  $128,925,144
------------------------------------------------------------------------

    Source: Financial statements of the Thrift Savings Plan [http://
www.tsp.gov/forms/financial-stmt.pdf].

                                 

    Chairman MCCRERY. Just to advise the witnesses of the order 
that I called, Mr. Pollock will be next, then Ms. Entmacher, 
Mr. Cavanaugh, and Ms. Reno. Mr. Pollock, you may proceed.

    STATEMENT OF ALEX J. POLLOCK, RESIDENT FELLOW, AMERICAN 
                      ENTERPRISE INSTITUTE

    Mr. POLLOCK. Thank you, Mr. Chairman, Ranking Member Levin, 
Members of the Subcommittee, I would like to start by taking a 
minute to look back into the insight of Congressman J.J. 
Pickle, who died this last Saturday and who was, of course, a 
Chairman of this Subcommittee, and who knew that in order to 
protect Social Security, you had to change it. Chairman Pickle 
told the Advisory Council on Social Security in 1995, ``The 
public knows that change in the Social Security program is 
necessary, and lack of action will be seen as failure in 
leadership, not as protecting interests.''
    Mr. Chairman, I think Chairman Pickle was absolutely right 
in 1995, and his words are absolutely right today. The public 
knows change has to happen, and the public knows something 
else: That in exchange for the money that they send in to 
Social Security, as Congressman Lewis was saying a minute ago, 
they would like to have an individual right where there 
actually exists an obligation for their retirement savings. 
Congressman Becerra made a similar point: there should be such 
an obligation. However, under the current Social Security 
program with the current trust fund, as we call it, the Supreme 
Court has made it clear there is no such obligation to any 
individual, again, as Congressman Lewis so rightly said. Now, 
we could create a clear and unquestionable obligation of the 
United States to the citizens individually. There is an 
instrument readily available to do that. It is called a United 
States Treasury Bond. That is an inviolable contract 
obligation.
    This brings us to linking Social Security to the 
philosophical ideal of ownership. Widely dispersed ownership 
throughout the society is deeply embedded, and one of the best 
elements of the American traditional political philosophy. It 
goes back to John Locke, the philosophical father of 
representative democracy, to the First Continental Congress, 
and among American Presidents, particularly to Jefferson and 
Lincoln. We have all kinds of programs in this country, and 
rightly so, to promote ownership in the form of home ownership. 
Ownership of retirement savings carries out the same principles 
and would carry out the same philosophy.
    Now, suppose there were a way to make at least part of 
Social Security a truly inviolable ownership obligation for the 
individual, which resulted in no cash shortfall to the 
Treasury, no investment risk to households, no default risk, no 
inflation risk, no increase in total government debt, and, 
moreover, it were a voluntary program where individuals had the 
choice to participate or not. I think everybody should support 
such a program. Who could be against it? I certainly think that 
the vast majority of the Americans who worry about what their 
individual rights in Social Security are would support this 
program.
    Now, there is, in fact, a way to do this. I call it your 
``personal lockbox.'' Here is how it would work. Consider the 
current way we use the Old-Age and Survivors Insurance (OASI) 
surplus, which was about $145 billion in 2004 and is expected 
to rise to $200 billion annually, or so, during the next 
decade. Americans pay their Social Security taxes. As we know, 
the Treasury actually gets the cash, and Treasury issues a bond 
to the trust fund. As we also know, this is an odd kind of 
bond. It is debt of the government to itself. It is an ``I-owe-
me,'' as somebody has said. As we all know, economically and 
financially, the size of the trust fund, whether it be large or 
whether it be zero, has absolutely no, zero, economic effect on 
the finances of the United States. So, now we need to make only 
one simple change: Americans pay the same Social Security taxes 
as now, Treasury gets the same cash they get now, the Treasury 
issues a bond, just as it does now, but that bond goes to your 
individual account, your personal lockbox.
    It is now an IOU, it is a real bond, it is a real asset for 
American families. It has become an inviolable contract. It can 
be inherited, and since it is a Treasury bond, it has no 
default risk. If we make it a TIPS, it will also have no 
inflation risk. Mr. Chairman, as we know, inflation is the 
greatest risk to retirement savings. So, putting TIPS into 
these personal lockboxes, I think, takes a huge step forward in 
making Social Security what I believe most Americans think it 
should be. It will also be an exceptionally low-cost, 
efficient, purely book entry system, very cheap to operate.
    There are many details in my written testimony. I would 
only say if we make this a voluntary option available to the 
American people to get TIPS in their own accounts in exchange 
for the surpluses, it will look to them just like payroll 
deductions to buy savings bonds. We will have very little 
educational effort, everyone will understand this immediately, 
and I would be willing, Mr. Chairman, to bet a large amount of 
my personal money that a large majority of Americans would 
volunteer for this program. Indeed, how could anyone oppose 
giving them this choice? Mr. Chairman, thank you very much for 
the opportunity to be here today.
    [The prepared statement of Mr. Pollock follows:]

  Statement of Alex J. Pollock, Resident Fellow, American Enterprise 
                      Institute Summary Statement

    Good morning, Mr. Chairman, Ranking Member Levin, and members of 
the Subcommittee. Thank you for the opportunity to testify today. I am 
Alex Pollock, a Resident Fellow at the American Enterprise Institute, 
and these are my personal views and recommendations.
    Linking ownership of property to liberty in a free society is 
deeply embedded in the American political philosophy, going back to the 
ideas of John Locke and the First Continental Congress. Personal Social 
Security accounts as vehicles for the expansion of ownership of 
retirement assets are very much in keeping with this American 
tradition.
    By creating ``personal lock boxes'' invested in Treasury inflation-
indexed bonds, Congress can make such accounts a reality in a highly 
efficient, low cost, low risk, sensible and easily understandable way. 
It would also make social security at least in part what most people 
think it should be: a retirement account building up your own personal 
savings.
    What is happening today, as we all know, is that some social 
security contributions are diverted into general government 
expenditures, evidenced by debt of the government to itself, or ``I owe 
me's,'' in the social security trust fund. In contrast, without 
diverting taxes or cash from the Treasury, personal accounts--``your 
personal lock box''--can be created with real Treasury bonds held by 
the public.
    Consider how the ``investment'' of the social security surplus 
currently works: Social Security taxes are collected by the Internal 
Revenue Service and deposited in the general fund at the Treasury 
Department, where they are spent on benefits but also on other federal 
programs. In exchange for these ``invested'' Social Security funds, the 
Treasury issues bonds to the Social Security trust fund. The bonds 
represent government liabilities to itself (``I owe me's''), rather 
than real obligations to the public.
    I suggest creating personal accounts with the annual social 
security surplus which protects it for retirement savings, without 
diverting cash from payroll taxes. This could be done by changing the 
current structure in one key respect: the Treasury would issue bonds 
directly to personal accounts, bypassing the confusing role of the 
trust fund, thereby creating ``your personal lock box'' of explicit 
government obligations.
    The personal accounts would be created by putting Treasury 
securities in them, not cash. The current trust fund is an unnecessary 
and confusing. The citizens and the U.S. Treasury are the only actual 
principals involved. Personal lock boxes would protect the social 
security surplus, make the accounting much clearer and more honest, and 
make the citizens direct owners of top quality retirement assets.
    There are perfect Treasury bonds for these accounts: Treasury 
Inflation Protected Securities (TIPS). Inflation poses the largest 
threat to retirement savings, and these default-free instruments also 
fully protect against that threat, thereby minimizing risk.
    Moreover, because these bonds operate on a book-entry basis, the 
program would have very low operating and administrative costs.
    According to the 2005 Annual Report of the Social Security 
Trustees, the OASI program (excluding the Disability Insurance program) 
ran a 2004 surplus of $145 billion. This represented over 30% of total 
OASI contributions of $473 billion. The 10-year intermediate case 
projection is for an aggregate OASI surplus of over $2 trillion. 
Instead of continuing the ``I owe me'' approach with this surplus, 
personal lock boxes could turn this $2 trillion into real assets of 
American households, free of both default and inflation risk.
    The Trustees' intermediate projection suggests that the OASI trust 
fund would stay approximately level, and the going-forward surpluses 
thus protected, if about half of the employee's social security tax 
were converted to TIPS in personal lock boxes each year. This would 
mean that Americans would receive real assets equal to 2.65 % (half 
of5.3 %) of income subject to social security tax each year. For a 
household with median 2003 income of $43,000, this would result in an 
initial year personal lock box account of over $1,100.
    I recommend the personal lock box as a purely voluntary program; 
individuals could elect either to remain in the current program or to 
receive TIPS in their personal accounts instead of future benefit 
payments of equal economic value. After a certain restricted period (I 
suggest five years), individuals could choose to reinvest their assets 
in other financial instruments, although I believe a many would simply 
stay with the TIPS ``default option.'' Ownership through personal 
accounts would also allow for account holders to bequeath their assets 
to future generations.
    Think how much more meaningful direct ownership of these Treasury 
bonds in a personal account--in ``your personal lock box''--would be 
for American individuals and families than the obscure operations of 
the current trust fund which few understand.
    In my opinion, TIPS would be an extremely popular alternative--
simple, easy to understand, and attractive. By analogy to the federal 
employees' thrift plan, it could be thought of as ``a G-Fund for 
everybody.''
    The personal lock box would result in greater and more widely 
distributed ownership of financial assets among American households. It 
would provide assets with no default risk and no inflation risk, with 
the ability to pass them on to future generations. It would establish a 
stronger and more honest financial relationship between government and 
citizens. Treasury securities are in fact inviolable contracts, in 
contrast to off-balance sheet future political formulas.
    I believe a large majority of Americans would prefer to accumulate 
inflation-protected retirement assets they actually own. They should be 
given this choice.

                      ``Your Personal Lock Box'':

          A New Approach to Personal Social Security Accounts

     Mr. Chairman, Ranking Member Levin, and Members of the 
Subcommittee, thank you for the opportunity to testify today. I am Alex 
Pollock, a Resident Fellow at the American Enterprise Institute, and 
these are my personal views and recommendations.
    By transforming Social Security, at least in part, to a program of 
greater personal property for the average American, voluntary personal 
accounts would be a key structural reform.
    But most current proposals for personal accounts also have serious 
disadvantages: they are complicated, to many people they are confusing 
and they require diverting a portion of payroll taxes away from the 
U.S. Treasury. How can there be effective management for millions of 
small accounts? Isn't the stock market too risky? Won't many people be 
confused by being forced to make choices they do not understand? Who 
can be sure the benefits are worth the costs and risks?
    There is, however, a better way to launch Social Security reform 
using private accounts and inflation-indexed Treasury bonds (or 
``TIPS''), which will deliver all of the benefits of personal accounts 
with none of the costs or risks cited by their opponents.
    I propose creating personal accounts, or ``Your Personal Lock 
Box,'' with an extremely simple and clear financial structure, without 
diverting any payroll tax receipts away from the U.S. Treasury, and 
with low cost and efficient operations. The results will be greater 
ownership of risk-free assets throughout American households, ability 
for inheritance, clear links between one's own efforts and retirement 
savings, and complete clarity in the dealings between the government 
and the citizens. The transition could begin promptly.
    The essential proposal is this: Social Security tax payments by 
individuals and employers, and Social Security tax receipts by the 
government would remain the same as they are now. Treasury would have 
the same cash receipts from Social Security taxes as it does now. But 
in exchange for the going-forward investment of OASI surplus 
contributions, Treasury would not issue bonds to the Social Security 
trust fund. Instead it would issue bonds--specifically, inflation-
indexed bonds or ``TIPS''--directly to the personal accounts of the 
individual citizens themselves, which would become in effect their own 
personal lock boxes. These accounts would not receive cash but would 
automatically receive the safest possible investment for retirement 
savings.
    This is proposed as a voluntary alternative covering the portion of 
Social Security taxes which represents mandatory savings of OASI 
surpluses. Everyone would be given the choice to participate in the 
proposed personal accounts or stay in the current Social Security 
program. I believe that a large majority would choose the personal 
accounts if they are designed as recommended, but this should be a 
purely voluntary option.
    This financial structure transparently shows the real transaction 
which is taking place between the two real principals involved: the 
American citizen and the U.S. Treasury Department. It cuts out the 
unnecessary and confusing role of the Social Security trust fund, which 
in fact is debt of the government to itself, or an ``I owe me.''
    The government's total obligations would not increase. Some 
Treasury debt would shift from being owned by the intra-governmental 
trust fund to being owned directly by the citizens themselves in their 
own personal lock box accounts. The bonds in the personal accounts 
would represent an increase in Treasury debt owned by the public, but 
would be issued, like bonds now sent to the government's trust fund, as 
automatic private placements.

Simplicity
    The simplicity of the proposed approach would remove from the 
current political debates many distracting issues, such as whether we 
could afford the transition costs, whether personal accounts would be 
too risky, whether Wall Street would reap a bonanza, and whether 
operating costs would be too high. It would make unnecessary the 
proposed delay in implementation until 2009.
    It would also remove a central objection made by the opponents of 
personal accounts: that Social Security must be a moral imperative, an 
inviolable promise and part of the social contract. Nothing could make 
Social Security more imperative, inviolable, and a contract than to 
turn it into a U.S. Treasury bond. Indeed, the only advantage which 
might be argued for the current Social Security structure over the 
proposed personal accounts is that the current structure leaves open 
the possibility for the government to renege on its promises and reduce 
benefits. This is presumably not an argument that opponents of personal 
accounts will wish to emphasize.
    How much of the current structure should be replaced by the 
proposed personal accounts? The answer reflects the fact that Social 
Security has two components: first, a mandatory savings program for 
retirement and old age, applicable to citizens of all levels of income; 
and second, a welfare or safety net program providing a minimum 
retirement income and disability insurance.
    The second component by definition requires commingling of funds 
and should remain as it is. This would include the disability portion 
of Social Security and the provision of a minimum retirement income for 
low income households.
    The proposed personal accounts apply to the first or mandatory 
savings component: which is what most Americans think their Social 
Security payments should be. I suggest that half of the employee's 
share of Social Security taxes, which represents mandatory saving of 
approximately the OASI annual surplus, should have available this 
personal account option.
    The simplicity of the proposed change to protect the surplus is 
easy to see by reviewing the current structure of Social Security and 
contrasting it with the proposal.

Current Structure for Social Security Surpluses
    The current Social Security structure handles surpluses with the 
following process:
    A. Cash from the citizen, both directly from wages and indirectly 
as employer contributions which could otherwise have been wages, is 
sent to the government as Social Security taxes.
    B. Social Security cash goes to the U.S. Treasury
    C. Treasury issues a Treasury debt obligation to the Social 
Security program. It is part of the total Treasury debt outstanding, 
but is an ``I owe me.''

A New Structure for Personal Accounts
    In the proposed structure, there would be no diversion of cash from 
the Treasury. Social Security payroll taxes paid to the government and 
cash received by the Treasury would stay the same as under the current 
structure. If voluntarily chosen by the citizen, the portion of these 
taxes representing the OASI annual surplus would be earmarked for 
personal lock box accounts. However, these accounts would not receive 
cash, but automatically receive an appropriate Treasury inflation-
indexed security.
    The surplus investment function would thus work as follows:
    A. Social Security taxes would be sent to the government, as they 
are now.
    B. Treasury's cash receipts would be the same as they are now. 
There would be no cash shortfall.
    C. Treasury would continue to issue a Treasury debt obligation, but 
to the citizen's personal account, not to the trust fund--thereby 
creating ``your personal lock box.''
    That is all. Thus the citizen would own a risk-free investment very 
well suited for retirement savings: an inflation-indexed Treasury 
security. Treasury debt owned by the public in personal accounts has 
increased, but debt owned by the trust fund has decreased. Treasury 
owes the citizen explicitly and clearly, rather than confusingly owing 
the government itself.
    Since the savings are now in the form of a directly owned, actual 
Treasury bond instead of future Social Security benefits, there must of 
necessity be an equivalent reduction in future benefits to offset the 
acquired Treasury security. The trust fund does not receive Treasury 
bonds but by the same taken has reduced future benefit liabilities. For 
the citizen, the replacement of future benefits with actual assets of 
course applies only on a going-forward basis, as the personal accounts 
grow. All benefits earned by past Social Security taxes, before the 
private accounts transition, would remain unchanged.
    The proposed structure is quite similar to a historically tried and 
true long-term savings program: payroll deduction for the purchase of 
U.S. savings bonds. It is also similar to a very popular option under 
the Thrift Savings Plan for federal government employees: the ``G 
Fund,'' which invests solely in U.S. Treasury obligations.
    Such analogies, as well as the basic simplicity of the structure, 
would make it easy for the public to understand. Would most people 
choose to create their own portfolio of Treasury inflation-indexed 
bonds rather than hoping for future payments from off-balance-sheet 
political formulas? I think they would--by a large majority.

Relation to Future Benefits
    If the economic value of the bonds acquired in the personal 
accounts is exactly equal to the economic value of the reduction in 
future off-balance-sheet future benefit promises, we would have created 
the many advantages of ownership, but the aggregate Social Security 
fiscal deficit would remain unchanged. However, this trade-off could be 
given a progressive structure, analogous to recent proposals for 
progressive changes to Social Security indexation formulas, for high-
income households.
    In other words, for the majority of households the TIPS exchange 
ratio would be 1 to 1, but for high income households it could be 
greater than 1 to 1. Since many of these households believe that in any 
case, their Social Security taxes will inevitably increase or their 
future benefits be reduced, or both, the trade in exchange for 
achieving personal accounts could be viewed as advantageous. The 
transition to personal accounts would then reduce the Social Security 
deficit, in addition to its other attractions.

The Specific Treasury Bond
    The perfect candidate for the Treasury obligations to be issued to 
the personal Social Security accounts is clear: Treasury Inflation 
Protected Securities (TIPS). TIPS by definition preserve purchasing 
power against inflation, the single greatest risk and an essential 
consideration for retirement savings.
    The TIPS would be issued in automatic private placements for each 
personal account. Because all the TIPS involved will be book-entry 
securities in fully automated form, small accounts and small amounts 
could be easily handled, and operating costs will be low.
    Suggestions for how the details of this would work follow. Details 
could obviously vary around the essential structure.
    The TIPS should have maturities based on the individual's expected 
retirement date. For example, a twenty-five-year-old with an expected 
retirement age of sixty-five might in the first instance receive a 
forty-year TIPS. Note that it is proposed to consider creating long-
term TIPS to match the needs of retirement savings. All interest and 
inflation adjustments should simply accrue, as with typical savings 
bonds, so there is no problem of investing small amounts of cash. 
Laddering maturities as discussed below would result in a sensible 
pattern of cash flow during retirement.
    The average real return of government bonds (i.e. the yield net of 
inflation) in the long term is approximately 3 percent. The long-term 
TIPS to be privately placed in the personal accounts with a restricted 
period could have a real yield of about this same 3 percent. In an 
average inflation of 2 or 3 percent, for example, this would result in 
a compound annual return of 5 or 6 percent, respectively. A 3 percent 
real yield would match the real 3-percent discount rate often used in 
calculations of the value of future Social Security benefits.
    For ownership to be effective, the TIPS received in the personal 
accounts must be negotiable securities. However, it would make sense to 
have a period after each private placement during which sale would be 
restricted. After that, the citizen would be entirely free to sell in 
order to make other eligible investments, if desired, provided of 
course that all proceeds and investments must stay in the retirement 
account until qualified for withdrawal.
    The appropriate length of the restricted period before the 
privately placed TIPS would become negotiable must be defined. A 
starting suggestion would be five years, to insure a smooth transition, 
while also allowing the future addition of private asset categories.
    The maturities of the TIPS should be based on expected retirement 
age but should not all mature at that date, which would cause a 
difficult decision point and large reinvestment risk. The idea of 
buying an annuity upon retirement does not address this problem, since 
if at that time interest rates are low, annuities will be unattractive 
to purchase--not to mention the need to address the credit risk of the 
annuity writer. A preferable approach would be to automatically ladder 
the maturities of the TIPS in the personal accounts to spread cash 
receipts from maturing bonds over the retirement years. Recall in this 
context that the safety net component of Social Security would also 
continue to function.
    Individuals who choose to continue working past retirement age 
would continue to accumulate assets in their personal accounts. This 
would provide an incentive to reduce the extended period of retirement 
which is a central cause of Social Security's fiscal deficit without 
having to mandate changes in retirement age that would naturally be 
inappropriate in many individual cases.
    In sum, the personal lock box accounts would represent a voluntary 
way to hold mandatory savings, while explicitly protecting the social 
security surpluses. Continuing to hold the TIPS past their restricted 
period would also be voluntary.
    But no investment decisions or risks would be forced upon the 
citizen. Especially considering those who might feel confused or 
intimidated, no action would be required to have a very sensible and 
safe investment, with zero credit risk and guaranteed inflation 
protection, very suitable for retirement savings, automatically 
provided. This means that there is a robust ``default case,'' an 
important element in a system of choices.
    A safe prediction is that a significant proportion of these 
securities would never be sold, but would be held to maturity. There 
would be no rush and no pressure on the individual to have to do 
anything, unlike the case of having to invest cash. In addition, the 
restricted period should comfort any observers who might fear the 
possibility, however unlikely, of a large initial outflow of TIPS into 
the market.

Benefits for an Ordinary Couple
    Suppose an ordinary couple signed up for the personal account 
option when they were both twenty-five years old, with a household 
income of $50,000 per year. What might their personal account 
retirement assets look like at age sixty-five, assuming the ``default 
case'' of simply holding their TIPS?
    As an example, assume the real yield on TIPS is 3 percent, average 
inflation of 2.5 percent, real wage increases of 1.5 percent, and half 
the Social Security tax represents mandatory savings devoted to 
personal accounts. At age sixty-five they would own investments 
totaling over $800,000. If they worked to age seventy in line with 
their greater expected longevity and health, the personal account 
investments would total $1.15 million.
    Now suppose two-thirds of the Social Security tax represents 
mandatory savings which generate TIPS for the personal account. At 
sixty-five, the investments would be more than $1 million; and at age 
seventy, more than $1.5 million.
    These would be real assets, really owned by ordinary Americans.

Conclusion
    The proposed approach would lead to personal Social Security 
accounts as a key transition and structural reform. It addresses all of 
the objections to private accounts, as follows:

         1.  There would be no cash shortfall to the Treasury.
         2.  There would be no increase in the total national 
        obligations. Treasury debt owned by the public would increase, 
        but Treasury debt owned by the trust fund would decrease. Off-
        balance-sheet future benefit liabilities would also decrease. 
        If the suggested progressive structure were adopted, future 
        liabilities would decrease by more than the value of the TIPS 
        issued, thus reducing the Social Security deficit.
         3.  There would be no need to market more Treasury debt--the 
        bonds involved would automatically be privately placed in the 
        personal lock box accounts.
         4.  No difficult choices would be imposed on individuals--if 
        they do nothing, a very safe and appropriate retirement 
        investment is automatically provided. The default case is 
        robust.
         5.  There is no pressure to take risk or ``roll the dice.'' 
        TIPS are the exact opposite of rolling the dice. In particular, 
        they directly address the biggest risk to retirement savings, 
        namely inflation.
         6.  The obligation of the government for social security 
        surpluses is made truly inviolable by becoming an explicit 
        Treasury bond.
         7.  The use of TIPS would allow a very low cost, efficient 
        book entry system.
         8.  With investments automatically provided, there is no 
        windfall for Wall Street, and small accounts can be handled 
        efficiently.
         9.  Appropriate long-term investments matched to retirement 
        needs are automatically provided.
        10.  The proposal would allow prompt implementation.

    Moreover, the idea is simple and easy to understand. As a voluntary 
alternative, I believe having ``your personal lock box'' would be 
readily chosen by a large majority of Americans.

                                 

    Chairman MCCRERY. Thank you, Mr. Pollock. Ms. Entmacher.

   STATEMENT OF JOAN ENTMACHER, VICE PRESIDENT AND DIRECTOR, 
     FAMILY ECONOMIC SECURITY, NATIONAL WOMEN'S LAW CENTER

    Ms. ENTMACHER. Thank you, Mr. Chairman. I appreciate this 
opportunity to testify on behalf of the NWLC. My written 
testimony addresses some of the system design issues presented 
when individual workers, particularly women, try to convert the 
proceeds of a private account into a secure and equitable 
income throughout retirement. Other witnesses have addressed 
some of these issues, my remarks will focus on the question of 
whether and how the proceeds of a private account are to 
provide for the spouses, surviving spouses, divorced spouses, 
and children who rely on the family insurance benefits that 
Social Security provides. These issues are of special 
importance to women. Nearly 14 million Americans receive a 
spousal benefit from Social Security, and 98 percent of those 
who do are women. So, women need to know when things like 
``personal lockboxes'' are described, what will be the rights 
of a spouse, a widow, a divorced spouse to the funds in those 
accounts if Social Security spousal benefits are cut as part of 
a private accounts plan, and many private accounts plans do cut 
those benefits.
    At retirement, for example, will married workers be 
required to purchase a joint and survivor annuity to protect a 
surviving spouse the way Social Security does? Without such a 
requirement, a widow could be left with absolutely nothing from 
his account to supplement her own probably lower Social 
Security benefit and private account. Adding insult to injury, 
she could be facing an additional reduction in her Social 
Security widow's benefit precisely because he elected to 
participate in a private account. Requiring the purchase of a 
joint and survivor annuity would help protect elderly widows 
from impoverishment and promote fairness. There are real trade-
offs, which perhaps is why the Administration has so far failed 
to answer this very basic design question. Under Social 
Security, a worker's benefits are not reduced because Social 
Security provides payments to a surviving spouse. A private 
account is a finite sum of money. Purchasing a joint and 
survivor annuity means the worker will get lower payments.
    Ms. Mitchell has suggested that accounts should be 
annuitized to provide a poverty level benefit for both the 
worker and spouse. This is not a requirement the President has 
talked about so far, possibly because, the fact is, many 
accounts simply will not be large enough to provide such 
benefits for both people. In addition, annuities are 
inflexible. They cannot respond to changes in marital status 
that may occur after an annuity is purchased. Whether one is 
widowed right before or right after annuitization could make a 
big difference in what the widowed individual receives.
    Then there are subsidiary questions: How large a survivor 
annuity should be provided, and should waivers be permitted? 
Second, will workers be permitted to leave an account to 
anyone, as the President has sometimes said, or would widows 
have the right to an inheritance from the account to mitigate 
the impact of cuts in survivor benefits? If a young widow does 
inherit account assets, would she have immediate access to the 
funds to help support her family and supplement their reduced 
survivor benefits, or would she be required to keep that money 
aside for her own retirement, as several plans provide? Along 
with an inheritance from the account, would she inherit the 
obligation to pay back Social Security with interest out of a 
reduction in her own Social Security benefits later? Would the 
requirement to pay back Social Security apply to other heirs, 
and would they have access immediately to the accounts? What 
about minor children and adult disabled children who are 
entitled to Social Security benefits on a worker's record, 
would they have a right to inheritance from an account to 
offset cuts in those benefits?
    Third, how would accounts be divided at divorce? One 
approach that has been suggested would divide the assets that 
accumulate in accounts during the course of a marriage equally 
between spouses. It sounds simple. It isn't. Simply 
implementing the division of accounts at divorce will require 
new reporting, verification, and dispute resolution mechanisms 
well beyond those needed to administer either the current 
Social Security System or the thrift savings system. About 
100,000 applications for Social Security each year involve 
establishing evidence of divorce. There are 10 times as many 
divorces each year in the United States. The SSA doesn't need 
to track marital status over the lifetime. It just figures it 
out when application is made for benefits, and those benefits 
are based on the worker's record. The TSP doesn't need to do 
this either. It just follows the instructions that it gets from 
a State divorce court. If a woman doesn't have a lawyer, and 
doesn't know to ask for a share of a thrift savings account in 
a divorce, she gets nothing. I would point out right now there 
is no national registry with information about marriages and 
divorces in the United States. To divide accounts based on 
contributions during marriage would require the creation of 
such a registry, and, apart from the cost, that would raise a 
lot of new confidentiality issues.
    Finally, one last point. In a system of voluntary personal 
accounts, what would happen if one spouse decides to 
participate and the other doesn't? How would any of the rules 
created apply? I think the simplest answer to these questions 
would be to focus on strengthening the safety net that Social 
Security currently provides, rather than shifting resources 
into private accounts. Thank you.
    [The prepared statement of Ms. Entmacher follows:]

   Statement of Joan Entmacher, Vice President and Director, Family 
             Economic Security, National Women's Law Center

    Chairman McCrery, Ranking Member Levin, and members of the 
Subcommittee, thank you for this opportunity to testify on behalf of 
the National Women's Law Center.
    My testimony will focus on the administrative and implementation 
issues that would arise at the point that money is to be paid out of 
private accounts created as part of Social Security. So far, most of 
the discussion of administrative issues has focused on how a new system 
of accounts would be created, how money would get into the accounts, 
and how investments during the working years would be managed. Far less 
attention has been paid to questions of how, if, and when money would 
be paid out from accounts to workers and their family members. \1\
---------------------------------------------------------------------------
    \1\ This testimony is informed by my work as a member of the expert 
study panel convened by the National Academy of Social Insurance to 
examine issues relating to the payment of benefits from individual 
accounts, two years of discussion with the thoughtful and knowledgeable 
experts on the panel and the NASI staff, and the report issued by the 
panel earlier this year: Reno, Graetz, Apfel, Lavery and Hill, eds, 
National Academy of Social Insurance, Uncharted Waters: Paying Benefits 
from Individual Accounts in Federal Retirement Policy (2005) 
[``Uncharted Waters '']. However, I am testifying today solely on 
behalf of the National Women's Law Center.
---------------------------------------------------------------------------
    The administrative issues that arise at the payout phase are 
critically important. If private accounts are supposed to make up for 
reductions in Social Security benefits that now provide basic income 
security for tens of millions of Americans and their families, a key 
question is, how can the proceeds of an individual account be converted 
into a secure, equitable, and adequate source of income for workers--
and their spouses and children--when workers retire, die, or are 
disabled?
    My testimony first discusses some of the issues that arise with 
payouts at retirement to an individual worker, then moves to some of 
the even more complex issues that arise because Social Security is a 
family insurance plan, not just a worker retirement program. \2\ The 
questions raised are difficult, and the tradeoffs presented are 
inevitable and painful, because an individual account simply cannot 
substitute for the range of insurance protections that it is possible 
to provide for workers and their families through Social Security.
---------------------------------------------------------------------------
    \2\ For more information about the family insurance benefits that 
Social Security provides, and their special importance to women, see 
Testimony of Nancy Duff Campbell, Co-President, National Women's Law 
Center, to the Subcommittee on Social Security, Committee on Ways and 
Means (May 17, 2005) [``Campbell Testimony''].
---------------------------------------------------------------------------
    There may be a temptation to think that resolving these issues can 
wait, because the retirement of workers who establish an account would 
be several years away. But that would be a serious mistake. Workers 
will need to know what the payout rules are before they can make a 
decision about whether or not to contribute to a private account. Well 
before they reach retirement age, some workers will divorce, some will 
become disabled, some will want access to the funds in ``their'' 
accounts, and some will die--all events that require decisions about 
payouts from private accounts, decisions that will have serious 
consequences for the wellbeing of Americans who rely on the safety net 
that Social Security provides for them and their families.

How Will Private Accounts Provide Secure Lifetime Income to Individuals 
        When They Retire?
    When a worker retires, dies, or becomes disabled, Social Security 
provides the worker and eligible family members with benefits that 
cannot be outlived, are adjusted annually for inflation, and do not 
fluctuate with the financial markets. In contrast, private accounts 
represent a limited pool of assets; retirees will face the challenge of 
deciding how to manage whatever they may have accumulated in an account 
to provide for their own retirement security, possibly for the 
financial security of a spouse or children, and for other goals. 
Administrative issues and financial risks do not end when a worker 
reaches retirement age. Retirees must make decisions in the face of 
multiple uncertainties: how long they or a spouse might live, how much 
prices will rise, and how the financial markets will perform.
    The challenges of managing private savings throughout retirement 
are not limited to accounts created within Social Security. But the 
stakes are higher--and the issues more complex--if private accounts are 
being relied upon to help make up for reductions in the basic income 
that Social Security provides to workers and their families.

      How will workers be protected against outliving the 
assets in a private account?

    With a plan that cuts Social Security benefits and relies on 
private accounts to provide basic income security to retirees for the 
rest of their lives, Congress will have to decide whether, when, and to 
what extent to require workers to purchase a life annuity to insure 
against the risk of outliving the assets in the account and becoming 
impoverished in old age. Alternatives to life annuities have been 
suggested as a way of preventing workers from exhausting their accounts 
too quickly, such as taking phased withdrawals over the period of their 
projected life expectancy. However, phased withdrawals do not provide 
the assurance of lifetime income; indeed, as the Congressional Research 
Service notes, on average, about 50 percent of those opting for phased 
withdrawals will live longer than expected and exhaust the funds in an 
individual account. \3\ Making the purchase of a life annuity 
mandatory, at least to assure income up to some specified level, would 
reduce income insecurity; broaden the annuity pool and increase annuity 
payments on average; and reduce marketing expenses. \4\
---------------------------------------------------------------------------
    \3\ Laura Haltzel, Congressional Research Service Report, Social 
Security Reform: President Bush's Individual Accounts Proposal 7 (April 
25, 2005).
    \4\ See Uncharted Waters.
---------------------------------------------------------------------------
    But requiring the purchase of an annuity poses tradeoffs. An 
annuity requires payment up front; once it is purchased, the assets 
used to buy it are no longer available. An annuitization requirement 
would reduce the risk of outliving assets, but also would reduce the 
control and choice workers could exercise over the accounts. An 
annuitization requirement also could eliminate the possibility of a 
bequest; the Bush Administration estimates that under its plan, 15 
percent of all retirees and 30 percent of retirees with lower lifetime 
incomes would have to spend all the assets in their accounts to bring 
the combination of reduced Social Security benefits and payments from 
the account up to the poverty level for their lifetimes, leaving 
nothing in their accounts for discretionary spending or a bequest. \5\
---------------------------------------------------------------------------
    \5\ [5] Associated Press, ``Survivor Benefits Face Cut, Official 
Says,'' (May 12, 2005).

      Will discrimination on the basis of gender be prohibited 
---------------------------------------------------------------------------
in annuities purchased from private accounts?

    Social Security pays monthly benefits on a gender-neutral basis. In 
contrast, in the private annuity market, if a man and woman purchase a 
life annuity with the same amount of money at the same time, the woman 
will get lower monthly payments for life.
    Federal law already bans differential pricing and benefits in group 
annuities or pensions that are part of an employment relationship. \6\ 
If Congress creates private accounts as part of Social Security, it 
must prohibit gender discrimination in annuities marketed to those with 
private accounts. An effective prohibition on discrimination will 
require more than passing legislation; regulatory oversight be needed 
to avoid the design and marketing of annuity products specifically to 
men or women to avoid the effect of the uniform pricing requirement. 
\7\
---------------------------------------------------------------------------
    \6\ See, e.g., City of Los Angeles Department of Power and Water v. 
Manhart, 435 U.S. 702 (1978).
    \7\ See Uncharted Waters at 78-80.

      How will annuities from private accounts provide 
---------------------------------------------------------------------------
protection against inflation?

    Social Security provides payments for life that are adjusted 
annually to keep up with increases in the cost of living. No private 
annuities currently offer full protection against inflation, and 
experts believe they are unlikely to evolve without the substantial 
involvement of the federal government, even if the market for private 
annuities expanded with the establishment of private accounts. \8\
---------------------------------------------------------------------------
    \8\ See Uncharted Waters at 85-86.

      How will workers be assured that annuity payments from a 
---------------------------------------------------------------------------
private company will continue for life?

    A person who purchases a life annuity pays the price up front, in 
exchange for a contractual promise to make payments for the life of the 
purchaser (and survivor, in the case of a joint and survivor annuity). 
The purchaser counts on the company providing the annuity to make good 
on the promised payments for years to come.
    Today, life annuities are provided by life insurance companies that 
are regulated by the states. States are responsible for setting 
solvency standards, monitoring compliance with those standards, and 
providing some protection in case an insurance company defaults. Every 
state has a ``guaranty fund'' to deal with defaults--but unlike the 
Federal Deposit Insurance Corporation, which ensures bank deposits, or 
the Pension Benefit Guarantee Corporation, which ensures defined 
benefit pensions, state guaranty funds are not pre-funded at all. \9\ 
In the event of a default, states make assessments against other 
companies doing business in the state. States also make their own rules 
about who is protected, and to what extent, by the state guaranty. \10\ 
For example, a retiree who purchases an annuity from a company doing 
business in state A, then moves to State B, may not be entitled to 
payments from State A's guaranty fund if the company defaults. \11\ The 
risk that a large insurance company will default is not hypothetical, 
as the case of the Executive Life Insurance Company of California and 
its New York subsidiary shows; as this case also shows, policy holders 
may suffer substantial losses in case of default. \12\
---------------------------------------------------------------------------
    \9\ Id. at 76-82.
    \10\ Ibid.
    \11\ Ibid.
    \12\ Ibid.
---------------------------------------------------------------------------
    If Congress creates private accounts that are designed to replace 
Social Security income, the federal government will have to intervene 
in the annuities market, and probably act as guarantor, to make sure 
that Americans get the payments they are counting on. But Social 
Security already provides annuities that are adjusted for inflation, 
nondiscriminatory, and secure; it would be far more efficient and 
effective simply to protect and strengthen Social Security.

      When will retirees be required to purchase the annuity?

    Investment risk does not end when a worker has stopped contributing 
to an account is ready to start drawing retirement income. Indeed, 
converting account assets to a life annuity magnifies market risk. The 
lifetime income that an annuity provides will be determined by the 
value of the assets and interest rates at the moment of conversion--and 
these can fluctuate greatly over short periods of time. The drop in the 
stock market and interest rates between 2000 and 2003 meant that a 
retiree purchasing an annuity in 2003 would have 60 percent less income 
than a retiree who purchased an annuity in 2000--after a lifetime of 
making similar contributions and investment choices. \13\
---------------------------------------------------------------------------
    \13\ [13] Calculations by Gary Burtless, Senior Fellow, Brookings 
Institution, quoted in Julie Kosterlitz, ``Cracking the Nest Egg,'' 
National Journal (April 22, 2005).
---------------------------------------------------------------------------
    These market fluctuations are a source of concern--and hardship--
for workers in or near retirement who look to their IRAs or 401(k)s to 
supplement their Social Security benefits during a market downturn. 
When private accounts are expected to replace part of the basic income 
provided by social insurance system, the risks are even greater.
    In Chile, interest rates fell dramatically in 1996 and 1997, 
greatly reducing the lifetime payments workers would get from their 
accounts. Under Chile's partially privatized system, which requires 
workers to annuitize their accounts when they claim retirement 
benefits, workers who had planned to retire--some of whom had already 
been pushed out of their jobs when they reached retirement age--were 
simply told, ``Don't.'' \14\ In the United Kingdom, protests by workers 
who were required to buy annuities under the UK's partially privatized 
system at a time when interest rates (and thus annuity payments) were 
down led the government, in 1994, to allow workers to defer their 
annuity purchase until age 75. \15\ But many workers don't have 
sufficient other resources to meet their needs while they wait for 
market conditions to improve. And there is no guarantee of improvements 
in the short term; indeed, in 1999, The Financial Times reported that 
the income obtainable from a private account had dropped 16 percent in 
the course of a year. \16\
---------------------------------------------------------------------------
    \14\ Ibid.
    \15\ Ibid.
    \16\ Ibid.
---------------------------------------------------------------------------
    Giving workers flexibility in the timing of an annuity purchase 
raises new issues and administrative challenges; indeed, an extended 
time frame could essentially negate an annuitization requirement. Many 
workers would be unable to get by on their greatly reduced Social 
Security benefits while they wait for market conditions to improve; 
but, if they are allowed to start withdrawing assets from the accounts 
while they wait to annuitize, there will be less left in the account to 
assure lifetime income for workers and their spouses.
    Deciding when to annuitize carries lifetime financial implications; 
to take advantage of the added flexibility, workers will need 
additional financial counseling (not that investment counselors are 
necessarily successful at predicting short-term changes in asset values 
and interest rates). Gradual annuity purchases could spread the risk of 
interest rate fluctuations over a longer period--but would not 
eliminate the risk, and would add to administrative costs and make 
payment levels uncertain. \17\
---------------------------------------------------------------------------
    \17\ Uncharted Waters at 70.
---------------------------------------------------------------------------
    Workers deciding when to annuitize could be subject to conflicting 
family pressures: children and other possible heirs may seek to delay 
the purchase of an annuity which would leave little or nothing for them 
to inherit; a spouse concerned about the depletion of assets may prefer 
the purchase of a joint and survivor annuity. And workers with lower-
than-average life expectancies would probably seek to delay the 
purchase of an annuity as long as possible; life annuities are not a 
good deal for people with shorter life expectancies. But if individuals 
with shorter life expectancies can effectively opt out of the annuity 
pool by delaying the purchase of an annuity for years, those who 
purchase annuities can expect lower payments.

How Will Private Accounts Help Make Up for Cuts in the Family Insurance 
        Benefits that Social Security Provides?
    The various family insurance benefits that Social Security 
provides, and their importance--especially to women and their 
families--are described in the May 17, 2005 testimony to this 
Subcommittee by Nancy Duff Campbell, Co-President of the National 
Women's Law Center. Her testimony also explains why private accounts 
cannot match the benefits that Social Security provides not just for 
retired workers, but for workers if they are disabled, and for the 
spouses and children of workers when workers retire, die, or are 
disabled.
    The best way to protect the safety net that Social Security 
provides for women and their families is to reject plans that would 
create private accounts out of Social Security, and work instead to 
strengthen and improve Social Security.
    However, if this Subcommittee is considering private accounts 
plans, it must consider how the rules governing the payouts from 
private accounts might mitigate--or exacerbate--the harm to family 
members and disabled workers from the plan's cuts to Social Security 
benefits--even though it will be impossible to protect them fully.

      Will married workers be required to purchase a joint and 
survivor annuity?

    Social Security assures the spouse of a retired worker a benefit 
equal to 50 percent of the worker's benefit; it assures the surviving 
spouse a benefit of 100 percent, assuming both spouses retire at full 
retirement age. Divorced spouses and divorced surviving spouses, if 
married to the worker for at least ten years, are entitled to the same 
benefits as current spouses. Social Security spousal benefits are paid 
in addition to the worker's benefit; they do not reduce the benefit 
that the worker, or the current spouse (or ex-spouse) of the worker 
receives.
    Because Social Security spousal benefits are calculated based on 
the worker's Social Security benefit, cuts in Social Security benefits 
for retired workers mean cuts in spousal benefits as well. Private 
accounts plans may cut spousal benefits twice: first, as part of a 
general benefit reduction applicable whether or not a worker has chosen 
to contribute to a private account (even if these cuts are designed to 
exempt workers with very low earnings, widows with very low income may 
face benefit cuts, because their benefits are based on the record of a 
worker who had earnings above the minimum level). Second, if a plan 
cuts benefits specifically for workers who contribute to a private 
account, benefits for the retired spouse and widow of a worker who 
contributes to an account are also likely to be cut further. \18\
---------------------------------------------------------------------------
    \18\ See Uncharted Waters at 174; no private accounts plan with an 
offset applied it only to the accountholder's benefit.
---------------------------------------------------------------------------
    In view of the importance of spousal benefits to women, now and in 
the future, it is disturbing that the Administration has so far failed 
to say whether its private accounts plan would require married workers 
to purchase a joint and survivor annuity to help make up for reductions 
in Social Security spousal benefits. Without such a requirement, a 
married worker (call him Michael) could convert all the assets in his 
account to a single life annuity, leaving his widow (call her Sarah) 
with nothing from the account: no household income from his annuity 
payments, no survivor payments for herself, and no inheritance. Yet she 
may be facing a deep reduction in her Social Security benefits 
specifically because Michael contributed to a private account. \19\
---------------------------------------------------------------------------
    \19\ For a further discussion of how annuitization choices could 
affect benefits for a couple, see Campbell Testimony.
---------------------------------------------------------------------------
    In addition to deciding whether to require the purchase of joint 
and survivor annuities, Congress also must decide their size and form. 
What percentage of the payment to the annuity purchaser should be 
required to be provided for a surviving spouse: 50, 67, 75 or 100 
percent? A higher survivor benefit means more income security for the 
widowed spouse--but lower payments when both are alive. If the spouse 
is several years younger, payments during the life of the annuity 
purchaser would be lower still.
    Requiring survivor annuities will provide additional protection to 
surviving spouses, but they cannot respond to changes in marital status 
as Social Security can. Someone who enters retirement as a single 
individual, purchases a single life annuity, then marries, cannot 
change it to a joint and survivor annuity. Someone who is widowed or 
divorced shortly after retirement cannot change a joint and survivor 
annuity to a single life annuity with higher payments. And whether one 
is widowed right before or after annuitization could make a big 
difference in what the widowed individual receives. \20\
---------------------------------------------------------------------------
    \20\ Uncharted Waters at 62-66.

      Will waivers of the right to a joint and survivor annuity 
---------------------------------------------------------------------------
be permitted?

    There is no procedure in Social Security for a spouse, surviving 
spouse, or divorced spouse to waive the right to spousal benefits. Nor 
is there any need for such a waiver procedure; because the payment of 
spousal benefits in Social Security does not reduce benefits for the 
worker or current spouse, the issue does not arise.
    Federal law requires that, in defined-benefit pension plans, the 
default pension payment to a married worker must be in the form of a 
joint and at least 50 percent survivor annuity, unless the spouse 
consents, in writing and before a notary, to a less generous or no 
survivor annuity.
    In the context of retirement plans that are designed to provide 
tiers of income on top of the basic Social Security benefit, allowing 
for informed waivers of the right to a joint-and-survivor payment 
balances the competing goals of protecting the rights and needs of 
spouses and giving couples flexibility to make the financial 
arrangements that best meet their goals. The spouse with the right to a 
survivor payment may have other financial resources available, or both 
spouses may prefer to receive higher income when both are alive, even 
at the risk that the surviving spouse will only have Social Security to 
rely on.
    However, in the context of a plan that cuts basic Social Security 
benefits, waivers would raise more difficult issues. The benefit cuts 
under some proposals could leave many workers with retirement income 
far below scheduled benefit levels, even if they maximized the payments 
from a private account by purchasing a single life annuity. \21\ 
Providing a survivor annuity for a spouse would reduce the worker's 
benefits even further. Workers might pressure their spouses to waive 
the survivor benefit; even without undue pressure, some spouses might 
waive their right to a survivor benefit to ensure a modest income for 
the couple now--and worry about the future later. If waivers are 
permitted, information about the options and their implications would 
need to be provided to both spouses, adding to administrative 
responsibilities and costs.
---------------------------------------------------------------------------
    \21\ See Campbell Testimony.

      Will workers be able to leave the account to anyone, or 
---------------------------------------------------------------------------
will surviving spouses have the right to inherit?

    The President has said repeatedly that under his plan, workers 
could leave an account to anyone, never suggesting that a surviving 
spouse would have the right to inherit. The Administration also has 
confirmed that its proposal would reduce benefits for widows and 
surviving children. \22\ Congress must decide whether a surviving 
spouse will have a right to inherit account assets, and the nature of 
that right. Would it apply to all the assets in the account? Could it 
be waived?
---------------------------------------------------------------------------
    \22\ Associated Press, ``Survivor Benefits Face Cut, Official 
Says,'' May 12, 2005.
---------------------------------------------------------------------------
    Other questions would arise if a spouse inherits account assets, 
whether by right or designation. If a worker died young and left the 
account to his widow, would she have immediate access to whatever small 
amount the account might contain to help support her family and 
supplement their reduced survivors' benefits? Or would she have to save 
them for her own retirement, as several plans propose? With Social 
Security, a widow is eligible for benefits based on the deceased 
husband's work record both while she is raising their children and at 
retirement--but funds in a private account can only be used once.
    If account assets went to someone other than a surviving spouse, 
would those heirs have immediate access to the funds they inherit?
    Would children have any inheritance rights in a parent's account, 
especially if a private accounts plan cuts their survivor benefits? If 
the children live in a different household than the widowed spouse--a 
not uncommon situation--how would their interests be balanced? What 
would happen if a child made a claim after all the assets in an account 
had been distributed--as could happen, if the deceased parent had not 
had contact with the child for several years? Would all children have 
inheritance rights, or only minor children and disabled adult children 
entitled to benefits on the parent's work record? If the latter, would 
the child's share be related to the number of years the child would be 
reliant on reduced survivor benefits, so that a toddler or disabled 
adult child would be entitled to a greater share of a parent's account 
than a school age child? These questions highlight the impossibility of 
expecting a private account to make up for cuts in the life insurance 
benefits that Social Security provides to surviving spouses, surviving 
divorced spouses, and children.
    If a plan provides that workers must pay back the money they 
contributed to a private account, with interest, out of a reduction in 
their Social Security benefits, would the widow inherit this debt along 
with any assets in the account? If so, would other heirs have the same 
obligation? Transferring a debt along with account assets would greatly 
diminish the value of any inheritance; in fact, if the account has done 
very poorly, the bequest could be a net liability. On the other hand, 
Social Security's finances will suffer if funds diverted from Social 
Security to private accounts are not reimbursed by those who benefit 
from them.

      How would accounts be divided at divorce?

    Social Security provides benefits for divorced spouses and divorced 
surviving spouses who have been married for at least ten years. 
Benefits for divorced spouses are calculated in the same way as 
benefits for spouses and surviving spouses, based on the full work 
history of the higher-earning spouse, not just the earnings during the 
period of the marriage.
    The question of how accounts should be divided at divorce raises 
fundamental legal questions. \23\ If private accounts are considered 
``property,'' they might be subject to state laws concerning marital 
property--which differ between community property and common law 
states, and among the states in each group. This would lead to 
different rights for spouses in community property and common law 
states, and for couples or individual spouses who move from state to 
state. Congress could create a uniform system of federal rules 
governing the division of accounts at divorce and other spousal rights 
issues--or explicitly certain issues to the operation of state law--but 
it must explicitly resolve the issue of whether state or federal law 
will govern.
---------------------------------------------------------------------------
    \23\ For an explanation and further discussion, see Uncharted 
Waters at 120-137.
---------------------------------------------------------------------------
    There are several possible approaches for dividing accounts between 
spouses. \24\ One approach would involve contribution splitting during 
marriage--that is, contributions to accounts made by either spouse 
during the marriage would be shared equally between the spouses' 
accounts. Because contributions would have been shared at the front 
end, there might be no further division of account assets at divorce. 
Another approach would divide the assets that accumulated in the 
accounts during the course of the marriage equally between the spouses. 
Or this 50-50 division could be the default, but other allocations 
could be allowed by agreement of the parties or a court order.
---------------------------------------------------------------------------
    \24\ Six options are identified in Uncharted Waters at 127.
---------------------------------------------------------------------------
    If a plan provides that workers must pay back the money they 
contributed to a private account, with interest, out of a reduction in 
their Social Security benefits, Congress also must decide if and how 
the debt, as well as account assets, should be divided at divorce. \25\
---------------------------------------------------------------------------
    \25\ See Uncharted Waters at 178-179.
---------------------------------------------------------------------------
    Implementing a system for dividing private accounts at divorce, 
even one that calls for the automatic 50-50 division of accumulations 
during the marriage at the time of divorce, will require new reporting, 
verification, and dispute-resolution mechanisms well beyond those 
needed to administer the current Social Security system. Social 
Security needs to review evidence of marriage, marriage duration, and 
divorce only at the point an individual applies to receive Social 
Security benefits as a divorced spouse or divorced surviving spouse. 
Approximately 100,000 new applications for Social Security in 2001 
involved evidence of divorce--but that is just one-tenth of the roughly 
one million divorces that occur each year in the United States. \26\
---------------------------------------------------------------------------
    \26\ Uncharted Waters at 133.
---------------------------------------------------------------------------
    Benefits for eligible divorced spouses are based on the full work 
history of the higher-earning spouse, so Social Security needs no 
additional information to calculate benefits for an eligible divorced 
spouse. To divide contributions and accumulations during only the 
period of the marriage would require historical records of year-by-year 
(or quarter-by-quarter) contributions and investment earnings, along 
with evidence and dates of marriage and divorce.
    For the entity administering a system of private accounts to obtain 
reliable information about marital status is no simple matter; no 
national registry of marriage and divorce information currently exists. 
\27\ Creating such a registry would require new resources and raise new 
confidentiality issues. Individuals could be asked to report changes to 
their marital status. But accountholders might fail to report a 
marriage, because they would prefer that a spouse not receive funds at 
their expense. Social Security minimizes conflicts and disputes, 
because the payment of benefits to a divorced spouse does not reduce 
payments for the worker or his or her current spouse.
---------------------------------------------------------------------------
    \27\ See Uncharted Waters at 134-135.
---------------------------------------------------------------------------
    Once an account is divided at divorce, the assets are gone. The 
spouse who is the net loser in the division will have less in an 
account to supplement reduced retirement benefits or provide for 
children or a new spouse.

      What if only one spouse participates in a private 
account?

    If participation in a private account is voluntary, there will be 
spouses who have made different decisions with respect to 
participation. Such differences cannot be eliminated by requiring 
married couples to make the same choice (and deciding whether 
participation or nonparticipation is the default if they cannot agree). 
Many Americans enter the labor force before they get married, and, 
under some plans, would make an irrevocable choice at that time; new 
marriages and remarriages would produce additional mismatches.

Conclusion
    This testimony raises some important issues about the critical 
payout phase of a private accounts plan. \28\ Some of these questions 
have been raised before--but many still have not been answered by the 
Administration. Before this Committee considers proposals that would 
radically change a program on which millions of Americans rely, it must 
address the fundamental questions: how would an account that fluctuates 
with the market provide a secure, basic retirement income, and how 
would an individual account make up for reductions in the family 
insurance benefits that Social Security provides.
---------------------------------------------------------------------------
    \28\ For a further discussion of these issues, and the many issues 
concerning payments to people with disabilities and their families 
which this testimony does not address, see Uncharted Waters.

---------------------------------------------------------------------------
                                 

    Chairman MCCRERY. Thank you. Mr. Cavanaugh.

 STATEMENT OF FRANCIS X. CAVANAUGH, FORMER EXECUTIVE DIRECTOR 
    AND CHIEF EXECUTIVE OFFICER, FEDERAL RETIREMENT THRIFT 
                        INVESTMENT BOARD

    Mr. CAVANAUGH. Thank you, Mr. Chairman, Members of the 
Subcommittee. I welcome this opportunity to discuss the 
administration of Social Security personal retirement accounts. 
My comments will focus on the President's current proposal for 
such accounts. The critical question, of course, is cost. 
Individual accounts are proposed to provide a higher investment 
return than would be realized by the Social Security Trust 
Fund. On this basis, individual accounts would not be feasible 
for the 68 million employees of 98 percent of the businesses in 
the United States. That is the 5.6 million small businesses 
with fewer than 100 employees. It just would not work.
    To understand the cost to small business, we must first 
understand why 85 percent of them do not now have retirement 
plans. A major reason is that the 401(k) industry has found 
that it cannot profitably provide services for a company for 
less than approximately $3,000 a year. This is after 10 years 
of competing with each other and the economies of scale. So, 
even though they enjoy these economies of scale when combining 
thousands of employers in their centralized computer system, 
they cannot reduce the cost below this. Further, significant 
economies of scale would not be realized by a central Federal 
TSP-type agency because of the fixed costs of reaching out to 
millions of small businesses. Nor can we assume that a new 
central government agency would be more efficient than the 
major 401(k) providers who now serve this market. Thus, the 
cost per employee of a company with ten employees would be 
$300, or 30 percent of the President's proposed initial annual 
individual account contribution of $1,000. Most U.S. companies 
have fewer than ten employees.
    Accordingly, the initial expense ratio for employees of the 
average-size business would be more than 3,000 basis points, or 
100 times the Administration's estimate of 30 basis points. 
This is not conjecture, Mr. Chairman, this is based on what the 
market is actually doing and saying right now. Obviously, since 
the administrative costs of individual accounts would exceed 
their estimated returns from investments, a substantial 
government subsidy would be necessary to make the individual 
accounts attractive to employees of small business. If all 
Social Security taxpayers eventually participated in the 
individual account program, the administrative costs would be 
more than $46 billion a year. In addition to the above costs, 
which are based on what the current providers are actually 
charging for establishing and servicing 401(k) plans, there are 
overwhelming practical obstacles to modeling individual 
accounts on the TSP or private 401(k) plans.
    First, the TSP is administered by just one employer, the 
U.S. government, with an extensive network of agency personnel 
payroll and systems staff to provide the essential employee 
education, retirement counseling, payroll deduction, timely 
fund transfers, error correction functions, and so on. These 
essential employer services in 401(k) plans could not possibly 
be performed by small business employers or by a new TSP-like 
central agency. It cannot be done. Second, the TSP is 
computerized, like all other large plans, with investments made 
promptly after contributions are deducted from the employee's 
paycheck. With individual accounts it would be up to 22 months 
after payday under current SSA procedures before the individual 
accounts could be credited and invested. Third, the TSP is 
balanced to the penny every day. The Social Security system is 
never balanced. Each year there are billions of dollars of 
unreconciled discrepancies. Fourth, the TSP and the Federal 
employee agencies have a very effective system of 
communication. The TSP mailings consistently have reached more 
than 99 percent of employees, but 25 percent of Social Security 
mailings are returned as undeliverable.
    Since individual accounts are certainly not feasible for 
employees of small business, the only practical way to give 
higher returns for Social Security beneficiaries generally, 
including small business employees, is to invest part of the 
Social Security Trust Fund in equities. The likely increase in 
trust fund earnings would be an effective way to help maintain 
the solvency of the trust fund. Every State in the United 
States has authorized public retirement fund investment in 
stocks which can now be done through broad-based index funds, 
which avoid the problem of direct government control over a 
particular company. There is no longer any reason not to invest 
the trust fund partially in equities. As is shown in the chart 
on page eight of my statement, there is even less government 
influence over private companies under the trust fund 
alternative than under the TSP or the Administration's plan.
    In conclusion, the Administration's plan for universal, 
individual accounts is not feasible. The way for the Social 
Security system to capture the higher returns available from 
investments in stock is to diversify the Social Security Trust 
Fund investments. The trust fund alternative, compared to 
individual accounts, would be less disruptive of financial 
markets, would save tens of billions of dollars a year in 
administrative costs, and could be effective virtually 
immediately rather than the 2009 starting date proposed for 
individual accounts. The multi-trillion-dollar transition costs 
of individual accounts would be avoided. The additional trust 
fund earnings would go a long way to strengthening Social 
Security finances and would thus reduce, if not eliminate, the 
need for significant tax increases or benefit reduction. Thank 
you for your attention. I would be happy to answer any 
questions.
    [The prepared statement of Mr. Cavanaugh follows:]

Statement of Francis X. Cavanaugh, former Executive Director and Chief 
     Executive Officer, Federal Retirement Thrift Investment Board

    Mr. Chairman and Members of the Subcommittee:
    I welcome this opportunity to discuss the administration of Social 
Security personal retirement accounts (PRA). My comments will focus on 
the Administration's current proposal for such accounts.
    I am a public finance consultant, but I speak only for myself. I 
have no clients with an interest in Social Security individual 
accounts. From 1986 until 1994, I was the first Executive Director, and 
thus the chief executive officer, of the Federal Retirement Thrift 
Investment Board, the agency that administers the Thrift Savings Plan 
(TSP) for federal employees. Before that, I was a financial economist 
in the Treasury Department for 32 years, and was the senior career 
executive responsible for developing federal borrowing, lending, and 
investment policies, including those for the Social Security and other 
federal trust funds.

The Administration's Proposal
    While there is no specific proposal before your committee, the 
Administration's current broad proposal, according to White House 
statements and press reports, provides a basis for at least a 
preliminary analysis of its administrative feasibility.
    The following features of the Administration's approach would have 
significant impacts on its feasibility:

      PRAs would be voluntary for all Social Security taxpayers 
under age 55, but would be mandatory for employers of employees who 
chose PRAs.
      A major purpose of PRAs would be to encourage savings by 
young and low-income workers and employees of small businesses who do 
not now have 401(k)s or other pension plans.
      The maximum amount of an individual's initial annual 
contribution to a PRA would be $1,000, which would increase by $100 a 
year, to 4 percent of pay eventually. It would take more than 30 years 
for the highest income individuals to be able to contribute the full 4 
percent of pay.
      Eligible investments for PRAs would be Treasury 
securities and stock and bond index funds, which would be similar to 
eligible investments of the TSP.
      PRAs would be centrally managed, apparently by a TSP-like 
agency with a part-time board, appointed by the President with the 
advice and consent of the Senate, and a full-time executive director 
and CEO appointed by the board. Following the TSP model, the board 
members and the executive director would be independent of the 
Administration, and would be fiduciaries required to act solely in the 
interests of the holders of the PRAs and their beneficiaries.
      Unlike contributions to 401(k)s or to the TSP, PRA 
contributions would not be eligible for matching contributions or 
exclusion from taxable income, and loans or withdrawals before 
retirement would not be permitted.

Cost Analysis
    A critical question, of course, is costs. PRAs are proposed to 
provide a higher investment return than would be realized by the Social 
Security trust fund. Thus PRAs would not be feasible if their 
administrative costs were so high as to offset the advantage of 
diversified investments in stocks and other securities that yield more 
than the Treasury securities in the Social Security trust fund.
    The Administration assumes that PRAs would earn an average 
investment return of 4.9% after inflation, and that administrative 
costs of.3%, that is, 30 basis points, would reduce the net return to 
4.6%, or 1.6% more than the assumed net return of 3% on the Treasury 
securities in the Social Security trust fund. Thus, if one accepts the 
Administration's assumptions, PRAs would outperform the trust fund 
investments so long as the administrative costs were less than 1.9%. In 
my view and that of many other economists, the 4.6% assumption is much 
too high; indeed, the Congressional Budget Office's estimate of the net 
return is reportedly only 3.3%.
    The Administration's estimate of 30 basis points is optimistically 
low; even the Cato Institute, a leading advocate of individual 
accounts, estimates PRA expenses at 55 basis points. Yet this higher 
estimate is also too low. Like so many others I have heard, these 
estimates are based mainly on experience with large 401(k)s for large 
organizations, like the TSP, \1\ with economies of scale and 
comprehensive payroll, personnel, and computerized systems support. 
They have little relevance to the likely costs of a universal system of 
PRAs. More than 85 percent of the 5.6 million small business employers 
in this country offer no pension plans at all and, accordingly, have 
none of the administrative apparatus to service them.
---------------------------------------------------------------------------
    \1\ The administrative cost, or expense ratio, of the TSP is 6 
basis points.
---------------------------------------------------------------------------
    To understand the costs of bringing PRAs to employees of small 
businesses, we must first understand why 85 percent of them do not now 
have retirement plans for their employees. Fortunately, the 401(k) 
industry has already done part of the job for us. Companies like 
Citigroup, Fidelity Investments, Merrill Lynch, State Street 
Corporation, and T. Rowe Price have been competing for two decades to 
provide investment, record keeping, counseling, and other 401(k) plan 
services to small businesses. They have found that they cannot 
profitably provide these services for a company for less than 
approximately $3,000 a year, even though they have for years enjoyed 
economies of scale from serving thousands of employers in their 
centralized computer systems. \2\ Further significant economies of 
scale would not be realized by a central TSP-type agency, because there 
would still be millions of small businesses or workplaces to be 
reached. Nor can we assume that a new central government agency would 
be more efficient than the major 401(k) providers who now serve this 
market.
---------------------------------------------------------------------------
    \2\ Francis X. Cavanaugh, ``Feasibility of Social Security 
Individual Accounts,'' AARP Public Policy Institute, Washington, D.C., 
Sept. 2002, pp. 4-6. The $3,000 charge is still common today. See ``Big 
Fees Hit Small Plans: Costs Take Huge Toll on Retirement Accounts of 
Firms With Fewer Than 50 Employees,`` Wall Street Journal, Oct. 31, 
2004, p. D1.
---------------------------------------------------------------------------
    Thus the cost per employee of a company with 10 employees would be 
$300, or 30 percent of the President's proposed annual PRA contribution 
of $1,000--and most U.S. companies have fewer than 10 employees. \3\
---------------------------------------------------------------------------
    \3\ See generally U.S. Department of Labor, Pension and Welfare 
Benefits Administration, ``Study of 401(k) Fees and Expenses,''Apr. 
13,1998. The study found that average charges by 17 major 401(k) 
providers for plans with 100 participants and $2 million in assets 
ranged from $114 to $428 per participant, and averaged $264. Id. at 51. 
Charges obviously would be much higher for much smaller plans.
---------------------------------------------------------------------------
    Even the largest business that is classified as a ``small 
business,'' one with 100 employees, would therefore have an expense 
ratio of at least 3 percent, which would be ten times the 
Administration's estimate of 30 basis points. And for the 60 percent of 
employers in this country that have fewer than 5 employees, the initial 
expense ratio would be more than 60 percent, that is, 6,000 basis 
points. In fact, commercial 401(k) providers routinely discourage small 
businesses from establishing 401(k) plans if they have fewer than 10 
employees and, in some cases, fewer than 25 employees.
    Obviously, substantial and continuing government subsidies would be 
necessary to make PRAs attractive to employees of small businesses. If 
all Social Security taxpayers participated in the PRA program, the 
administrative costs would be more than $46 billion a year (155 million 
participants times more than $300 per account), which would be a 
subsidy to PRA administrators for performing an uneconomic function. 
These figures are reinforced by a number of studies, including those 
cited in a review of administrative costs by the Employee Benefit 
Research Institute. \4\
---------------------------------------------------------------------------
    \4\ See, e.g., Employee Benefit Research Institute, Issue Brief No. 
23, Nov. 1998. See also Ellen E. Schultz, ``Poodle Parlor Retirement 
Plans,'' Wall Street Journal, Nov. 13, 1998, p. C1.
---------------------------------------------------------------------------
    I recommend that your committee secure the testimony of individuals 
from financial institutions that are actually providing 401(k) services 
to the nation's businesses, large and small. Give them a specific set 
of assumptions to cost out that reflects the makeup of our country's 
5.7 million employers subject to Social Security--of which 98% are 
small business employers of 68 million employees. \5\ Then and only 
then will you know whether the Administration's proposal--or anything 
similar--will produce reasonable net investment returns, or, in the 
alternative, how much of a government subsidy would be necessary to 
achieve them.
---------------------------------------------------------------------------
    \5\ Patrick Purcell, Congressional Research Service, ``Social 
Security Individual Accounts and Employer-Sponsored Pensions,'' Feb. 3, 
2005, pp. 3, 5.
---------------------------------------------------------------------------
Critical Administrative Problems
    In addition to the above costs, which are based on what the current 
providers are actually charging for establishing and servicing 401(k) 
plans, there are overwhelming practical obstacles to the creation and 
maintenance of PRAs. Because President Bush seemed to idealize the 
Thrift Savings Plan--the largest of all 401(k)-type plans--as the model 
for PRAs in his February 2005 State of the Union message--and because 
many others have done so as well--I would like to point out the 
considerable dissimilarities between the TSP and the Administration's 
proposal. (Most of these dissimilarities would hold true for a 
comparison between any large corporate 401(k) plan and the proposal.)
    Too Many Small Employers. The TSP is administered by just one 
employer--the U.S. Government--with extensive personnel, payroll, and 
systems staffs to provide the essential employee education, retirement 
counseling, payroll deduction, timely funds transfers, and error 
correction functions. The Thrift Investment Board is only a wholesaler 
of services; the federal employing agencies deal with the individual 
employees participating in the plan. In fact, the TSP statute directs 
the Office of Personnel Management to provide for the training of TSP 
counselors for each federal agency.
    The Administration's plan is intended to reach all employees, but 
it makes no provision for the performance of what are now essential 
employer functions in 401(k) plans. They could not possibly be 
performed by small business employers who are now responsible only for 
the relatively simple payroll deduction and transmission of Social 
Security taxes to the IRS. Since most businesses have fewer than ten 
employees, they do not have the experience or resources to support the 
new plan. These are barbershops, beauty salons, garages, restaurants, 
laundries, lawn services, households, nanny services, and other very 
small businesses that could not be expected to meet the high fiduciary 
standards required of those responsible for educating and counseling 
employees, for presenting a new plan in the context of the employer's 
existing pension or other benefits, and for the timely and accurate 
transfer of funds for investment. The new TSP-like agency obviously 
could not provide such employer-type services to deal with tens of 
millions of diverse employees, either directly or on a contract basis.
    Consider, as but one example of several profound administrative and 
legal issues, that about 650,000 businesses go out of business each 
year. By whom and how would the enforcement of contributions by 
delinquent or bankrupt employers be prosecuted? (Judicial remedies for 
denial of TSP benefits must, in general, be pursued by the affected 
individual TSP participant in the federal court system.) For that 
matter, by whom and how would breach-of-fiduciary-duty suits be brought 
against ``mom-and-pop'' fiduciaries? Can the employer of a housekeeper 
or a manicurist be expected to exercise the ``care, skill, prudence, 
and diligence'' demanded of every 401(k) plan fiduciary by current law? 
\6\ What would be the measure--and the limit--of their personal 
liabilities, say, for untimely or inaccurate investment of their 
employees' contributions? These questions only scratch the surface of 
the inevitable pathology of plan administration--pathology that, even 
if represented in small percentages among 155 million Social Security 
participants, would result in enormous absolute numbers.
---------------------------------------------------------------------------
    \6\ See Employee Retirement Income Security Act (ERISA), 29 U.S.C. 
Sec. 1104(a); Federal Employees' Retirement System Act (FERSA), 5 
U.S.C. Sec. 8477(b)(1).
---------------------------------------------------------------------------
    Untimely Investments. The TSP is computerized, like all other large 
plans, with investments made for each employee's account on the same 
day that contributions are deducted from the employee's paycheck. 
Social Security taxes are deducted on paydays, but many small 
businesses send them to the IRS only once each quarter. In 2003, 72 
percent of employer reports to the Social Security Administration were 
submitted on paper. Moreover, individual Social Security taxpayers are 
identified only once each year, with their employer's annual income tax 
filings; and it would be up to 22 months after payday, under current 
SSA procedures, before individual PRAs could be credited.
    Furthermore, the Administration's proposal is to pay PRAs the same 
annual return, regardless of when contributions were actually made 
during the year. Thus a contribution in January would not earn any more 
than a contribution of a similar amount in December. During a year of 
highly volatile markets, the attempted explanation of this provision to 
millions of outraged participants with irregular tax payments, because 
of illness, seasonal, temporary, or other periods of unemployment, 
would be a daunting challenge to the plan's telephone counselors.
    Unbalanced Accounts. The TSP is balanced to the penny every day. 
The Social Security system is never balanced. Each year there are 
billions of dollars of unreconciled discrepancies between Social 
Security taxes paid to the IRS and reported to the SSA. These 
discrepancies are tolerated because they generally have little impact 
on the ultimate calculation of employee benefits. Such discrepancies 
are never tolerated by financial institutions responsible for timely 
investment of individual funds. Theoretically, PRA contribution errors 
might be largely corrected by a rigorous examination of employer 
records. Yet the error correction procedures, including retroactive 
adjustments of investment gains or losses in volatile markets, could 
bring the entire system to a screeching halt.
    Inevitable Account ``Leakage.'' Unlike the TSP, the 
Administration's plan would prohibit loans and emergency withdrawals, 
and would require individuals to purchase annuities on retirement. I 
find it inconceivable, however, that Congress--or an Administration--
would long be able to resist calls for emergency access to funds before 
a worker's retirement, and in lump sum amounts. Suppose, for example, 
that an individual has suffered a devastating personal financial loss, 
such as thousands experienced in last year's Florida hurricanes in the 
destruction of their homes. Would these persons be told that they may 
not access their PRA balances to mitigate such dire misfortunes? What 
about a catastrophic illness, leaving a family's breadwinner unable to 
work? Could such persons be denied their account balances to sustain 
spouse and children? I don't think so. There are, of course, scores 
more such examples, and with 155 million potential participants, you 
can be sure that they all would arise. Administering the inevitable 
emergency withdrawal or loan program would add enormously to the cost 
of the Administration's plan.
    Communication Problems. The TSP has a very effective communications 
system, because it can rely on the federal employing agencies to 
distribute plan materials and to educate and counsel their employees. 
Even so, the TSP found it necessary to have the central record keeper 
for its 3 million accounts maintain a staff of more than 200 telephone 
counselors to respond directly to questions from individual 
participants. Since more than 200 million Social Security taxpayers and 
retirees eventually would be eligible for PRAs, the required number of 
telephone counselors would be more than 13,000, based on the TSP 
experience, and probably much higher because of the special PRA 
deficiencies noted above. \7\ Also, TSP mailings consistently have 
reached more than 99 percent of participants, but 25 percent of SSA 
mailings are returned as undeliverable.
---------------------------------------------------------------------------
    \7\ Fidelity Investments, a major 401(k) provider, has estimated 
that the administration of a 401(k)-type plan for Social Security 
taxpayers would require a total staff of 100,000. See Employee Benefit 
Research Institute, Issue Brief No. 23, Nov. 1998, p. 166.
---------------------------------------------------------------------------
    Congress would undoubtedly insist that every effort be made to 
advise all Social Security taxpayers of the PRA benefits Congress 
intended to provide them. The TSP sent summary plan documents to all 3 
million eligible employees, which required 18 trailer trucks of printed 
materials. Similar documents would have to be sent eventually to the 
more than 200 million Social Security-covered employees and retirees.
    The eventual costs of such massive efforts at this point are 
unknown, but they clearly would have a significant impact on PRA 
expenses.
    Small Employer Antipathy. Even if small businesses were able to 
perform normal employer functions for PRAs, would they want to? PRAs 
would be voluntary for employees but, if employees elect to have PRAs, 
mandatory for their employers.
    The TSP and 401(k) plans generally are enthusiastically sponsored 
and supported by the large employers who offer them as a major benefit 
for their employees, and as a means to move away from defined benefit 
retirement plans that require employers to bear substantial investment 
risks. The major attractions of the TSP and 401(k)s generally are the 
matching employer contributions and the immediate tax benefit from 
excluding employee contributions from taxable income. The ability to 
borrow or withdraw funds to meet emergency needs is also a significant 
benefit. PRAs, as currently proposed, would offer none of these 
benefits, and would be a relatively unattractive product that employers 
might be reluctant to support, especially small employers who do not 
have any pension plans. Moreover, it would be unrealistic to expect 
small-business employers to act as large corporate employers do in 
assuming the costs of investment losses because of, say, employer error 
in transmitting funds for timely investment of 401(k) accounts, or for 
myriad other commonplace employer errors. These serious concerns for 
small businesses would have to be addressed during congressional 
hearings on PRA proposals. (See the examples of legal issues on page 5 
above.)

The Trust Fund Alternative
    Since PRAs are certainly not feasible for employees of small 
businesses--the vast preponderance of the business community--the only 
practical way to give them the higher returns available from equity 
investments is to invest part of the Social Security trust fund in 
equities. That way, the overwhelming administrative costs and practical 
problems of the Administration's plan would be avoided. The total 
administrative cost of having the Social Security trust fund invest in 
the private funds proposed for PRAs would be no more than one basis 
point, based on the actual costs of market investments by the Thrift 
Savings Plan. The likely increase in trust fund earnings would be an 
effective way to help maintain the solvency of the trust fund without 
having to resort to significant increases in Social Security taxes or 
reductions in benefits.
    Every state in the United States has authorized public retirement 
fund investment in stocks. Yet the federal government still clings to 
the old notion that governments should not have an ownership stake in 
private companies, which made some sense when individual stocks were 
involved. Today's broad-based index funds, however, remove the investor 
from direct control over particular companies. Small business employees 
should not be denied the benefits of portfolio diversification in the 
Social Security trust fund simply because the federal government has 
not kept up with the states in understanding the evolution of financial 
markets.
    Less Government Influence Over Private Companies. As shown in the 
following chart, there is even less government influence over private 
companies under the trust fund alternative than under the TSP or the 
Administration's plan.


               Government Influence Over Private Companies

                                                                Social
                                  Thrift     Administration    Security
                                 Savings          Plan        Trust Fund
                                   Plan                      Alternative

Selection of stock and bond    Government   Same             Same
 index funds                    decides
Selection of fund managers     Government   Same             Same
                                decides
Selection of private record    Government   Same             N/A
 keeper                         decides
Selection of auditors and      Government   Same             N/A
 consultants                    decides
Selection of annuity           Government   Same             N/A
 providers                      decides
Selection of allocations       Individuals  Individuals      Government
 among index funds              decide       decide           decides


    N/A--not applicable. (There would be no need for private record 
keepers, auditors, consultants, or annuity providers for trust fund 
investments.)
    Sppecial Benefits for Trust Fund.Unfortunately, some political 
leaders have convinced many of the public that the Social Security 
trust fund is not really invested because it has been ``looted,'' and 
that the trust fund consists of ``worthless IOUs.'' Nothing could be 
farther from the truth, and such statements betray an apparent 
ignorance of federal finance in our highest circles of government. The 
trust fund is fully invested in the best securities in the world--U.S. 
Treasury obligations. Private trust funds invest in Treasury securities 
in the open market, but the Social Security trust fund buys its 
Treasury securities directly from the Treasury, which is more efficient 
than if the Treasury were to issue the securities in the market and 
then buy them back for the trust fund.
    Moreover, the trust fund actually gets a much better deal than the 
private funds that buy Treasuries in the market. The trust fund, by 
law, may redeem its securities before maturity at par value, rather 
than at the sometimes deep market discounts suffered by private 
investors during periods of rising interest rates. Also, since the 
trust fund gets its securities directly from the Treasury, it avoids 
the market transaction costs which private investors must pay. Finally, 
the law requires the Treasury to pay the trust fund an interest rate on 
all of its investments in Treasuries equal to the average yield on 
long-term Treasury marketable securities. This is a significant benefit 
to the trust fund, since long-term rates are generally much higher than 
short-term rates. Thus in recent years, private investors have been 
earning about two percent on their short-term Treasuries, while the 
Social Security trust fund was earning about four percent on 
effectively the same maturities. The public seems to be totally unaware 
of these subsidies to the Social Security trust fund, which have been 
there for many decades.
    Trust Fund Dedicated to Social Security. The assets of the Social 
Security trust fund consist of investments in Treasury securities 
solely for future beneficiaries. Yet political leaders from both 
parties complain that the Treasury has ``spent'' the trust fund surplus 
on government programs. What on earth do they expect the Treasury to do 
with the money--bury it in the Treasury's back yard? The Treasury also 
spends the money it raises by issuing Treasury securities in the 
market. Does that mean that the private investors in Treasuries are 
also being ``looted'' by the Treasury? Of course not. The scandal would 
be if the Treasury left the trust fund uninvested and not earning 
interest. Then the Secretary of the Treasury would be in effect saying 
``I don't owe you,'' and that indeed would be a worthless IOU.
    So why do government officials find fault with perfectly sound 
financial practices? From ignorance, as I suggested earlier?--or is it 
is because they are trying to hide the real problem, which is the 
unique way the Social Security program is treated in the budget? Social 
Security expenditures are excluded from the budget and thus from the 
restraints on other government spending, which is proper since they are 
entitlements, and cannot be restrained under existing law. But the 
Social Security surplus is then, inconsistently, included in the 
calculation of the overall budget deficit, for the sole purpose of 
appearing to have achieved deficit, and thus spending, reduction. Then, 
having committed this accounting farce, officials have the audacity to 
complain that the misleading budget treatment of the trust fund 
surplus--which they could change--makes it available to finance other 
programs. The problem here is not the financing of the trust fund, but 
the political gimmickry of its budget treatment.

Conclusion
    In conclusion, the Administration's plan for universal PRAs is not 
feasible, and it should not survive the process of responsible 
Congressional hearings. The only practical way for the Social Security 
system to capture the higher returns available from investments in 
stocks is to diversify Social Security trust fund investments. The 
trust fund alternative, compared to PRAs, would involve less government 
influence over private companies, would be less disruptive of financial 
markets, would save tens of billions of dollars a year in 
administrative costs, and could be effective virtually immediately, 
rather than the 2009 starting date proposed for PRAs. The multi-
trillion dollar transition costs proposed by PRA proponents would be 
avoided. The additional trust fund earnings would go a long way toward 
strengthening Social Security finances, and would thus reduce, if not 
eliminate, the need for significant tax increases or benefit 
reductions.
    Thank you for your attention. I would be pleased to answer any 
questions.


                                 

    Chairman MCCRERY. Thank you, Mr. Cavanaugh. Ms. Reno.

 STATEMENT OF VIRGINIA P. RENO, VICE PRESIDENT FOR INCOME AND 
     SECURITY POLICY, NATIONAL ACADEMY OF SOCIAL INSURANCE

    Ms. RENO. Thank you, Mr. Chairman, and thank you for the 
opportunity to be here today. I will present findings from a 
new study panel report from the NASI that is solely about how 
money would be paid from individual accounts. I think you all 
have a copy of the report available to you. This was the work 
of 27 top experts in various fields over a two-and-a-half year 
period led by bipartisan cochairs, Michael Graetz of Yale Law 
School, who was a top Treasury official in the George Herbert 
Walker Bush Administration, and Ken Apfel, who is now at the 
LBJ School in Texas, and was Commissioner of Social Security 
during the Clinton Administration. The ten chapters in this 
report cover a host of payout issues about retirement, about 
annuity markets, early access to the money, disability, spousal 
rights, children's benefits, offsets, tax treatment and 
financial demographics. I will focus today simply on the 
retirement payout issues.
    At retirement we face four kinds of risks. We don't know 
how long we will live, how long our spouses will live, how our 
investments will do, or how prices will rise. Social Security 
covers these risks by automatically adjusting benefits for 
inflation and paying spousal benefits. Individual accounts, in 
and of themselves, do not provide for those risks, but 
retirees, as Dr. Mitchell said, can buy annuities that turn 
their savings into a lifetime income. That is why annuities 
that pay out over a lifetime are so important. For a retiree, 
the good news about a life annuity is that your money will last 
as long as you live. The bad news about an annuity is you have 
to pay the full price up front. That is how they work. You 
can't change your mind. So, all of the money used to buy an 
annuity is not available to leave to heirs.
    In a sense, an annuity is a chance to trade ownership for 
income security once you get to retirement. It is important to 
recognize also that more security in retirement will cost more. 
If a retiree wants inflation protection, his annuity will start 
out at a lower level. If he wants to make sure his wife has a 
survivor benefit, his annuity will start out again at a 
somewhat lower level. Essentially, when choosing these 
features, the retiree gets to choose current income over future 
security. An important question our panel looked at is who 
would provide inflation-indexed annuities since all seem to 
agree that inflation indexing is very important. Our panel 
concluded that it would take some help from the Federal 
Government to have inflation-indexed annuities on a widespread 
basis. The government might issue a large volume of TIPS that 
my colleague mentioned, or it might simply issue the annuities 
directly to retirees. In either case, the government would be 
receiving a large amount of money for these transactions, 
either from TIPS investors for private annuities, or from 
annuity purchasers themselves. An important question that 
hasn't received much attention is how would the government 
manage, invest, or spend that volume of money it receives to 
facilitate inflation-indexed annuities?
    Another key question is, will the government insure the 
insurers if annuities are provided privately? States now 
regulate insurance companies that provide life annuities. If 
Federal law required people to buy annuities, there might be an 
expectation that the Federal Government would also guarantee 
the solvency of those insurance companies. If so, it is 
important to think through what that guarantee would look like. 
Would it resemble the Federal Deposit Insurance Corporation or 
the Pension Benefit Guaranty Corporation? Essentially, that is 
a question that hasn't been fully resolved.
    To wrap up, life annuities offer retirees the choice to 
trade ownership for retirement security or to trade current 
income for future security. The decisions have lifelong 
consequences for the retiree and his or her family. Retirees 
may need help to make informed choices. Finding and paying for 
good consumer education will be important. In terms of consumer 
education, there remains confusion today about exactly what we 
call annuities. Two different products are called annuities, in 
fact, and they are very different. Deferred annuities are 
investment products. They don't guarantee income for life. Life 
annuities are quite different. The deferred annuities are, in 
fact, more actively marketed and are a much bigger market than 
are life annuities. We are hearing stories today about how 
elders are being misled about deferred annuities and end up 
losing some of their lifetime savings.
    To conclude, our cochairs have both emphasized that these 
payout issues with individual accounts are first-order 
questions, and it is really important that the Congress pay 
attention to them. We are glad that you are doing that today. 
Thank you.
    [The prepared statement of Ms. Reno follows:]

 Statement of Virginia P. Reno, Vice President for Income and Security 
              Policy, National Academy of Social Insurance

    Mr. Chairman and members of the Subcommittee, thank you for this 
opportunity to testify before you today. The National Academy of Social 
Insurance is a non-profit, non-partisan organization of the nation's 
leading experts on Social Security, Medicare, and related programs. Our 
mission is to promote sound policymaking on social insurance through 
research, education and the open exchange of ideas.
    I will present findings of our new study panel report, Uncharted 
Waters: Paying Benefits from Individual Accounts in Federal Retirement 
Policy. Twenty-seven top experts contributed to this study. It was led 
by bi-partisan chairs, Mike Graetz of Yale law school and a top 
Treasury official in the George H. W. Bush Administration, and Kenneth 
Apfel of the LBJ School at the University of Texas and Commissioner of 
Social Security in the Clinton administration. The report is solely 
about how money would be paid out of individual accounts. Why are 
payouts important? Our co-chairs Mike Graetz and Ken Apfel said it 
best:
    ``Payouts are important because the central goal of Social Security 
is to assure some level of adequate income.'' ``To date, payouts have 
been largely neglected. Yet it is crucial that policymakers resolve 
issues in this report if they decide to add universally available 
individual accounts to our current system of providing retirement 
income.''
    Our expert panel did not seek to agree on whether individual 
accounts in Social Security are a good idea. Nor did they agree on a 
blueprint for how to design payouts. Rather, they did agree that issues 
in this report are the right ones and how the questions get answered is 
critically important. The ten chapters cover such questions as:
    (a) How would retirees get the money? (b) If annuities are 
required, who would provide them, and how? (c) Could workers get the 
money before they retire? (d) What rights would spouses have? (e) How 
would accounts affect disability benefits? (f) How would children be 
affected? (g) How might ``offsets'' affect payouts? (h) How would 
payouts be taxed? I will focus on retirement payouts. First, I 
summarize what we called ``financial demographics'' that set the stage 
for considering accounts.

Financial Demographics
    Social Security is the bedrock of income security for millions of 
Americans. The 47 million beneficiaries account for about one in six 
Americans living in one in four U.S. households. About two in three 
beneficiaries age 65 and older rely on Social Security for half or more 
of their total income. Women without husbands are the most reliant on 
Social Security; three in four such women over age 65 get half or more 
of their income from Social Security. For 44 percent of these women, 
Social Security is nearly all they have, making up 90 percent or more 
of their income.
    Despite beneficiaries' reliance on Social Security, the benefits 
alone do not provide a comfortable level of living. The average benefit 
for a retired worker was about $955 a month, or $11,500 a year in 
January 2005. Average benefits are somewhat lower for disabled workers 
($894) and elderly widows ($920). Benefits for future retirees will 
grow somewhat more slowly than earnings, which will cause replacement 
rates to decline over the next 20 years as the age for full retirement 
benefits rises from 65 to 67. Benefits for 65-year-old retirees will 
replace a smaller share of prior earnings than is the case today or at 
any time in the last 30 years. Because Social Security is not in long-
run financial balance, other changes might be enacted that will further 
lower benefits or raise revenue.
    Employer-sponsored pension plans have covered about half of 
private-sector workers over the past 25 years. These plans are shifting 
away from the defined benefits that dominated the 1970s and 1980s to 
defined-contribution or 401(k)-type plans. The newer plans give workers 
more choices about whether to participate and how much to contribute; 
workers can take the accounts with them when they change jobs; and they 
have choices about when and how to withdraw the money. At the same 
time, workers finance more of the plans themselves and bear the 
investment risk that employers took on in defined-benefit plans.
    In 2001, about half of all U.S. families owned a tax-favored 
retirement account. The median balance of those accounts was $29,000. 
Older households had somewhat larger tax-favored savings, with a median 
value of $55,000 for the 59 percent of families age 55-64 who had such 
accounts. Tax-favored savings are concentrated among high-income 
households; families in the top 20 percent of the income distribution 
held two-thirds of all tax-favored retirement savings.
    The heavy reliance on Social Security among retirees up through the 
middle of the income distribution, the shift away from defined-benefit 
pensions, and increased use of 401(k) plans amplifies the importance of 
payout options that convert savings into guaranteed incomes during 
retirement.

Financial Risks for Retirees
    Retirees face four kinds of risks to their financial security. They 
don't know how long they will live, how long their spouses will live, 
how their investments will perform, or how much prices will rise in the 
future. Social Security covers these risks by paying benefits for life, 
with automatic cost of living increases, and automatic survivor 
benefits. Individual accounts, in and of themselves, do not cover these 
risks. But retirees can buy life annuities that turn their savings into 
guaranteed income for life. That is why life annuities are important.

The Life Annuity Trade-Off
    From the retirees' perspective, the good news in buying a life 
annuity is that your income will last as long as you live. The bad news 
is that you pay the full purchase price up front, and the purchase is 
irrevocable. All the money used to buy the annuity is no longer 
available to leave to heirs. For example, if you use $40,000 to buy a 
life annuity and die a few months later, that $40,000 is gone. The 
insurer uses the money from people who die early to cover the costs of 
paying annuities to those who live a long time.

More Protection Costs More
    If a retiree wants his annuity to keep up with the cost of living, 
it will start lower. If he wants it to continue to pay his widow after 
he dies, his payment will start out lower. For example, in one set of 
assumptions, a 65-year-old retiree with $10,000 could buy a simple 
annuity of $80 a month. If it is to keep up with the cost of living (at 
say 3 percent a year), it would start out lower, $62 a month. If it 
would continue to pay as long either the retiree or his wife lived, it 
would start out at $50 a month. (These prices assume everyone would be 
required to buy annuities. If annuities were optional, they would pay 
less than shown here because of what is known in the insurance world as 
adverse selection.)

Guarantees Might Provide for Heirs
    Some annuities guarantee payments to a death beneficiary if the 
retiree dies shortly after buying an annuity. A 10-year certain annuity 
assures payment for 10 years if the annuitant dies in less than 10 
years. This feature will lower the initial monthly payment to the 
retiree (from $62 to $58, in the above example, if the annuity is 
inflation-indexed).

Changes in Marital Status
    In general, life annuities cannot be rewritten to shift from a 
single-life to a joint-life annuity if one marries after retirement. 
Nor can one ``undo'' the purchase of a joint-life annuity and shift to 
a single life annuity if a marriage ends shortly after buying an 
annuity. This could affect married couples' decisions about whether and 
when to buy annuities. It could also produce very different incomes for 
widows depending on whether they were widowed before or just after 
buying annuities. For example, if John dies before buying an annuity, 
Mary could inherit his account, combine it with her own, and buy a 
single life annuity with the total amount. Consider this the base case. 
If instead, John and Mary both bought joint-life annuities, each would 
start out with a monthly payment that is about 81 percent of what a 
single-life annuity would provide. When one died, the survivor would 
receive 81 percent as much as the base case. The key point is that the 
timing of annuity purchase interacts with the timing of widowhood to 
produce very different results for retirees who are otherwise in 
similar circumstances.

Who Would Provide Inflation-Indexed Annuities?
    Inflation-indexed annuities are very rare in the private insurance 
market. Our panel concluded that some help from the federal government 
would be needed to develop such a market. If the federal government 
increased the supply of long-duration (say 30-year) Treasury Inflation-
Protected Securities (TIPS), insurers could use them to hedge inflation 
risk. Alternatively, the government could sell inflation-indexed 
annuities directly to retirees.
    In either case, the government could be receiving a large amount of 
funds (from TIPS buyers or from retirees' annuity premiums) that would 
represent contractual obligations to make long-term inflation-indexed 
payments. For example, in a universal system, funds backing annuities 
funded with 2 percent of workers earnings could amount to about 15 
percent of GDP when the system is fully mature. A key question for 
policymakers is, ``how would the government manage (or spend or invest) 
the large volume of funds it received from TIPS buyers or annuity 
purchasers?''

Insuring Insurers
    Currently, state governments have sole responsibility for 
regulating insurance companies and guaranteeing their solvency. If 
federal law requires or encourages retirees to buy annuities, it might 
also be expected to guarantee the solvency of the insurance companies 
that have made long-term commitments to retirees. How that solvency 
guarantee would be organized is an important question. Would it 
resemble the Pension Benefit Guaranty Corporation, or the Federal 
Deposit Insurance Corporation, or other federal models?

Recap of Annuity Choices
    The design of retirement payout rules will confront inevitable 
tension between offering choice and providing security. Hard and fast 
rules might ensure some level of security, but will also create 
pressure for exceptions. As accountholders approach retirement, they 
will face choices such as:

(a)

Whether to buy a life annuity at all;

(b)

How much of one's account to spend on a life annuity;

(c)

Whether the annuity would be indexed for inflation;

(d)

When to buy a life annuity;

(e)

Whether to buy a guarantee feature and, if so, what kind;

(g)

If joint-life annuities are optional for unmarried retirees, whether to buy 
one and with whom; and

(h)

If joint-life annuities are offered or required for married retirees, which 
type to buy and what level of benefit to provide the secondary annuitant.

    Decisions on these questions will have lifelong consequences. To 
the extent that retirees have choices, it will be important that they 
receive advice and assistance to understand the consequences of 
different courses of action for themselves and for the well-being of 
their spouses, dependents, and potential heirs. Organizing and paying 
for trustworthy advice could become an important issue in a new system 
that envisions many choices in the purchase of life annuities.
    Finally, in the realm of consumer education, there is a risk of 
confusion about what we mean by ``annuity.'' Two different products are 
called ``annuities'' and they are very different. Deferred annuities 
are investment products, not insurance. Only life annuities guarantee 
payments for life. Deferred annuities are far more common and more 
actively marketed by financial advisors, brokers or agents. In recent 
weeks, a few news reports have pointed to problems in the marketing of 
deferred annuities to seniors, who end up losing part of their life 
savings because they bought products that were not appropriate to their 
circumstances. (Moregenson 2005, Kirchheimer 2005).
    Sources:
    Reno, Virginia P., Michael J. Graetz, Kenneth S. Apfel, Joni 
Lavery, and Catherine Hill (eds.), (2005). Uncharted Waters: Paying 
Benefit from Individual Accounts in Federal Retirement Policy, Study 
Panel Final Report, Washington, DC: National Academy of Social 
Insurance.
    Morgenson, Gretchen, ``Who's Preying on Your Grandparents?'' New 
York Times--Business Section, Sunday, May 15, 2005
    Kirchheimer, Sid, ``Deviled Nest Eggs,'' Consumer Alert, AARP 
Bulletin, June 2005.

                                 

    Chairman MCCRERY. Thank you, Ms. Reno. We have, we think, 
one vote on the floor. We are going to run over to the floor, 
vote, and come right back. The Committee will be in recess 
until we return.
    [Recess].
    Chairman MCCRERY. The hearing will come to order. Thank you 
all for your patience. We are glad to be back to ask you a few 
questions about your testimony, which was excellent. We 
appreciate very much your being with us today to share those 
views with us. Mr. Pollock, I appreciate your thoughtful 
approach to establishing personal accounts using the Social 
Security surplus. In your proposal, you recommend having 
workers initially invest in inflation-indexed Treasury bonds. 
Why do you recommend this particular type of bond? How large is 
the market now for that type of bond? Would having individuals 
invest solely in those specific forms of Treasury securities 
affect the market in any way?
    Mr. POLLOCK. Mr. Chairman, I view the biggest risk to 
retirement savings to be inflation. Especially if you own 
straight bonds, if we should get into a situation of monetary 
expansion which results in inflation, it is a way to lose the 
real purchasing power of your savings. With the inflation-
protected securities, you are automatically covered by the 
inflation indexation of the bonds against this single biggest 
risk. For decades, inflation-indexed securities have been 
promoted as the right idea by economists. They have been 
introduced by many countries including this country, during the 
nineties. They have been extremely popular. There are many 
mutual funds that offer them to the public, and they have been 
welcomed by public investors. I picture the TIPS in personal 
accounts being a private placement, another similarity to what 
happens now with the trust fund. The bonds now are privately 
placed with the Social Security trustees. This would be a 
private placement into a personal account, so, there would be 
no need to sell the bonds, nobody would have to beg foreign 
investors to buy them; in the first instance, you wouldn't move 
the market.
    I also recommend in my detailed comments that there be some 
restricted period, I suggest perhaps 5 years, where these 
couldn't be sold, just in case--I don't think it would happen, 
but just in case--there would be an outflow from the accounts 
that would be market moving. If you had a 5-year period, 
another advantage is that then these accounts would 
automatically buildup to a size where they would become 
economically easier to evolve into other investment 
opportunities if we wanted to at that point.
    Chairman MCCRERY. Now, talk to us about the advantage--I 
assume you think there is an advantage--in issuing these 
Treasury securities in the names of individuals, as opposed to 
simply issuing Treasuries to the Social Security Trust Fund. 
Expound upon that.
    Mr. POLLOCK. I think the single most important advantage, 
Mr. Chairman, is that it is a philosophical advantage, as I 
said in my testimony. It is consistent with the deep and 
extremely important trends in American political philosophy of 
trying to create widespread ownership throughout American 
society, where the ordinary citizen is an owner of property as 
part of the American social miracle. This plan would promote 
widespread ownership of retirement savings, as I said before, 
just as we promote widespread home ownership. So, I think there 
is a huge advantage in the fundamental philosophical message we 
are giving.
    Now, from the individual's point of view, you are, in 
addition, gaining an absolute obligation of the government, a 
true right, in exchange for the very large amount of money you 
are giving over to the Social Security system, as opposed to a 
future political formula which may or may not be there. As we 
know from public surveys, a very large number of people, 
especially young people, simply don't believe they will ever 
get money from Social Security. We can say that the citizen 
knows that if all you do is raise taxes to pay Social Security 
benefits, net, you haven't done anything but taken money and 
given it back. So, this plan would establish a true obligation; 
you would own a U.S. Treasury bond. I go on in my detailed 
testimony--and I can comment if you want me to, Mr. Chairman, 
as to how that might further link to actually decreasing the 
long-run Social Security deficit, which I think it could do in 
an efficient way.
    Chairman MCCRERY. So, I believe you are saying, Mr. 
Pollock, that by taking an equivalent amount in Treasury 
bonds--an amount in Treasury bills equivalent to the amount of 
the cash surplus in Social Security, and putting those in the 
names of individuals, you are at least protecting, for those 
individuals, that surplus to be paid as part of Social 
Security?
    Mr. POLLOCK. Yes, sir. You are protecting it by making it 
into a real Treasury obligation, a real asset that they 
actually own.
    Chairman MCCRERY. Thank you. Dr. Mitchell, you and other 
members of the President's Commission to Strengthen Social 
Security concluded that it would be reasonable to establish 
personal accounts along the lines of the Federal Thrift Savings 
Plan. Why did you use the TSP as a model?
    Dr. MITCHELL. We used the TSP as a model for several 
reasons. One is, it is very clear, efficient and low cost; as 
was testified to by Mr. Purcell earlier today, the fact that it 
has large scale means that it is quite inexpensive to offer. 
Also, the fact that it has relatively few investment choices, a 
reasonable array--I would personally add, probably, a lifecycle 
fund, maybe a TIPS fund--but a reasonably small number of 
funds, makes it inexpensive. Also, the fact that the plan is 
fairly well communicated, fairly well understood, it has been 
around a long time, and it seems to have some history behind it 
offers an example of how it might work.
    Chairman MCCRERY. Thank you. Mr. Cavanaugh, in your 
testimony, you talk about the administrative expenses being, 
perhaps, very high if we were to try to go to personal accounts 
in Social Security; and you back up your assertion by pointing 
to administrative expenses of small 401(k) plans, small defined 
contribution plans, and, particularly, small employer plans. 
While I think what you say is true in terms of the cost of 
those plans, I am not sure how that is relevant to establishing 
personal accounts in Social Security. After all, the Social 
Security system itself is kind of like a big 401(k) plan. There 
are personal contributions collected by the employer and sent 
to the central government. What most personal account proposals 
propose is for basically the same thing to happen, except 
instead of the money--all the money going into a central pot, 
it goes into individual accounts, which as we heard in 
testimony from the SSA this morning--the SSA seems to indicate 
that we have the capacity to do that at a much lower cost than 
you have estimated. Do you want to comment on that, and maybe 
clear that up?
    Mr. CAVANAUGH. Yes. What you are saying, I think, was what 
many people in the 401(k) industry believed a decade or so ago, 
because they had already got all the business they could get 
from large corporations where everybody had a 401(k). So, they 
started to try working with smaller businesses and bringing 
them in, to provide them the services, and they had the notion 
that they could get economies of scale. I think they did a 
great job. They bring in businesses, some not too small, and 
some of the smaller ones, and they get thousands of them into 
their computerized system, and that way they get the economies 
of scale, just like the TSP.
    But, what they find is that in order to do this, they have 
to deal with each one of these businesses and their employees. 
You have to go in there and explain the plan document. You have 
all sorts of explanations of different investment options; and 
you have a fiduciary responsibility to make sure that you don't 
give investment advice, but at the same time you give 
information. It is a very tricky business. It requires 
professional people to do it, and that is why they can't do it 
for less than 3,000 bucks a year, even though they have these 
tremendous economies of scale that they have achieved over the 
last 10 years and fierce competition.
    The difference between that and the TSP is, the TSP is just 
one employer, the government, not five-and-a-half million. The 
first thing I did when I started the TSP was to announce that, 
Hey, I am a wholesaler, I am not going to do the retail. I left 
it to the Federal agencies and their field offices all over the 
world to do the face-to-face; the consultation, the advice, and 
all that sort of thing, passing on the forms, helping people 
fill them out.
    Congress required that Office of Personnel Management train 
trainers in every Federal agency for this purpose, a very 
demanding task. So, they did the retail for me. They had 
systems in place, they had the computers, they had the payroll 
offices. Small businesses have none of that, and that is 
fundamentally why you just can't reach them. Sixty percent of 
businesses in the United States have less than five employees, 
which means the cost, or expense ratio, would be twice what I 
testified on. You can't get there from here.
    Chairman MCCRERY. Couldn't we address some of those 
concerns, though, in a different way from the 401(k) model? 
Since this is going to be basically a government-administered 
plan, as was suggested by Mr. Lockhart this morning, the SSA 
itself could play a role in educating workers as to the choices 
in their Social Security account. Couldn't we relieve that 
burden from the small employers?
    Mr. CAVANAUGH. Yes. I think politically, as well as a 
practical matter, virtual guarantees have been given to small 
business: You are not going to have to do anything more than 
what you are doing now. There are two problems there. What they 
are doing now, while adequate for the purposes of the Internal 
Revenue Service (IRS) and the SSA, is totally inadequate for an 
investment function when the SSA cannot reconcile $10 billion a 
year to individual accounts. You can accept that in Social 
Security because of the way they credit the benefits; you 
cannot accept that in a financial institution. I balance the 
TSP to the penny every day, and that is the way banks have to 
operate. So, the information system is just totally inadequate.
    When you talk about alternatives--of course, in my 
testimony, I was focusing, as I said, on the Administration's 
plan, which is the 401(k)-type structure, there is little 
difference between the two. I think the reason why you find 
very few proponents of alternatives like Individual Retirement 
Accounts (IRAs), Great Britain and Chile stumbled doing that 
sort of thing, is because you have millions of people out there 
that are financially unsophisticated, at the mercy of a whole 
bunch of financial institutions. Certainly in the United 
Kingdom they learned never to do that again.
    Chairman MCCRERY. Right.
    Mr. CAVANAUGH. So, then you get into Simple IRAs, which 
some small businesses use, but they require employer 
contributions, which a little business just can't handle. So, 
what you get down to is--even though you say to small 
businesses, you don't have to do any more than what you are 
doing--what they are now doing is inadequate for investment 
purposes. Second, if they are not going to do the normal 
employer functions in a 401(k), who is going to do them? I can 
take you through the traces on it. You wind up with a 401(k) 
provider having to do them, people like Fidelity Investments, 
and T. Rowe Price, and Vanguard, and all those people; they 
will go in and they will do them for you as they have been 
trying to do. They will come out with the same economics. You 
still have to pay 3,000 bucks a year--more, actually, than 
that, because right now the 401(k) industry has pushed the 
frontier, getting into more and more businesses, and some 
smaller ones. Beyond that it is a jungle which they can't 
penetrate; little tiny businesses, you just can't deal with 
them. So, you would be back in the hands of the 401(k) 
providers, God bless them, but you can't ask them to do it 
without a government subsidy.
    Chairman MCCRERY. Well, it seems to me it is kind of apples 
and oranges when we are talking about a central government 
system and lots of 401(k)s. I understand. I get your--one of 
your points that the educational process would be difficult, 
but I don't know that it would be so difficult that we couldn't 
do it if we managed it correctly. Do any of the other panelists 
have a thought on this subject? Mr. Cavanaugh has brought up 
something that is provocative. Mr. Pollock.
    Mr. POLLOCK. Mr. Chairman, I think one of the advantages of 
the sort of plan that I was proposing, which is the creation of 
book entry government securities, is that it solves a lot of 
the administrative problems. It makes a purely book entry, 
computerized system, possible, which would have very low costs, 
even granting the existence of problems at Social Security, 
which Frank brings up. A second reason is that, as an education 
project, it is exceptionally simple. It is essentially the same 
as acquiring savings bonds. I think if you explain to the 
American people that this is in exchange for part of your 
Social Security; and I stress the ``part,'' because we are 
talking about the surplus. We are not talking about disability 
insurance or the other welfare aspects of Social Security; we 
are talking about the savings aspects of Social Security. For 
that we are creating an actual savings program, which looks 
just like a payroll deduction, to acquire a savings bond. In 
particular, it looks like an I Bond, which is an inflation-
indexed savings bond. People will understand that, and I think 
they will like it very much.
    Chairman MCCRERY. Mr. Purcell.
    Mr. PURCELL. I think I agree with both Mr. Pollock and 
Frank Cavanaugh. You certainly could have a low-cost system of 
individual accounts, one that is essentially invisible to the 
employer, they would do nothing more than they do now. However, 
that system would not look like the TSP. The fundamental 
roadblock is that, although, as was testified earlier today by 
Mr. Lockhart, more than 60 percent of W-2s are submitted 
electronically, more than 70 percent of employers submit W-2s 
on paper. Four out of 5 employers in the United States have 
fewer than 10 employees. They could eventually do all of that 
electronically; maybe in 10 or 20 years, and then it will be 
very easy, relatively easier, to design a system that somewhat 
resembles the Thrift Plan.
    Today, you have employers submitting large blocks of tax 
money to the local Federal Reserve Bank, which consists of both 
income tax withholding and Social Security taxes. The Federal 
Government doesn't know how much of that money is payroll tax 
and how much is Social Security tax, much less how much is mine 
versus a co-worker, until the employer files the W-2 once a 
year. That is what slows things down. You can't allocate those 
taxes into specific investments until you know how the employee 
wants to invest it; and if that is done on paper, it is going 
to be done very slowly.
    So, you could have a system that is built entirely on the 
current payroll tax structure, but you are going to have a 
delay; and the current delay, according to the SSA, could be 7 
to 22 months in getting a worker's payroll tax into the account 
that they choose to put it in. They will probably have the 
opportunity to redirect that investment from the bond fund to 
the stock fund once a year. They wouldn't be able to log onto a 
Web site and monitor their account daily and redirect their 
investments; that is what you can do with the TSP. So, it is 
simply -it is not a matter of, Can it be done? Certainly it can 
be done. There will be tradeoffs between choice and control by 
the participant, versus cost borne either by the employer or 
through a government subsidy.
    Chairman MCCRERY. Bear in mind that carving out part of the 
payroll tax is not the only way to fund personal accounts. So, 
don't get caught in that trap that everything has to be in that 
universe. It doesn't. There could be default options until the 
employee's paperwork was received by the Federal Government. 
So, I think, again, we can get around those----
    Mr. PURCELL. I agree completely. It is simply a matter of 
the public understanding that this is what it will look like, 
and is that okay with you. If it is, it will work.
    Chairman MCCRERY. But, mechanically, and from a cost 
standpoint, I still believe we can solve those problems without 
the kinds of administrative expenses Mr. Cavanaugh is talking 
about. Ms. Entmacher.
    Ms. ENTMACHER. I did want to point out that if the accounts 
are supposed to be divisible at divorce, that is, the 
contributions that are made during the course of the marriage 
are supposed to be divided at the time of the divorce, there 
would be additional information that would need to be collected 
throughout about marital status. The timings of contributions 
would have to be related to the information about marital 
status. The contributions during marriage would have to be 
segregated from the amounts that were in the account prior to 
the marriage, which would be separate property. Then, you sort 
of start looking at, okay, what goes in during the marriage and 
what are the accumulations on that; and if that is the divorce 
division that is planned for, then you would have to know that 
from the front end, and plan for it.
    As I say, there is no system that currently does that. The 
TSP doesn't do it. We don't have a national registry. There is 
no national center that collects this information, even on a 
statistical basis, much less with respect to individuals. There 
would, quite frankly, be concerns if the Federal Government 
said, Okay, we are going to start collecting information about 
the marriages and divorces of all Americans--confidentiality 
issues. So, I think that--this is why so many of these payout 
issues have to be confronted at the front end, because you 
might say, Well, that is what we are going to do and then 
discover you don't really have enough information to implement 
a fair division of accounts.
    Chairman MCCRERY. Why couldn't the personal retirement 
account be an asset like any other asset in a divorce, and have 
the court decide how it is divided and communicate that to the 
SSA?
    Ms. ENTMACHER. Well, that is an extremely expensive way of 
doing it for the individuals. What happens today with many of 
those assets is that people who are not represented by lawyers 
don't get a fair share. I think what most people don't realize 
is that in a majority of divorce cases, at least one person is 
not represented by a lawyer. So, it is going to be very 
haphazard whether anyone would get a share of an account. It is 
particularly troubling because we don't know how much benefits 
might be reduced because a spouse contributed to an account. 
So, leaving it to a court to decide--plus you would get 
different results all over the country. Some States have 
community property, some States don't.
    Chairman MCCRERY. You get that now, though, with assets.
    Ms. ENTMACHER. Yes, and it results in a lot of women not 
getting a fair share of assets at the point of divorce. If we 
are talking, again, not about accounts that supplement Social 
Security benefits, but that are designed to replace them, we 
are talking about the core benefit, the benefit that people 
rely on to stay out of poverty--not a supplement. We are 
talking about the bottom half of the income distribution. The 
401(k)s and IRAs held mostly by higher income people. This is 
the basic tier of retirement income, so, it is especially 
important to make sure that women get a fair share.
    Chairman MCCRERY. I don't disagree with you, but, couldn't 
we--in order to cut through some of that disparity from State 
to State, jurisdiction to jurisdiction, overlay a Federal 
requirement with the disposition of these accounts so that the 
court would have to take into account the Federal requirement, 
and make that part of the judgment?
    Ms. ENTMACHER. Well, but the issue has to come to the 
attention of the State court, and figuring out--what does that 
mean? Do you--if you are not a community property State, do you 
use community property principles to divide the account? Where 
do you get that? You then have to turn to the SSA, or the TSP, 
or whoever is administering the accounts for information and 
say, ``Okay, tell us how much money went in during the course 
of the marriage and what were the accumulations?'' The 
information still has to be generated by somebody for the court 
to do that.
    Chairman MCCRERY. Well, again, I agree with you that there 
are complications, and believe me, we have looked at these. I 
think there are ways to work all those out to provide the 
maximum amount of protection for both spouses. You make some 
good points, and we need to pay attention to that.
    Ms. ENTMACHER. Thank you.
    Chairman MCCRERY. Mr. Levin.
    Mr. LEVIN. I will get back to the TSP. Let me just ask you, 
Mr. Pollock, right now we have a Social Security system that is 
progressive, right, replacement rate differs according to 
income? Under your approach, that would not be true?
    Mr. POLLOCK. Under my approach, the progressivity would 
remain the same.
    Mr. LEVIN. How is that?
    Mr. POLLOCK. It remains the same because, as is explained 
in my detailed submission, Congressman, in exchange for getting 
TIPS that you actually own in your own account today, you give 
up an equivalent value of future benefits; that is, 
economically you must do that, that is your choice. The future 
benefits, what I call the ``exchange rate'' between the future 
benefits and the TIPS today, have built into them, and into the 
exchange rate, all of the formulas which govern Social Security 
benefits. So, the result of all of that progressivity stays 
exactly the same.
    If I may comment further, Congressman--suggest that if we 
want to use this promotion of ownership and creation of real 
assets in exchange for Social Security payments, we want to use 
that, in addition, to reduce the deficit, we could make that 
exchange rate itself more progressive; that is to say, for 
higher income people you would give up more than a dollar's 
worth of present value of future benefits for a dollar today. 
It is my belief that since higher-income people doubt very much 
that they will ever receive the scheduled benefits, they would 
choose to go into the program.
    The result--excuse me, if I could just finish the thought. 
The result would then be that in addition to creating ownership 
and true assets, we would also be reducing the Social Security 
actuarial deficit. The liabilities of the Social Security 
program would be falling faster than the odd assets in the 
trust fund, and its finances would be improving.
    Mr. LEVIN. Indexing into it, you are suggesting--I don't 
know what you mean by ``higher income.'' Let's suppose that 
everybody with 20,000 and above was classified as higher. 
Disability would be separated out?
    Mr. POLLOCK. Congressman, in all of my mathematical work on 
this, I start by excluding the disability insurance program, 
dealing only with OASI, because disability insurance, by 
definition, is insurance--you must commingle funds to have 
insurance work.
    Mr. LEVIN. Survivor benefits?
    Mr. POLLOCK. Is the same. I think there is a little 
confusion, if I may say so, when we talk about creating 
personal accounts or personal lock boxes, because people tend 
to characterize all of Social Security as one thing, whereas, 
in fact, Social Security is two quite different things.
    Mr. LEVIN. Okay. I just have limited time.
    Mr. POLLOCK. I am sorry. You agree with that point, though, 
I think, Congressman, that it is part insurance or a welfare 
program and it is part a savings program.
    Mr. LEVIN. I wouldn't call it a welfare program. I think 
the public--you size up the public very incorrectly, I think 
the public has spoken pretty loudly about personal accounts. It 
is interesting, we are spending a lot of time on the TSP, 
though what I think has been suggested in the last 24 hours 
doesn't involve TSP. So--and maybe they listened, Mr. 
Cavanaugh, to your admonitions about the complexity of it. 
There is plenty of complexity with this concept. I guess they 
are including two bonds and somehow an increase in solvency; we 
will see how they work out without taking general revenue 
moneys or making promises as to the future. We spent a lot of 
time on the TSP. It is in the President's plan; it is not, 
apparently, in this concept. Ms. Entmacher, just if you would, 
the greatest dangers you think of what Mr. Pollock has 
suggested are what?
    Ms. ENTMACHER. I think the greatest danger is really 
embodied in this idea that it is an individual lock box. First 
he talked about, well, that doesn't apply to disability, and I 
wonder whether he is including the family disability benefits 
in that as well. Particularly women, when they reach retirement 
age and take benefits at 62 and above, many, many women rely on 
the higher benefit that they get as a spouse and as a widow. 
Millions of working women, women who work in the paid labor 
force, who have earned their own Social Security benefit, get a 
supplement that Social Security provides to spouses and widows.
    Despite the fact that more women are in the paid labor 
force for longer periods of time, it is still the fact that 
they earn less than men and take more time out of the labor 
force. So, when they reach retirement, their income security 
really depends on the higher benefits that they get as spouses. 
What has been referred to as the individual savings component 
of Social Security--they benefit from the fact that it is not 
an individual savings component; it is a family savings 
component that protects workers and their spouses and surviving 
spouses. So, I think that is the biggest concern. We could, in 
a few years, see a return of the poverty rates, particularly 
among elderly women who already are the large majority of the 
elderly poor, to what they were back before Social Security was 
created.
    That is my nightmare. When I look at the benefits that 
women will get under Social Security in the future, 40 percent 
of women are going to be receiving spousal benefits 40 years 
from now; they are still going to be relying on them. When I 
look at what women have in their 401(k)s and IRAs, it is 60 
percent of what men have. They are still disadvantaged. So, if 
it is every man for himself in this brave new world, I worry. 
That is my biggest worry.
    Mr. LEVIN. My time is up. Let me just ask Dr. Mitchell 
about the Hirschel Organization, and they would have to buy--
what would keep them out of poverty? We are talking about 20, 
30 years, right, for the average woman?
    Dr. MITCHELL. Right.
    Mr. LEVIN. Twenty or 30 years; we are talking about a long 
period of time. How do you calculate an annuity so that you are 
sure it will keep people out of poverty when you are not sure 
what the poverty level is 5, 10, 15, 20 years from now; how do 
you do that?
    Dr. MITCHELL. Well, practically speaking, the poverty line 
in the United States is defined according to a real standard, 
that is, an inflation-indexed standard. So, what you would do, 
practically speaking, is compute the annuity that you would 
take from your personal account, make it an inflation-indexed 
annuity, add the traditional defined benefit pillar that you 
would still keep getting from Social Security, which is also 
inflation indexed, and if the two of those were projected to 
exceed the poverty line, which is also inflation indexed, then 
it would work out. So, technically speaking, it is easy to do.
    Mr. LEVIN. I am going to ask you further questions, because 
the way that Social Security is now indexed, I think people 
would come out differently. I think it is difficult to 
calculate the annuitization so, it would really work, but my 
colleague is here, and so I may ask you to give us an answer 
for the record.
    Dr. MITCHELL. If you could restate the question, that would 
help me, please.
    Mr. LEVIN. I will do that.
    Chairman MCCRERY. Mr. Neal.
    Mr. NEAL. Thank you, Mr. Chairman. Mr. Pollock, I 
appreciated your references to Jake Pickle. For those of us who 
served with him, nobody could more succinctly take care of an 
argument than Jake Pickle. Having said that, I think everybody 
at the table would agree, the President's plan really has gone 
nowhere. In fact, the numbers have gone down the more time he 
spends trying to explain his proposal. As Jake Pickle would 
have said, ``That dog won't hunt.'' Why is the President having 
such difficulty in getting traction with his plan?
    Mr. POLLOCK. Congressman, I would prefer to answer that 
question by saying, how can we improve the plan? I think the 
way to improve it is to make it exceptionally simple and easy 
to understand. The objections to the President's plan, as I 
read them, have been, first of all, it is risky. We are going 
to invest in equities. People have compared investing in 
equities to gambling in Las Vegas and so on. Second, people say 
it will be exceptionally expensive to operate. I am speaking of 
administrative costs. Thirdly, they say you are going to take 
this cash away from the Treasury, and--have to borrow it from 
Asian central banks or something. There is a fourth objection, 
which is that people will be confronted with confusing and 
complex decisions that they don't wish to make, and Wall Street 
will be all over them with dozens of options and things.
    I think if we had a simple plan, which had none of these 
problems, which doesn't suggest to people they will be in Las 
Vegas with the equity market, that they are not obliged to make 
confusing and difficult decisions they don't wish to make, that 
the administrative costs are very cheap, and that it doesn't 
deprive Treasury of any money, they would like it a great deal. 
At least that is my bet, as I said before.
    Mr. NEAL. Thank you. Dr. Mitchell, you indicated during 
your testimony that you thought that private accounts would be 
a good idea based upon a very limited number of opportunities 
to invest?
    Dr. MITCHELL. Correct.
    Mr. POLLOCK. Mr. Chairman, may I be excused?
    Mr. NEAL. Feel free to disagree with this, but I must tell 
you, having been here the last 17 years, the same people in 
this Congress who are suggesting that there be a limited number 
of opportunities to invest, that they all be safe, anybody who 
knows how Washington works--that if we were sitting here five 
or six years ago with the dot-coms, that same chorus would be 
yelling to put that money in riskier investments. Do you know 
what? In this Congress there would be a sympathetic ear for 
trying that. Now, I try to remind people for those of us who 
were here during the Savings and Loan difficulty, that that is 
precisely what happened. You will not be able to rein in this 
Congress on the issue of safe investments. Feel free to comment 
or disagree and maybe suggest how you would argue that we would 
do it then. I think you know of what I speak.
    Dr. MITCHELL. I can answer on two fronts. As you know, in 
the private pension arena, defined contribution plans are 
required to have a minimum of three options, and typically you 
find a stock index fund, a bond index fund, and some sort of a 
cash money market fund. Moreover, in the private sector you 
will typically find employers putting employees into a default. 
If you choose not to choose, then we are going to put you in 
something. It has been a money market in the past, or slowly 
but surely, life-cycle funds are growing in popularity, where 
you start out, maybe, a little bit riskier; the older you get, 
you move into a more conservative account.
    When we were talking on the Social Security Commission in 
2001, we grappled with the issue of cost versus choice in great 
detail. What we proposed was that nobody should be allowed to 
buy a single company stock. I think with Enron and World Com 
and Tyco, in everybody's minds that simply would not be 
permissible. These would have to be indexed accounts, so that 
you wouldn't be able to pick just high-tech or just low-tech, 
but you would have to diversify. I really like the idea of 
life-cycle. I think you are right, most people don't have the 
time, the energy, or perhaps even the education to go and 
decide how to allocate their portfolios, day in and day out; 
plus they probably shouldn't be doing that. So, one thing we 
have seen, for example, is that countries all over the world 
that have defined contribution plans have moved to life-cycle 
funds.
    Mr. NEAL. Do you think there is some validity to the point 
that I raised about the safeguards that have to be built into 
this? Only because, given that S&L issue, when one tracks what 
happened, it is very simple to conclude that people got into 
doing that business when they shouldn't have been in. We have 
deregulated that industry, no questions were asked, and the 
disaster awaited the American people. I am very concerned that 
if this is not meticulously addressed, we run the risk of 
having the dot-com industry drive retirement arguments around 
here. Maybe I can go to Mr. Cavanaugh for a second, Mr. 
Chairman, if I could. I know my time has expired. Mr. 
Cavanaugh, could you talk about how difficult it would be to 
administer a series of small accounts?
    Mr. CAVANAUGH. You want me to talk about the difficulties 
in administering?
    Mr. NEAL. Yes.
    Mr. CAVANAUGH. Well, the problem is, 60 percent of 
businesses have less than five employees; they don't have 
personnel offices and payroll offices. As Mr. Purcell was 
pointing out, 72 percent of them are reporting on paper; they 
are not electronic. They don't have the education, and they 
have been virtually assured by the Administration, as I 
understand it, that they wouldn't have to do any of the things 
that employers in 401(k) plans do. They just can't do it. They 
don't have the education. We are talking about people that cut 
your grass and cut your hair and run the local garage and 
beauty salon and whatever. These are not institutions that can 
handle complex financial matters. The industry--this is not 
speculation--is trying to get in there, the 401(k) industry and 
get these people signed up, and they find they just cannot do 
it at a low enough cost. There are too few people to spread the 
administrative expenses over; it just doesn't work.
    Mr. NEAL. Thank you. Thank you, Mr. Chairman.
    Chairman MCCRERY. Thank you, Mr. Neal. Mr. Levin, do you 
have anything else?
    Mr. LEVIN. No, thank you very much.
    Chairman MCCRERY. Yes, thank you all very much. Mr. 
Pollock, by the way, told us prior to the hearing that he had 
to leave at 12:45. He was not upset with anyone at the witness 
table or, I don't think, with the Members.
    Mr. LEVIN. Can I just say that Mr. Pomeroy isn't here 
because of a Base Re-Alignment Commission commission hearing in 
his home State. He wondered where he should go, and I said it 
is your judgment, but we understand if you decide to attend a 
base closing in your home State. Congresswoman Stephanie Tubbs 
Jones had an emergency she had to take care of, that is why she 
is not here.
    Chairman MCCRERY. Thank you for that information. Thanks, 
once again, very much, and the hearing is adjourned.
    [Whereupon, at 12:50 p.m., the hearing was adjourned.]
    [Submissions for the record follow:]

           Statement of Robert L. Clark, Dillingham, Arkansas

    Mr. Chairman, thank you for the opportunity to provide written 
testimony to the Committee on the need for enhanced financial education 
for American worker in order for them to better plan for their 
retirement income.
    Throughout our economy, more and more American workers are being 
required to assume greater responsibility for their own retirement 
saving. The continuing trend toward greater use of defined contribution 
plans means that a larger proportion of the labor force must make key 
decisions on whether to participate in employer-provided pension plans, 
how much to contribute to these plans, and how to invest plan assets. 
The changing nature of defined benefit plans allowing lump sum 
distributions and the conversion of traditional defined benefit plans 
to cash balance plans are providing more options for workers as they 
change jobs and have the opportunity to access pension funds. Plan 
terminations and reduced generosity of existing plans mean that workers 
must decided whether to increase their own saving in voluntary 
supplemental retirement plans offered by their employers or in their 
own private savings plans. Adoption of individual accounts in 
conjunction with other Social Security reforms will further extend the 
choices individuals face in planning for their retirement and increase 
the responsibility that each of us has for our own retirement income.
    Financial education and a comprehensive understanding of investment 
options are necessary if workers are to successfully achieve their 
retirement objectives. As public and private pension systems shift more 
responsibility to individuals, who bears the responsibility for 
providing an appropriate level of financial knowledge to American 
workers? Is it the responsibility of each worker to acquire sufficient 
knowledge to make appropriate saving decisions? Should firms that offer 
retirement plans be required to provide financial education programs 
for their workers? If in the future, Social Security offers workers the 
option of placing a portion of their contributions into personal 
accounts and deciding how these funds will be invested, does the 
Federal government bear some responsibility to provide adequate 
financial knowledge to Social Security participants so that they can 
make informed choices?
    Available survey data indicates that many workers lack the required 
knowledge to determine how much to save each month to accumulate the 
desired level of income in retirement. In addition, they lack basic 
information on how to invest their retirement funds. The challenge for 
workers, firms, and the government is how can the level of financial 
literacy be improved so workers have the needed information to 
successfully plan for retirement in the coming years.

Employer Pensions and Private Savings
    If individuals have insufficient knowledge concerning the saving 
process, they are unlikely to be able to make optimal retirement plans. 
A lack of financial education may result in workers starting to save 
too late in life and saving too little to reach their retirement goals. 
As a result, they are unlikely to achieve the desired balance between 
consumption while working and consumption in retirement. In addition, a 
lack of information concerning the risk-return distribution of various 
investments might lead workers to misallocate their retirement 
portfolios.
    In a series of recent papers, my co-authors and I have examined the 
response of individuals to participation in retirement education 
seminars. Our research has shown that financial education can produce 
significant changes in how individuals think about and plan for 
retirement. In financial educational programs, workers may learn that 
they have based their desired retirement age and income in retirement 
on insufficient saving and investment behavior. After participating in 
educational programs, many workers tended to revise their goals and 
concluded that they wanted to alter their savings behavior.
    Individuals with low desired retirement ages often reported that 
they raised their expected retirement ages based on the information 
provided in the seminars. In addition, participants that entered the 
seminars with low retirement income goals tended to increase their 
retirement income targets toward a level more consistent with having 
retirement incomes similar to their net income while working. Many 
workers decided to alter their retirement saving behavior and planned 
to open new retirement savings plans and increase contributions to 
existing plans.
    Another important finding is that plans for changes in retirement 
savings made during the seminar were not immediately acted on by many 
respondents. Whether due to inertia, myopia, or changed circumstances, 
many participants who expressed a desire to change their savings 
behavior at the time of the educational program failed to do so in the 
months following the seminar. Thus, it would be useful if participants 
in such programs are offered the opportunity to change their 
contributions to retirement plans at the conclusion of educational 
programs so that they can immediately institute their desired changes.
    Our results indicate that there are significant differences in the 
reaction of individuals to the information presented in the seminars. 
Younger workers were more likely to indicate that they planned changes 
in their retirement savings as were women and participants in clerical 
and blue collar positions. Further research is needed to explore the 
actual responsiveness of participants to educational programs, the 
reasons why desired actions are not taken, and what policies would 
increase the link between desired changes in retirement plans and the 
actions necessary to achieve new retirement goals.
    The results of our studies are interesting and have direct policy 
implications for plan sponsors and workers. The analysis indicates that 
financial education matters and ignorance is not bliss in the area of 
retirement planning. Quality educational programs encourage workers to 
reassess their retirement goals, to make more realistic plans, and to 
change their behavior in order to achieve their objectives. Follow 
through on plans made during a seminar remains problematic and 
introducing methods for immediate action would be useful additions to 
educational programs. Employees, employers, and appropriate government 
agencies should consider how to enhance the financial knowledge of 
American workers so that they will be better able to handle the 
increasing responsibility for their own retirement savings.

Implications for Personal Accounts in Social Security
    Surveys of Americans reveal a dearth of financial knowledge. The 
lack of understanding of financial mathematics, investment options, and 
the level of savings needed for an adequate retirement income means 
that many Americans save too little, start too late, and are unaware of 
the implications of their actions for their standard of living in 
retirement. As life expectancy increases and employer pensions place 
more responsibility on workers for their own retirement income, the 
state of financial education becomes more important. The inclusion of 
individual accounts as a component of Social Security further enhances 
the importance of developing financial education programs for working 
Americans.
    Inadequate financial education will have adverse affects if workers 
are offered personal accounts as part of Social Security. Workers who 
do not understand risk and uncertainty, financial mathematics of 
compounding, and expected rates of return to investment choices are 
likely to make poor decisions concerning the use of personal accounts. 
If Congress chooses to enact Social Security reforms that include 
personal accounts, I believe that Congress should develop financial 
educational materials and programs that provide appropriate information 
for American workers. Such materials should help them make the best 
choices within a future Social Security system.
    This financial information and education should be unbiased and 
provide a clear picture of the investment choices. With such 
information, workers can choose to allocate a part of their 
contributions to these account and make appropriate investment choices 
that will maximize their retirement income or to refrain from selecting 
this option. Without such education, many workers will make poor 
choices based on inappropriate or bad information. As this Committee 
considers Social Security reforms, I recommend that you include as part 
of your deliberations how to provide high quality financial information 
and education to workers so that they will be able to adequately deal 
with the new environment.
    The research showing that individuals respond to employer-provided 
financial education by altering their retirement goals and private 
savings behavior has important implications for Social Security 
personal accounts. Results of my research studies are reported in the 
publications cited below. Thank you for this opportunity to discuss 
with you the importance of financial education and its impact on 
retirement savings.

Research Papers on Financial Education
    Clark, Robert and Madeleine d'Ambrosio. July 2002. ``Saving for 
Retirement: The Role of Financial Education,'' TIAA-CREF Institute 
Working Paper 4-070102-A. Published on-line in Retirement Implications 
of Demographic Family Change Symposium, Society of Actuaries, http://
www.soa.org/library/monographs/retirement_systems/m-
rs_2_tableofcontents.html.
    Clark, Robert and Madeleine d'Ambrosio. 2003. Ignorance Is Not 
Bliss: The Importance of Financial Education, TIAA-CREF Institute 
Research Dialogue No. 78.
    Clark, Robert, Madeleine d'Ambrosio, Ann McDermed, and Kshama 
Sawant. 2003. ``Financial Education and Retirement Savings.'' Paper 
presented at Sustainable Community Development: What Works, What 
Doesn't and Why, a conference sponsored by the Federal Reserve System. 
Washington, March 2003.
    Clark, Robert, Madeleine d'Ambrosio, Ann McDermed, and Kshama 
Sawant. 2004. ``Sex Differences, Financial Education, and Retirement 
Goals'' in Olivia Mitchell and Stephen Utkus (eds.), Pension Design and 
Structure, Oxford, UK: Oxford University press, pp. 185-206.
    Clark, Robert and Sylvester Schieber. 1998. ``Factors Affecting 
Participation Rates and Contribution Levels in 401(k) Plans,'' in 
Olivia Mitchell and Sylvester Schieber (eds.), Living with Defined 
Contribution Plans. Philadelphia: University of Pennsylvania Press, pp. 
69-97.

                                 

   Statement of Craig Copeland, Ph.D. and Jack L. VanDerhei, Ph.D., 
                  Employee Benefit Research Institute
    Individual Social Security Accounts: Administrative Issues

    We are pleased to provide this written testimony on administrative 
issues within proposed Social Security reforms. All views expressed are 
our own, and should not be attributed to the Employee Benefit Research 
Institute (EBRI). Established in 1978, EBRI is committed exclusively to 
data dissemination, policy research, and education on financial 
security and employee benefits. EBRI does not lobby or advocate 
specific policy recommendations; the mission is to provide objective 
and reliable research and information. All of our research is available 
on the Internet at www.ebri.org
    President Bush has made a strong push for the inclusion of 
individual accounts within the Social Security system. The inclusion of 
individual accounts brings forth many issues that are not currently 
present under the traditional defined benefit system structure of 
Social Security. While individual accounts have been debated heavily on 
its political attractiveness, the details of how the system would 
operate or the administration of the individual accounts have not been 
thoroughly discussed. However, the discussion on this issue has greatly 
expanded from the first time the Employee Benefit Research Institute 
focused attention on these administrative issues in their November 1998 
EBRI Issue Brief.
    The political and policy debate has been on going, but from a 
practical perspective, the debate so far has virtually ignored any 
specific considerations about how to administer such accounts. Any 
discussion of whether to create individual accounts must also address 
the basic but critical questions of how they would work: Who would run 
them? What would they cost? Logistically, in what form are they 
possible?
    This testimony presents an overview of the most salient 
administrative issues facing the current Social Security reform 
debate--issues that challenge policy makers to carefully think through 
how their proposals could be implemented, in order to achieve their 
policy goals.
    Speeches, media articles, books, and television reports have 
frequently suggested that if the federal employee Thrift Savings Plan 
can work, and if private employers can make 401(k) plans work, then 
individual accounts in (or in addition to) Social Security can be 
easily administered. There is a way to design a system of individual 
accounts that could potentially be administered in a cost-effective and 
timely way--but for a variety of inescapable reasons, that system most 
likely will bear little or no resemblance to a modern 401(k) plan. If a 
typical Internet-based 401(k) with easy access to account information 
and investment options can be described as the ``Porsche'' of 
retirement savings plans, then the public should realize that a 
workable, cost-effective individual account within Social Security is 
most likely to look like a ``Model T'':

      401(k) plans typically offer an average of 14 actively 
managed investment choices (the `Porsche' offers virtually unlimited 
options through `mutual fund windows' and `self-directed brokerage 
accounts'), versus a very limited number (five for the Federal Thrift 
Savings Plan) of index investment options for a startup universal 
individual account system (probably one initially).
      401(k) plans typically offer daily access through the 
telephone, and the `Porsche'' offers 24/7 internet-based self-
management with immediate access to account information, updated daily, 
versus something closer to an annual account statement for a startup 
universal individual account system.
      401(k) plans typically offer participants loans or 
hardship withdrawals from their accounts, with the `Porsche' providing 
the ability to `do it yourself' on the internet; a startup universal 
individual account system would likely find it impossible, setting 
aside desirability arguments, to offer either
      Workers' 401(k) contributions typically come out of every 
paycheck, with rapid crediting to investment accounts; a startup 
individual account system tied to Social Security would involve `bulk' 
contributions, with annual reporting of contributions to the Social 
Security Administration at the worker level, with crediting as much as 
18 months later as the paper is processed. `Porsche' 401(k) plans do 
both contributions and allocations on an every pay period fully 
automated basis. The startup individual account system could not, as a 
majority of employers file with the government on paper.
      401(k) plans allow participants to modify their 
contributions regularly, with `Porsche' plans allowing it 24/7 on the 
internet for next pay period implementation, versus a more likely once 
a year when the employee fills out their withholding form (per 
employer) for a startup individual account system.
      `Porsche' 401(k) plans rely upon employers and 
administrators to be completely automated with computer interface of 
all data; a startup individual account system would have to allow 
employers to continue using pen-and-paper reports--as most currently 
do--if there was a desire to avoid high new employer administrative 
costs.

    The issues and options in administering individual accounts raise 
concerns that cut across ideology. The object of this report is neither 
to dissuade the advocates nor support the critics of individual 
accounts. Rather, it is to bring practical considerations to a 
political debate that must ultimately deal with the pragmatic 
challenges of designing individual accounts that would not be too 
complex for participants to understand, nor too burdensome for small 
employers to administratively support, nor too difficult for a record 
keeper to administer, nor too expensive for low- and moderate-income 
participants to afford.
    The major findings in this analysis include:

      Adding individual accounts to Social Security could be 
the largest undertaking in the history of the U.S. financial market, 
and no system currently exists that has the capacity to administer such 
a system . The number of workers currently covered by Social Security--
the largest single entitlement program in the nation--is at least four 
times higher than the combined number of all active tax-favored 
employment-based retirement accounts in the United States, which are 
administered by hundreds of entities.
      Direct comparisons between employment-based retirement 
savings plans and Social Security reform are difficult at best . Social 
Security covers workers and businesses that are disproportionately 
excluded from employment-based plans. Because of these differences, a 
system of individual Social Security accounts would be more difficult 
to administer than employment-based plans, and total administrative 
expenses could be larger relative to benefits due to most employers not 
using automatic payroll systems, large numbers not using direct 
deposit, the vast millions of short service and young workers that are 
not included in either public or private employer savings plans, and 
the high relative cost of even one phone conversation with the holder 
of an account (commonly estimated to be an average of $10 per phone 
conversation).
      Credit-based systems such as the current Social Security 
program are less difficult to administer than cash-based systems, which 
must account for every dollar. Inherent in the individual account 
debate is generally the presumption that individual account benefits 
would be based on cash contributions and investment returns. The 
current credit-based system tolerates small errors in wage reporting, 
because they rarely affect benefits. But every dollar counts in a cash-
based individual account system. To ensure that benefits are properly 
provided, an individual account system would require more regulation, 
oversight, and error reconciliation than the current Social Security 
program.
      Social Security individual accounts cannot be 
administered like 401(k) plans without adding significant employer 
burdens--especially on small businesses. Under the current wage 
reporting and tax collection process, it would take at least seven--19 
months for every dollar contributed to an individual's account to be 
sorted out from aggregate payments and credited to his or her IA. This 
seven--19 month ``float period'' could result in substantial benefit 
losses over time. Options for preventing such losses involve difficult 
trade-offs, such as increased government responsibility, increased 
complexity, greater employer burdens, and/or investment restrictions 
for beneficiaries. Elimination of this ``float period'' by requiring 
faster action by small employers would lead to significant new 
administrative burdens and costs.
      If legally considered personal property, the individual 
accounts of married participants could pose significant administrative 
challenges. Social Security today must obtain proof of marriage only at 
the time spousal benefits are claimed. But some individual account 
proposals would require contributions to be split between spouses' 
individual accounts, requiring records on participants' marital status 
to be continuously updated to ensure that contributions are correctly 
directed. Also, dealing with claims on individual account contributions 
in divorce cases could place individual account record keepers in the 
middle of spousal property disputes.
      The current body of knowledge is too uncertain, and in 
general the proposals to date are too vague, to make an objective 
estimate of how much an individual account system would cost to 
administer or whether it would succeed in accomplishing its policy 
goals . Uncertainty exists over how individual account proposals would 
address key policy areas affecting administrative cost and complexity, 
how administrative costs operate in the current employer-sponsored 
retirement arena, and how lessons from the employment-based system 
apply to Social Security reform. For instance, in July 2001 the Federal 
Retirement Thrift Investment Board terminated and sued a contractor for 
failure to design a workable administrative system after nearly three 
years of effort. Given the relatively small size of the Federal Thrift 
Savings Plan (less than 3 million participants) compared with the total 
U.S. workforce (more than 148 million), a great deal could be learned 
by policy makers from this apparent system upgrade failure.
      Individual account benefits would be highly sensitive to 
administrative costs, according to results using the SSASIM policy 
simulation model. Workers born in 1976 and 2026 would receive between 
14 percent and 23 percent lower total benefits under high 
administrative cost assumptions \1\ than under low-cost assumptions, 
indicating that additional research on administrative costs is 
essential to assessing how--or whether--IAs could achieve the lower-
cost assumptions. Proposals to use a flat percentage administrative 
charge could approach the lower-cost assumptions if the system had a s
---------------------------------------------------------------------------
    \1\ For further details about these points, see ``Individual Social 
Security Accounts: Administrative Issues.'' EBRI Issue Brief no. 237, 
September, 2001. (www.ebri.org/publications/ib)

---------------------------------------------------------------------------
                                 

                Statement of Paul A. Cyr, Greene, Maine

    Thank you for giving me this opportunity to write to you.
    I am fifty-eight and a half years old and I work for the state of 
Maine Department of Transportation as a Highway Worker II. My jog is 
driving truck--winter and summer. And when I am not actually driving I 
am doing heavy physical work. I've worked for the state of Maine for 
approximately ten and a half years. The first three and a half years 
were as a Highway Worker I. The job was mainly flagging for eight to 
ten hours a day, and at forty--seven years old it wasn't easy standing 
on hot top all day or being out in 10 below zero weather. While doing 
this job I've had the driving public swear at me for holding them up 
for three minutes, or people going by and hollering at me to get a real 
job. There have also been numerous times when I was almost hit by cars 
and in some cases the drivers actually laughed about it. I took the 
insults and obscenities thinking I would just do what I had to and it 
would pay off in the long run--that when I retired I would have a 
pension from this job to go along with my social security.
    Before I went to work for the state of Maine I was a sheet metal 
journeyman. I spent four years going to school at night to get my state 
license as a sheet metal worker. I worked for a company while going to 
school and stayed with them for approximately twenty-five years. After 
that I worked for another metal shop for about four years until they 
filed for bankruptcy. And than it was another metal shop company for 
three years until I was let go for supposedly not being able to keep up 
with the younger people. These companies had very few benefits--no 
pension, no paid vacations, no bonuses, some paid holidays, and some 
had very limited health insurance. During all those years I installed 
duct work in dirty paper mills, in buildings with asbestos, and out in 
the cold and heat. I was paying into Social Security during this time 
and thought I would have a SS check when I retired. I couldn't put 
money into savings for retirement because it took all I had to be able 
to get by and pay my bills.
    Than after working with no benefits or pension all those years and 
being out of work for two years, I landed a job with the state of 
Maine. After being a state worker for seven years I went to a 
retirement seminar and learned about the GPO and WEP. I was very upset 
by what I was told, so I went to my Social Security office in Auburn, 
Maine. The person I spoke to told me I would loose about a third of my 
Social Security benefits because I worked for the state of Maine. To 
add insult to injury, she than told me that I should not have taken 
this job, but should have found work somewhere else! But as I told her, 
I had been out of work for two years when I got the job with the state 
and figured that with their pension and Social Security I would be able 
to get by when I retired. But now the way I feel they might as well 
bury me in my work clothes, because I'll probably be working until I 
die.
    From what I see the future does not look very good for me. With 
taxes going up all the time, as well as the cost of living (gas, 
lights, insurances, etc.), I'll never have enough to make ends meet if 
I retire. Also, there has been a woman in my life for many years but we 
can't get married, because if we do and she draws any of my state of 
Maine pension after I die she will get penalized on her Social 
Security. And I won't do that to her. Why should she get penalized 
anyway?
    In my opinion the GPO and WEP should be repealed. Just give me the 
money that I earned and put into Social Security over the years--all of 
it, no more and no less. And do not penalize my fiancee either. Please, 
please repeal these two unjust laws so people like myself that have 
worked so hard for so many years (four of which were serving this 
country in the military) can end our working days with dignity an a 
sense of self worth and accomplishment.

                                 

          Statement of Ike Jones, America's Community Bankers

    America's Community Bankers (``ACB'') \1\ is pleased to submit this 
written statement in connection with the Subcommittee's eighth hearing 
on ``Protecting and Strengthening Social Security.'' ACB commends 
Chairman McCrery and Chairman Thomas for their leadership in crafting 
social security reform legislation. ACB firmly believes that personal 
accounts should be a part of any social security reform legislation, 
and that a ``Community Bank Option'' should be available to workers 
under a system of personal accounts.
---------------------------------------------------------------------------
    \1\ America's Community Bankers is the member driven national trade 
association representing community banks that pursue progressive, 
entrepreneurial and service-oriented strategies to benefit their 
customers and communities. To learn more about ACB, visit 
www.AmericasCommunityBankers.com.
---------------------------------------------------------------------------
Personal Savings Accounts
    The debate over creating a more solvent Social Security retirement 
system has gained momentum over the past few years, especially with the 
Administration putting the issue front and center. Because Social 
Security operates almost entirely as a pay-as-you-go system, it is 
highly sensitive to the dramatic demographic changes that are 
increasing the average ages of our population. Increased life 
expectancies mean more retirees collecting benefits for more years. The 
resulting decline in the ratio of workers to beneficiaries is pushing 
the system toward insolvency, and personal savings accounts may be one 
way to create a more secure Social Security system.
    For several years now, ACB has supported allowing workers the 
choice of investing at least a part of their Social Security taxes in 
personal accounts. This approach will give individuals greater control 
over how their retirement security funds are invested and can create a 
more solvent system. Social Security reform should give workers the 
option of relying on their community banks and the investment products 
those banks offer, including insured deposits, in addition to those 
investment options available on Wall Street.

Community Banks Are Experienced Investment Advisers
    Community banks already offer a variety of retirement investments, 
including FDIC-insured Individual Retirement Accounts (IRAs), 
certificate of deposits (CDs), mutual funds and annuities. Because 
community bankers understand the creation of wealth and their 
customers, they are in an excellent position to help consumers choose 
appropriate investments for these personal accounts.
    ACB believes that any Social Security reform should ensure that 
consumers have the option of seeking advice on their personal accounts 
from people they already trust--their local community banker.

FDIC-Insured Accounts (the Community Bank Option)
    For decades, American workers have trusted FDIC-insured financial 
institutions with their retirement savings. As of the end of 2003, 
banks and thrifts managed $246 billion in retirement funds invested in 
IRAs and Keogh plans.
    FDIC-insured retirement accounts should be an option for workers 
choosing personal accounts. Workers of all ages could benefit from an 
FDIC-insured deposit option or what is now being called the Community 
Bank Option. Some workers, while wanting a greater return than the 
Social Security program currently provides, will be wary of investing 
all of their retirement funds in equities and other retirement products 
that carry much higher risk. For these workers, a long-term deposit 
account would be the most appropriate investment for all or part of the 
funds freed up under Social Security reform legislation to invest in 
personal accounts.
    Certainly, as workers near retirement age, their tolerance for 
taking risks in the stock or bond markets will dramatically decrease. 
ACB believes that FDIC-insured accounts would benefit these older 
workers not only as a place to invest new funds, but also as a safe 
product in which to rollover funds from riskier personal account 
products.
    In addition, FDIC insurance has given millions of American families 
the confidence for over 70 years that the money they deposit in banks 
will be there when they need it. Allowing FDIC-insured accounts as an 
option under Social Security reform would encourage workers to choose 
the personal account option and increase support for reform among 
consumers.

The Facts Support FDIC-Insured Account Option
    Many Americans depend on FDIC and NCUSIF-insured IRAs as part of 
their retirement savings plans. According to a January 2005 report of 
the Employee Benefit Research Institute, American families had $270 
billion invested in FDIC and NCUSIF-insured IRAs as of the end of 2003. 
These funds represent 9 percent of all IRA assets.
    In addition, a 2001 Federal Reserve survey of consumer finances 
showed that 12.3 percent of all American households held insured IRA 
deposit accounts. The survey also indicates that households in all age 
brackets rely on these insured accounts for retirement savings, not 
just households headed by older Americans. American workers should also 
have insured deposit accounts as an option under any program of 
personal investment accounts.
    Age Distribution of IRA Accounts at Insured Institutions

------------------------------------------------------------------------
                                     Percent of all
                                      households in   For each age group
                                     each age group     IRA deposit at
     Age of head of household         that have IRA         insured
                                       accounts at      institutions as
                                         insured       percent of total
                                       instutions          IRA funds
------------------------------------------------------------------------
< 35                                5.3               25.0
------------------------------------------------------------------------
35-49                               11.1              116.4
------------------------------------------------------------------------
50-64                               17.0              18.4
------------------------------------------------------------------------
65-69                               18.5              28.0
------------------------------------------------------------------------
> = 70                              16.1              27.4
------------------------------------------------------------------------
All Ages                            12.3              21.2
------------------------------------------------------------------------
Data were provided by Federal Reserve Board staff based on the 2001
  Survey of Consumer Finances (the most recent survey available)
  published in 2003.
IRA accounts refer to all IRA and Keogh accounts, excluding IRA-SEPs and
  similar accounts maintained as part of an employer-provided retirement
  benefit.
Insured institutions include banks, thrifts, and credit unions.
IRA deposits refer to the sum of IRA account balances of all household
  members at all insured institutions used by the household.

Community Bank Option Invests In Local Communities
    Allowing investments in FDIC-insured accounts returns money to the 
local communities where taxpayers live and work. Community banks will 
invest these funds in their local communities by providing loans to 
local businesses, mortgage loans to families, education loans to 
students headed off to college, and in many other ways. If this option 
is adopted, it could result in reduced rates for these loans.
    Increase Deposit Insurance
    Providing a substantial increase in deposit insurance coverage for 
all retirement accounts would further enhance the community bank role. 
Currently, the FDIC provides up to $100,000 of deposit insurance for 
the retirement accounts (e.g. IRAs and Keoghs) of a depositor in a 
bank. The shift in America from defined benefit plans to IRA and 
401(k)-type savings has increased the burden on individuals to manage 
their own assets. Retirement assets often exceed the current $100,000 
coverage limit by substantial amounts. A substantial increase in FDIC 
coverage of retirement accounts would strengthen the viability of the 
insured deposit account option. And it is important to note, the FDIC 
is funded through assessments on banks and savings associations, not 
taxes.

Conclusion
    ACB strongly believes that FDIC-insured retirement accounts, or the 
Community Bank Option, should be available for workers choosing 
personal accounts. Workers of all ages could benefit from an FDIC-
insured deposit option. Some workers will be wary of investing their 
retirement funds in equities and other retirement products that carry 
risk. For these workers, an FDIC-insured long-term deposit account 
would be the most appropriate investment for all or part of the funds 
freed up by Social Security reform.

                                 

           Statement of Renee Lancon, West Hills, California

    Our community is proud of our strong work ethic. We believe in 
rewarding and supporting hard work and citizens who make contributions 
to our society. That's why our congressional delegation simply must 
support legislation (Representatives McKeon (R-CA) and Berman (D-CA) 
have introduced the Social Security Fairness Act of 2005 (H.R.147) to 
address the unfair cuts to the retirement benefits of public employees 
who have dedicated their lives to serving their communities and their 
country. I urge our Senators and Representatives to support legislation 
to address these discriminatory penalties for public service: 

Government Pension Offset (GPO) and Windfall Elimination Provision 
(WEP), and on the issue of mandatory Social Security coverage
    Too much is at stake to ignore this common-sense legislation!
    I have recently retired and was shocked to find out when I went to 
the Social Security Office last week that I cannot collect ANY of my 
spouse's Social Security OR my own earned Social Security!
    I earned my own 40 quarters when I worked in jobs other than 
teaching and cannot collect my own! I also cannot collect on my 
husband's because I chose to spend my career teaching!!!
    I also moved from one state to another to follow my husband's 
career and was not permitted to leave my ``retirement'' in the previous 
state, so have worked as a teacher since 1962, and retired recently and 
only have ``credit'' for 20 years of teaching! It is NOT enough to live 
on!!!
    I was supposed to be able to collect $500 a month IF I had not been 
a teacher!!!
    The Social Security would have helped enormously! I also help 
support two children, one of whom is handicapped!!!
    This Offset is unfair, inequitable and discriminatory. Most 
teachers are women and this affects them enormously! We have a severe 
teacher shortage in Los Angeles and in many other places and this 
hinders our recruitment efforts even more!
    And, the ``pension'' I do collect, I paid into!!! I also saved in a 
403B. I did all the things that I thought would allow me independence 
after a long career of public service! But, NOW, I am on the brink of 
poverty after dedicating myself to teaching special needs children for 
almost 40 years!!!
    Nine out of 10 public employees affected by the GPO lose their 
entire spousal benefit, even though their deceased spouse paid Social 
Security taxes for years. The WEP causes low-paid public employees who 
work both inside and outside the Social Security system to lose up to 
60 percent of their Social Security benefits. The loss of these 
benefits may make some people eligible for poverty-based assistance, 
such as food stamps.
    Please join
    I have also attached additional information from National Education 
Association's (NEA)
    On behalf of the 2.7 million members, we would like to thank you 
for the opportunity to submit comments on the Government Pension Offset 
(GPO) and Windfall Elimination Provision (WEP), and on the issue of 
mandatory Social Security coverage. We commend the Subcommittee for 
holding this important hearing on a matter of great concern to 
educators and other public employees.
    NEA strongly supports complete repeal of the Government Pension 
Offset and the Windfall Elimination Provision, which unfairly reduce 
the Social Security and Social Security survivor benefits certain 
public employees may receive. We oppose requiring public employees to 
participate in Social Security. Our testimony will cover both of these 
issues.

The Government Pension Offset: A Devastating Loss of Benefits for 
        Widows and Widowers
    The Government Pension Offset reduces Social Security spousal or 
survivor benefits by two-thirds of the individual's public pension. 
Thus, a teacher who receives a public pension for a job not covered by 
Social Security will lose much or all of any spousal survivor benefits 
she would expect to collect based on her husband's private-sector 
earnings.
    Congress and the President agreed in 1983 to reduce the spousal 
benefits reduction from a dollar-for-dollar reduction to a reduction 
based on two-thirds of a public employee's retirement system benefits. 
This remedial step, however, falls well short of addressing the 
continuing devastating impact of the GPO.
    The GPO penalizes individuals who have dedicated their lives to 
public service. Nationwide, more than one-third of teachers and 
education employees, and more than one-fifth of other public employees, 
are not covered by Social Security and are, therefore, subject to the 
Government Pension Offset.
    Estimates indicate that nine out of 10 public employees affected by 
the GPO lose their entire spousal benefit, even though their deceased 
spouse paid Social Security taxes for many years. Moreover, these 
estimates do not include those public employees or retirees who never 
applied for spousal benefits because they were informed they were 
ineligible. The offset has the harshest impact on those who can least 
afford the loss: lower-income women. Ironically, those impacted have 
less money to spend in their local economy, and sometimes have to turn 
to expensive government programs like food stamps to make ends meet.
    NEA receives hundreds of phone calls and letters each month from 
educators impacted by the GPO. Many are struggling to survive on 
incomes close to poverty, fearing they will be unable to cover their 
housing, medical, and food expenses on their meager incomes. For 
example, consider the following stories:
    From NEA member Frances in Louisiana:
    ``My husband, a Baptist minister, passed away [in 2001] after 
paying Social Security for 42 years. At times we had to take a second 
loan on our home to pay the Social Security. Now, I had to pay back the 
loan, but discovered that I will not get benefits because I receive a 
small teacher retirement''
    From NEA member Stella in Colorado:
    ``I am a 72-year-old widow. . . . I was happily married to the same 
man for 391/2 years. My husband was a World War II disabled veteran who 
worked and paid into Social Security for 50 years. . . . He passed away 
11 years ago thinking I would be able to receive his Social Security 
and Veterans Widow pension. . . . But now I'm living in poverty.''
The Windfall Elimination Provision: A Shocking Loss of Earned Benefits
    The Windfall Elimination Provision reduces the earned Social 
Security benefits of an individual who also receives a public pension 
from a job not covered by Social Security. Congress enacted the WEP 
ostensibly to remove an advantage for short-term, higher-paid workers 
under the original Social Security formula. Yet, instead of protecting 
low-earning retirees, the WEP has unfairly impacted lower-paid retirees 
such as educators.
    The WEP penalizes individuals who move into teaching from private-
sector employment, or who seek to supplement their often insufficient 
public wages by working part-time or in the summer months in jobs 
covered by Social Security. Educators enter the profession often at 
considerable financial sacrifice because of their commitment to our 
nation's children and their belief in the importance of ensuring every 
child the opportunity to excel. Yet, many of these dedicated 
individuals are unaware that their choice to educate America's children 
comes at a price--the loss of benefits they earned in other jobs.
    While the amount of reduction depends on when the person retires 
and how many years of earnings he or she has accumulated, many public 
employees can lose a significant portion of the Social Security 
benefits they earned in other jobs. Like the GPO, the WEP can have a 
devastating impact on educators' retirement security. For example:
    From NEA member Carolyn in Kentucky:
    ``I started a direct sales business from my home at nights and 
weekends to supplement my teacher retirement. I earned my necessary 
quarters, reached my 62nd birthday, and then learned of the Windfall 
Elimination Provision. I was told that I was eligible to receive 
approximately $158 monthly; however, because of the WEP, this would be 
reduced to $78 a month. By the age of 65, my payments had risen to $84, 
but after paying $66 for Part B of Medicare, I now have $18 to deposit. 
I have been forced because of the economics of the day to return to the 
classroom to substitute teach for a paltry sum of $61 a day. . . . This 
is certainly not the American dream I had in 1956 to become a 
teacher!''
The ``Double Whammy'': Educators Impacted by Both the GPO and WEP
    Many NEA members report that they are subject to double penalties--
losing both their own benefits and spousal benefits due to the combined 
impact of the GPO and WEP. For example NEA member Martha from Texas 
reports:
    ``By 1978, when I started my teaching career, I had already earned 
my 40 quarters of Social Security and over the years depended on these 
benefits as part of my retirement. I should be entitled to $415 a month 
at the age of 62. However, because of the Windfall Elimination 
Provision, I will now be entitled to $206 a month, and this reduction 
in my earned retirement is a big loss. [In addition], according to the 
Social Security Administration, I should be entitled to approximately 
$970 a month for widow's benefits. However, because of the Government 
Pension Offset, I can only receive $21 a month. Both the Government 
Pension Offset and Windfall Elimination Provision are devastating to 
teacher retirees and me.''

The National Impact of the GPO and WEP: Undermining Teacher Recruitment 
        Efforts
    The GPO and WEP have an impact far beyond those states in which 
public employees like educators are not covered by Social Security. 
Because people move from state to state, there are affected individuals 
everywhere. The number of people impacted across the country is growing 
every day as more and more people reach retirement age.
    Perhaps most alarming, the GPO and WEP are impacting the 
recruitment of quality teachers to meet urgent national shortages. 
Record enrollments in public schools and the projected retirements of 
thousands of veteran teachers are driving an urgent need for teacher 
recruitment. Estimates for the number of new teachers needed range from 
2.2 million to 2.7 million by 2009.
    At the same time that policymakers are encouraging experienced 
people to change careers and enter the teaching profession, individuals 
who have worked in other careers are less likely to want to become 
teachers if doing so will mean a loss of Social Security benefits they 
have earned. Some states seeking to entice retired teachers to return 
to the classroom have found them reluctant to return to teaching 
because of the impact of the GPO and WEP. In addition, current teachers 
are increasingly likely to leave the profession to reduce the penalty 
they will incur upon retirement, and students are likely to choose 
other courses of study and avoid the teaching profession.
    The GPO and WEP also impact other critical public services fields, 
including police and firefighters. Our nation can ill-afford to allow 
the very real fear of poverty in retirement to force talented, 
dedicated individuals out of these professions.

The GPO/WEP Solution: Total Repeal
    Representatives McKeon (R-CA) and Berman (D-CA) have introduced the 
Social Security Fairness Act of 2005 (H.R.147). This bipartisan 
legislation, which already has over 260 cosponsors, would eliminate the 
GPO and WEP, thereby allowing public employees, like all other 
employees, to collect the benefits they earned and need. NEA urges the 
Subcommittee, and the entire House of Representatives, to take 
immediate steps toward passage of the McKeon-Berman Bill.

Mandatory Coverage: An Unwise and Unnecessary Approach
    NEA's position on repeal of the Government Pension Offset and 
Windfall Elimination Provision should not in any way be interpreted as 
support for requiring public employees to participate in Social 
Security. NEA strongly opposes mandatory coverage. Instead, NEA simply 
believes that educators should be able to receive the benefits they or 
their spouse earned by working in covered employment, without 
jeopardizing their public pension.
    Many existing public employee programs are tailored to meet the 
needs of specific employee groups. Forcing educators into Social 
Security would jeopardize these state and local plans. In addition, 
Social Security trust funds can be invested only in U.S. Treasury 
bonds. State and local governments permit a greater diversity of 
investment options, thereby potentially achieving a greater rate of 
return.
    Mandatory coverage of educators would also increase the tax burden 
on public-sector employers. Ultimately, these increased tax obligations 
would lead to difficult choices, including reducing the number of new 
hires, limiting employee wage increases, reducing cost-of-living 
increases for retirees, and reducing other benefits such as health 
care.
    Finally, mandating coverage of educators will not solve the Social 
Security system's financial difficulties. The amount of money gained by 
mandating coverage would be relatively small and would not solve the 
long-term Social Security crisis. Requiring new state and local 
employees to pay into Social Security would enable the federal 
government to continue borrowing money from Social Security trust 
funds, and, therefore, could exacerbate financing problems.
    We thank you for your consideration of these comments.

                                 

  Statement of Olivia S. Mitchell \1\, Wharton School, University of 
                Pennsylvania, Philadelphia, Pennsylvania
---------------------------------------------------------------------------
    \1\ Olivia S. Mitchell is the International Foundation of Employee 
Benefit Plans Professor and Professor of Insurance and Risk Management; 
she is also Executive Director, Pension Research Council and Director, 
Boettner Center for Pensions and Retirement Security, all at The 
Wharton School of the University of Pennsylvania. The views offered 
here are solely those of the author and do not represent those of any 
institutions with which she is affiliated.
---------------------------------------------------------------------------
    Mr. Chairman and members of the Subcommittee: Thank you for the 
opportunity to appear here today. My name is Olivia S. Mitchell, and I 
am a Professor of Insurance and Risk Management at The Wharton School 
at the University of Pennsylvania.
    As you know, Social Security faces imminent insolvency, with 
payroll tax revenues threatening to fall below benefit payments within 
6 years. The present system also contains many inequities and anomalous 
redistribution patterns, and it offers current workers a surprisingly 
low and very risky return. \2\
---------------------------------------------------------------------------
    \2\ Cogan and Mitchell (2002).
---------------------------------------------------------------------------
    The bipartisan Commission to Strengthen Social Security (CSSS), on 
which I served in 2001, believed that offering two separate tiers under 
a reformed Social Security program, each with its own function, would 
improve the overall program's transparency and equity. Social adequacy 
was to be the principal objective of the traditional defined benefit 
piece, while individual equity was seen as the goal of a personal 
accounts component.
    My testimony before this Subcommittee today focuses on two aspects 
that must evaluated in designing a Personal Retirement Account element 
as part of a reformed Social Security: (1) administrative fees and 
charges, and (2) payout issues. My views derive from the research 
literature on administrative fees and payout issues, particularly 
regarding how Personal Retirement Accounts might be invested and how 
the funds at retirement might be deployed. The views I offer are my own 
and do not represent those of any institutions with which I am 
affiliated.
    My conclusions are that the voluntary Personal Retirement Accounts 
(PRAs) should be formulated so that:

      They offer participants some investment choice while 
still being relatively inexpensive;
      They standardize disclosure regarding fees and charges so 
participants can understand and compare them;
      They require retirees to annuitize part of their 
retirement assets in their Personal Accounts, so that the combined 
benefit payments from Social Security will keep them out of poverty.
Administrative Fees and Charges
    Experience with public and private pension plans the world over 
indicates wide disparity in reported administrative fees and charges 
across systems. Several lessons are worth highlighting:

      Measuring pension expenses requires standardized 
reporting and disclosure standards. Pension systems often structure 
their charges in bewildering ways. For instance, fees can be levied as 
flat commissions, a percent of contributions, or a percent of the 
fund's annual yield. \3\ Such complexity makes it difficult for plan 
participants to compare fund performance. A sensible response, adopted 
by many Latin American pension supervisors, is to require disclosure 
using a standardized table for reporting charges. This has the effect 
of increasing the information available to participants and hence, 
making the market more competitive. A more problematic tactic adopted 
by the UK, for example, is to set a national fee cap. This may limit 
competition and reduce participants' focus on holding down costs.
---------------------------------------------------------------------------
    \3\ Mitchell (1998).
---------------------------------------------------------------------------
      Scale is important in keeping costs down. \4\ Larger 
money managers benefit from scale economies, centralized fund 
administration, and centralized collection of contributions. For 
example, in Australia, retail financial service providers charge three 
times more in pension fees and charges than do institutional managers 
of corporate pensions. While there is little agreement on the minimum 
size of a cost-effective pension, managers of large defined 
contribution plans such as the Federal Thrift Savings Plan which covers 
civil servants and military employees, my University's retirement plan 
(TIAA-CREF and Vanguard), and others, charge pension participants 
annual fees between 0.1-0.4% of assets under management. These fees are 
well below what savers pay in typical Individual Retirement Accounts.
---------------------------------------------------------------------------
    \4\ See Mitchell (1998), Bateman and Mitchell (2004), and 
Whitehouse (2005).
---------------------------------------------------------------------------
      Private retirement systems might seem to be more costly 
than Social Security, but this is a misleading conclusion as they 
generally offer more and different services. Some have suggested that 
the current U.S. Social Security system is one of the lowest-cost 
programs around. Nevertheless, Social Security does not provide the 
wide range of services provided by modern managers of asset-backed 
retirement accounts. For instance, the government program does not 
invest in the capital market, it holds no insurance-type reserves even 
though it offers disability and survivors' insurance, and it takes a 
very long time--more than a year--to post workers' contributions to 
their records. \5\ By contrast, privately managed fund providers would 
and can do better by taking advantage of modern technology.
---------------------------------------------------------------------------
    \5\ Mitchell (1998).

    Taking these and other factors into account, I and other Commission 
members concluded that it would be reasonable to establish personal 
accounts along the lines of the Federal Thrift Saving Plan. 
Accordingly, and for a few years into the system, a central Governing 
Board would be charged with collecting contributions, managing records, 
and selecting private-sector managers who would invest participant 
assets via a competitive bidding process. This Board could either 
handle record-keeping and benefit payments itself, or these functions 
could be outsourced via a competitive process.
    We also proposed that investment choices in the personal accounts 
would be limited but diverse. The options suggested include:

      a Government Securities Investment fund (mainly short-
term U.S. Treasury securities);
      a Fixed Income Index Investment fund (tracking a U.S. 
bond market index);
      a Common Stock Index Investment fund (tracking the 
Standard & Poor's 500 Index of large-company stock);
      a Small Capitalization Stock Index Investment fund 
(tracking the Wilshire 4500 stock index); and
      an International Stock Index Investment fund; and
      a fund that invests in Government Treasury Inflation-
Protected Securities.

    At some later date, plan participants might be permitted to move 
their investments to licensed, supervised, private money managers 
offering an approved set of low-cost investment options. The benefit 
levels that might be expected from alternative investment approaches 
for Personal Retirement Accounts appear in Table 1, along with a 
comparison of current benefits, payable benefits, and scheduled 
benefits.
    The Office of the Chief Actuary at Social Security estimated that 
the proposed CSSS approach would be quite inexpensive, costing only 
about 0.3% of assets annually.

  Table 1: Monthly Social Security Benefits Under Alternative Scenarios
                 Projected to 2052 (CSSS  Model  2  $01)



I. Lifetime low-wage earner\*\
Today's benefit                               $637
Projected Benefit With Personal Account:
      Low yield                               867
      Medium yield                            1,050
      High yield                              1090
Current Program Payable                       713
Scheduled benefit                             986
II. Lifetime medium-wage earner\*\
Today's benefit                               $1,052
Projected Benefit With Personal Account:
      Low yield                               1,204
      Medium yield                            1,525
      High yield                              1,595
Current Program Payable                       1,179
Scheduled benefit                             1,628
III. Lifetime maximum-wage earner\*\
Today's benefit                               1,366
Projected Benefit With Personal Account:
      Low yield                               1,565
      Medium yield                            1,907
      High yield                              1,983
Current Program Payable                       1,557
Scheduled benefit under current law           2,151

\*\ These categories, developed by Social Security actuaries, are
  specified (in $01) such that a lifetime ``low'' earner would have
  averaged approximately $15,900 per year, whereas the medium earner
  averaged $35,300 per annum and the high earner $56,400.

Source: Cogan and Mitchell (2003)

Payout Issues
    When considering how to structure payouts from voluntary Personal 
Accounts under a reformed Social Security system, naturally the 
question arises as to whether and how access to the funds should be 
permitted. CSSS members agreed that pre-retirement access to the money 
should not be allowed to `leak' out before retirement, as early 
consumption would likely increase the chances that the elderly would 
then have to rely on old-age antipoverty programs. Yet, as the 
Commission pointed out, ``a clear appeal of personal retirement 
accounts is that they grant workers ownership over their own assets.'' 
After weighing competing arguments, we concluded that personal accounts 
should be preserved until the nationally-agreed on early retirement 
age, consistent with current Social Security policy which does not 
permit pre-retirement access to old-age benefits.
    By contrast there is more discussion regarding appropriate designs 
for the pension decumulation process under Personal Accounts. This 
refers to the process by which older participants access their 
retirement assets, how they invest their money during retirement, and 
whether annuities--which are financial products designed to cover the 
risk of retirees outliving their assets--should play a central role. 
Regarding post-retirement fund management, my Commission recommended 
several methods of drawdown including phased withdrawals and annuities, 
as well as possibly lump sums.
    To highlight the importance of longevity risk, Table 2 shows that a 
65-year old U.S. male can anticipate living to age 81, but he has 
almost a 20% chance of living to age 90 or beyond. A woman of the same 
age can expect to live to 85, but she has more than a 30% chance of 
living to age 90 or older (Table 2). In other words, people face 
substantial risk of outliving their life expectancy, implying 
substantial uncertainty regarding how long one must conserve and spend 
retirement assets, combined with a high probability of running out of 
money.

   Table 2: Remaining Life Expectancy and Survival at Age 65 (in 2000)

                     Remaining Life                         Men    Women

Expectancy (years):                                        16.4    19.6
Probility of Surviving to Age:
         70                                                 88%     92%
         75                                                74      82
         80                                                56      69
         85                                                36      51
         90                                                18      31
         95                                                 6      14
        100                                                 1       4


Source: Mitchell and McCarthy (2004)


    A life-long annuity can help protect against this risk, by paying a 
premium to an insurer who then pools a number of people with similar 
longevity expectations. Though some have argued that such insured 
products seem expensive, my research shows that the ``money's worth'' 
(MW) of such life income products is rather substantial. The MW refers 
to the discounted cash flow of the lifetime payments received divided 
by the product premium. For example, Table 3 shows that U.S. purchasers 
of an immediate single-life annuity would expect back 93 cents on the 
dollar from a life annuity; in exchange purchasers have the insurance 
value that they will never outline their lifetime benefit payments. The 
MW ratios are similar in Australia, Italy, and the UK.


 Table 3. Money's Worth of Single Premium Nominal Life Annuities for 65-
  Year Olds: An International Comparison (using country Treasury yield
                    curves and annuitant life tables)

                                 Australia    Canada   Italy   UK    US

Men                             0.986        1.014    0.958   0.96  0.92
                                                               6     7
Women                           0.970        1.015    0.965   0.95  0.92
                                                               7     7

Source: Derived from Mitchell and McCarthy (2004)


    These issues are complex and potentially politically delicate, 
since some workers will fail to accumulate much in their accounts over 
their worklives; also some retirees might anticipate relatively lower-
than-average life expectancies, making forced annuitization seem 
punitive.
    In balancing the various choices for payout design, the Commission 
concluded that partial annuitization should be mandated so that ``the 
yearly income received from an individual's Social Security benefit 
plus the joint annuity (if married) would protect either spouse from 
falling below the poverty line during retirement'' (CSSS 2001). Any 
funds above those needed to buy the minimum annuity could be accessible 
as a lump sum and/or bequeathed at death. This approach has the dual 
benefit of both protecting the retiree from falling below the poverty 
line while still allowing some access to the funds accumulated in the 
Personal Retirement Account.
    Remaining design issues include how retirees would learn about 
annuity products, who would sell them, and whether the private 
insurance market can do a good job meeting market demand. To date, 
relatively few consumers have purchased payout annuities, making it a 
bit difficult to forecast how the market will develop. Several key 
issues will have to be decided:

      Which annuity products will be offered and to whom?

    Currently private insurers in the U.S. offer a wide and very 
complex array of annuity products, including immediate versus deferred 
benefit payments; fixed nominal payouts versus programs with escalating 
or variable payouts; and term certain versus other payment periods. 
Also annuities offered through company pensions are mandated to use 
unisex mortality tables whereas retail annuities do not.
    A logical lesson from the behavioral finance literature is that it 
would be sensible to establish a ``default'' payout format such as a 
joint and survivor inflation-linked or escalating life annuity, which 
retirees would automatically receive unless they specifically opted for 
something else. As a case in point, retirees in the UK are required to 
annuitize their pension assets at age 75; in Germany, workers with 
assets in so-called Reister-pensions may take 20% of their accumulated 
assets in a lump sum, another 20% in a phased withdrawal format; but at 
age 85, the retiree must annuitize his balance and the benefit may not 
be lower than the periodic payment received before that age.
    Of course, since many retirees are not accustomed to thinking about 
longevity risk, they would require financial education to help them 
clearly understand the costs and benefits of different ways to manage 
their Personal Retirement Account assets.

      Which annuity providers will be allowed in the market, 
and how will they be regulated?

    Evidence from other countries adopting personal accounts indicates 
that private insurers can and do offer the types of products that 
retirees want. For instance, in Chile, middle and upper income workers 
generally prefer the annuity payout over a phased withdrawal approach 
to retirement drawdowns.
    Nevertheless, there will likely have to be some governmental 
oversight over the annuity market. In Mexico, for instance, all 
insurers are required to bid on all retirees, and when issuing annuity 
bids, the companies may learn only a retiring worker's age and sex (but 
not his identity, his health status, or his account balance). This 
reduces the chances of ``cherry-picking'' rich retirees or those 
anticipated to die soon.
    Another issue has to do with whether unisex mortality tables would 
be required for the annuities. Doing so, of course, involves 
redistribution of wealth away from shorter-lived men and toward longer-
lived women, which is already true in the current Social Security 
System. Requiring joint and survivor benefits as a default would render 
this issue less important quantitatively.

      What role, if any, would the federal government have?

    As an alternative to building up private annuity markets, some have 
suggested that the federal government might directly sell the mandatory 
annuities under the new system. \6\ While this might hold down some 
costs, it can cause other problems. For example, there could be 
political interference associated with investing the annuity reserves--
amounting to 15% of GDP at maturity--and it raises questions about 
whether the reserves could truly be saved, or whether they would be 
`spent' akin to Social Security Trust Fund assets. Further, the 
government would then have responsibility for mortality and capital 
market risk, which would likely be incorrectly priced and managed.
---------------------------------------------------------------------------
    \6\ NASI (2005).
---------------------------------------------------------------------------
    One key role for the federal government in this context has to do 
with tax and transfer policy. For instance, pension and Individual 
Retirement Account assets are protected in bankruptcy but are divisible 
in divorce; whether the same treatment would be afforded PRA annuities 
and assets has yet to be determined. Conversely, annuity flows and lump 
sums are generally `counted' when retirees apply for SSI and Medicaid 
benefits; payouts are taxed as income. Whether and how PRA assets and 
annuities are to be treated for tax and transfer purposes--as well as 
others (e.g. the estate tax vulnerability of the PRA assets if the 
worker or spouse dies) will take additional work to get it right.
    Another role for the government is to enhance the range of 
investments available to insurers providing the products. \7\ Many 
writers have noted the key role of federal government provision of 
inflation-indexed bonds sufficient to meet market demand. Expanding 
their supply would allow private insurers to offer the kinds of indexed 
annuity products that would give retirees better protection against 
inflation, which is a source of substantial retirement insecurity.
---------------------------------------------------------------------------
    \7\ Bodie et al. (2002).
---------------------------------------------------------------------------
Conclusioin
    My testimony has focused on the role of administrative fees and 
charges in a PRA type approach, and also on payout considerations after 
retirement. I conclude that voluntary Personal Retirement Accounts can 
be designed so as to provide participants with some investment choice 
while still being relatively inexpensive; they can build in incentives 
for competition among fund managers, including disclosure regarding 
fees and charges; and they can sensibly require retirees to annuitize 
part of their retirement assets in their Personal Accounts, so that the 
combined benefit payments will keep them out of poverty.
    Thank you for your interest and I am happy to answer any questions 
you may have about my remarks.
References
    Bateman, H. & O.S. Mitchell. ``New Evidence on Pension Plan Design 
and Administrative Expenses.'' Journal of Pension Finance and 
Economics. 2004: Vol 3(1): 63-76.
    Bodie, Z., B. Hammond, and O.S. Mitchell, eds. Innovations in 
Financing Retirement. Philadelphia, PA: University of Pennsylvania 
Press, 2002.
    Brown, J.R., O.S. Mitchell, J.M. Poterba. ``The Role of Real 
Annuities and Indexed Bonds in an Individual Accounts Retirement 
Program.'' In Risk Aspects of Investment-Based Social Security Reform. 
Eds. J. Campbell and M. Feldstein. 2000: 321-360.
    Brown, J., O.S. Mitchell, J. Poterba, and M. Warshawsky. The Role 
of Annuity Markets in Financing Retirement. MIT Press, 2001.
    Commission to Strengthen Social Security (CSSS), Strengthening 
Social Security and Creating Personal Wealth for all Americans, Final 
Report, Washington, D.C., December 2001.
    Cogan, J.F. & O.S. Mitchell. ``Perspectives from the President's 
Commission on Social Security Reform.'' Journal of Economic 
Perspectives. 17(2). Spring 2003.
    Mitchell, O.S. ``Administrative Costs of Public and Private Pension 
Plans''. In Privatizing Social Security, Ed. M. Feldstein. NBER. 
Chicago: University of Chicago Press, 1998: 403-456.
    Mitchell, Olivia S. & David McCarthy. ``Annuities for an Ageing 
World''. In Developing an Annuities Market in Europe. Eds. E. Fornero & 
E. Luciano. Elgar, 2004: 19-68.
    NASI Uncharted Waters: Final Report. http://www.nasi.org/info-
url_nocat2718/info-url_nocat_show.htm?doc_id=212573
    Whitehouse, E. Testimony Before the Subcommittee on Social Security 
of the House Committee on Ways and Means, Washington, D.C. June 16, 
2005.

                                 

          Statement of Robin Sewell, Littleton, Massachusetts

    I worked to earn 40 quarters and had social security payments taken 
out of my paycheck for yeas. As I completed my undergraduate degree and 
master's degree in my late 20's and 30's, I started another career as 
an employee for the town of Littleton as a teacher over 17 years ago. 
Because I began my teaching career later in life, I will have a very 
difficult time putting in my years of service to even come close to 
reaching a moderate percentage from my municipal retirement.
    And, if my husband should pre-decease me, my widow benefits, even 
though he paid the maximum for social security each year, will also be 
severely impacted.
    Please consider that the current provisions penalize those 
``qualified'' workers who have a career change or start working for a 
municipality later in life.

                                 
