Social Security Reform: Issues for Disability and Dependent	 
Benefits (26-OCT-07, GAO-08-26).				 
                                                                 
Many recent Social Security reform proposals to improve program  
solvency include elements that would reduce benefits currently	 
scheduled for future recipients. To date, debate has focused	 
primarily on the potential impact on retirees, with less	 
attention to the effects on other Social Security recipients,	 
such as disabled workers and dependents. As these beneficiaries  
may have fewer alternative sources of income than traditional	 
retirees, there has been interest in considering various options 
to protect the benefits of disabled workers and certain 	 
dependents. This report examines (1) how certain elements of	 
Social Security reform proposals could affect disability and	 
dependent benefits, (2) options for protecting these benefits and
how they might affect disabled workers and dependents, and (3)	 
how protecting benefits could affect the Social Security program.
To conduct this study, GAO used a microsimulation model to	 
simulate benefits under various reform scenarios. GAO also	 
interviewed experts and reviewed various reform plans, current	 
literature, and GAO's past work.				 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-08-26						        
    ACCNO:   A77648						        
  TITLE:     Social Security Reform: Issues for Disability and	      
Dependent Benefits						 
     DATE:   10/26/2007 
  SUBJECT:   Aid for the disabled				 
	     Beneficiaries					 
	     Cost analysis					 
	     Cost of living					 
	     Dependents 					 
	     Disability benefits				 
	     Disability insurance				 
	     Federal social security programs			 
	     Persons with disabilities				 
	     Program evaluation 				 
	     Retirement 					 
	     Retirement benefits				 
	     Social security beneficiaries			 
	     Social security benefits				 
	     Trust funds					 
	     GAO High Risk Series				 
	     Old Age Survivors and Disability			 
	     Insurance Program					 
                                                                 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-08-26

   

     * [1]Results in Brief
     * [2]Background

          * [3]History and Development of DI and Dependents' Benefits
          * [4]Development of a Single Benefit Formula
          * [5]Benefits under Current Law
          * [6]Social Security's Financing

     * [7]Certain Benefit-Reducing Reform Elements Could Have a Substa

          * [8]Benefit-Reducing Reform Elements Take Different Forms
          * [9]Most of the Social Security Reform Elements We Considered Wo
          * [10]Reform Elements Reduce Median Lifetime Benefits for Disabled
          * [11]Disabled and Dependent Beneficiaries Are Not Uniformly Affec

     * [12]Options Protecting the Benefits of Disabled Workers and Depe

          * [13]Many Options Exist for Protecting Social Security Benefits,
          * [14]Some Benefit Protections Could Restore Benefits to Levels ne
          * [15]Protections for Dependent Benefits Could Be Targeted to Cert

     * [16]Some Protection Options May Create New Costs and Unintended

          * [17]Protecting the Benefits of Disabled Workers and Dependents W
          * [18]Certain Protections May Create Incentives for Individuals to
          * [19]Certain Protections May Affect the Work Decisions of Benefic

     * [20]Conclusions
     * [21]Matter for Congressional Consideration
     * [22]Agency Comments
     * [23]Microsimulation Model

          * [24]Description
          * [25]Evaluating Reform Proposals
          * [26]Assumptions and Limitations

               * [27]2007 Social Security Trustees' Assumptions

          * [28]Specification of Disabled Workers and Dependents

     * [29]Description of Social Security Reform Elements

          * [30]Longevity Indexing
          * [31]Price Indexing
          * [32]Progressive Price Indexing
          * [33]Increase the Number of Computation Years Used in the Initial
          * [34]Reduction in the Cost-of-Living Adjustment

     * [35]Description of Protection Options
     * [36]Options Modeled to Protect Benefits of Disabled Workers

          * [37]Full Exemption
          * [38]Partial Exemptions
          * [39]Partial Exemption Type I--Kolbe-Stenholm
          * [40]Partial Exemption Type II--Graham

     * [41]Data Reliability
     * [42]Calculating Old-Age Benefits
     * [43]Calculating Disability Benefits
     * [44]Calculating Benefits for Dependents of Retired and Disabled
     * [45]Calculating Survivors' Benefits
     * [46]GAO's Mission
     * [47]Obtaining Copies of GAO Reports and Testimony

          * [48]Order by Mail or Phone

     * [49]To Report Fraud, Waste, and Abuse in Federal Programs
     * [50]Congressional Relations
     * [51]Public Affairs

Report to Congressional Requesters

United States Government Accountability Office

GAO

October 2007

SOCIAL SECURITY REFORM

Issues for Disability and Dependent Benefits

GAO-08-26

Contents

Letter 1

Results in Brief 3
Background 5
Certain Benefit-Reducing Reform Elements Could Have a Substantial Impact
on Disabled Workers and Dependents 13
Options Protecting the Benefits of Disabled Workers and Dependents Could
Mitigate the Effects of Benefit-Reducing Reform Elements 23
Some Protection Options May Create New Costs and Unintended Incentives for
the Social Security Program 35
Conclusions 39
Matter for Congressional Consideration 40
Agency Comments 40
Appendix I Methodology 42
Microsimulation Model 42
Description of Social Security Reform Elements 45
Description of Protection Options 48
Options Modeled to Protect Benefits of Disabled Workers 50
Data Reliability 51
Appendix II Calculating OASDI Benefits 53
Calculating Old-Age Benefits 53
Calculating Disability Benefits 56
Calculating Benefits for Dependents of Retired and Disabled Workers 58
Calculating Survivors' Benefits 59
Appendix III Comments from the Social Security Administration 61
Related GAO Products 64

Tables

Table 1: Reform Elements, Their Solvency Impacts, and the Percentage of
Disabled Workers and Dependents with Reduced Total Lifetime Benefit Levels
for the 1985 Cohort 16
Table 2: Options for Protecting Disabled Worker and Dependent Benefits
from Social Security Reform 24
Table 3: Child Beneficiaries 34
Table 4: Total Lifetime and Average Monthly Benefits of Two Simulated
Individuals Who Began Receiving Benefits at Age 62, Exemptions Apply to
Disabled Workers (1985 Cohort) 38
Table 5: Options for Protecting Disabled Worker and Dependent Benefits
from Social Security Reform 48
Table 6: Dependent Benefits 59

Figures

Figure 1: Social Security Beneficiaries by Beneficiary Status, 2007 6
Figure 2: Disabled Worker and Dependent Benefits Timeline 8
Figure 3: Evolution of the OASDI Program 9
Figure 4: Projected Median Lifetime Benefits for Disabled Workers (1985
Cohort) 18
Figure 5: Projected Median Lifetime Benefits for Those Who Have Ever
Received Dependent Benefits (1985 Cohort) 19
Figure 6: Percentage Range of Projected Reductions in Total Lifetime
Benefits for Disabled Workers (1985 Cohort) 21
Figure 7: Percentage Range of Projected Reductions in Total Lifetime
Benefits for Individuals Who Have Ever Received Dependent Benefits (1985
Cohort) 22
Figure 8: Projected Median Lifetime Benefits under a COLA Reduction of 1
Percentage Point (1985 Cohort) 28
Figure 9: Projected Median Lifetime Benefits under Price Indexing (1985
Cohort) 30
Figure 10: Projected Median Lifetime Benefits under Longevity Indexing
(1985 Cohort) 31
Figure 11: Projected Median Lifetime Benefits under Progressive Price
Indexing (1985 Cohort) 32
Figure 12: Simulated Total Lifetime Benefits for Two Similar Individuals
Who Began Receiving Benefits at Age 62, Exemptions Apply to Disabled
Workers (1985 Cohort) 38
Figure 13: Calculating Benefits for Retirees 54
Figure 14: Calculating Benefits for Disabled Workers 57

Abbreviations: 

AIME: Average Indexed Monthly Earnings: 
COLA: cost-of-living adjustment: 
CPI: Consumer Price Index: 
CPI-U: consumer price index for urban consumers: 
CPI-W: consumer price index for urban wage earners and clerical 
workers:  
CSSS: Commission to Strengthen Social Security: 
DAC: disabled adult children: 
DI: Disability Insurance: 
FRA: full retirement age: 
GEMINI: Genuine Microsimulation of Social Security Accounts: 
MINT3: Modeling Income in the Near Term: 
OASDI: Old Age, Survivors, and Disability Insurance: 
OCACT: Office of the Chief Actuary: 
PENSIM: Pension Simulator: 
PIA: primary insurance amount: 
PSG: Policy Simulation Group: 
SI: Survivors'Insurance: 
SSAB: Social Security Advisory Board: 
SSASIM: Social Security and Accounts Simulator: 

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office
Washington, DC 20548

October 26, 2007

The Honorable Charles B. Rangel: 
Chairman: 
The Honorable Jim McCrery: 
Ranking Member: 
Committee on Ways and Means: 
House of Representatives: 

The Honorable Michael R. McNulty: 
Chairman: 
The Honorable Sam Johnson: 
Ranking Member: 
Subcommittee on Social Security: 
Committee on Ways and Means: 
House of Representatives: 

The Honorable Sander M. Levin: 
House of Representatives: 

Social Security forms the foundation of retirement income, providing old
age benefits to millions of Americans. However, Social Security is more
than a retirement program; it provides benefits to survivors and other
dependents as well as to disabled workers. For all these recipients,
benefits are determined through a common formula. Therefore, the prospect
of altering the benefit formula to address Social Security's financial
shortfall would affect all beneficiaries, not just retired workers.
Moreover, changes to the benefit formula could have a large impact on
those beneficiaries who are least able to compensate for any reduction in
benefits. For example, disabled workers may have fewer alternative sources
of income--especially earnings-related income--than do retired workers, to
offset any planned reduction in benefits. In addition, certain dependents,
such as older widows, may rely more heavily on Social Security benefits
and may not have the means to offset reductions caused by reform. Hence,
benefit reductions related to reform could potentially affect the welfare
of disabled workers and dependent beneficiaries more substantially than
those of retired workers. Social Security forms the foundation of
retirement income, providing old age benefits to millions of Americans.
However, Social Security is more than a retirement program; it provides
benefits to survivors and other dependents as well as to disabled workers.
For all these recipients, benefits are determined through a common
formula. Therefore, the prospect of altering the benefit formula to
address Social Security's financial shortfall would affect all
beneficiaries, not just retired workers. Moreover, changes to the benefit
formula could have a large impact on those beneficiaries who are least
able to compensate for any reduction in benefits. For example, disabled
workers may have fewer alternative sources of income--especially
earnings-related income--than do retired workers, to offset any planned
reduction in benefits. In addition, certain dependents, such as older
widows, may rely more heavily on Social Security benefits and may not have
the means to offset reductions caused by reform. Hence, benefit reductions
related to reform could potentially affect the welfare of disabled workers
and dependent beneficiaries more substantially than those of retired
workers.

Nonetheless, the debate over Social Security reform has focused primarily
on the effects proposed reforms would have on retirees--with little
discussion of how they might affect disabled workers and dependents. Given
interest in protecting these beneficiaries, this report discusses
Nonetheless, the debate over Social Security reform has focused primarily
on the effects proposed reforms would have on retirees--with little
discussion of how they might affect disabled workers and dependents. Given
interest in protecting these beneficiaries, this report discusses (1) how
certain elements of Social Security reform plans would affect disability
and dependent benefits, (2) some options for protecting these benefits and
how these options might affect disabled workers and dependents, and (3)
how protecting these benefits could affect the Social Security program
itself.

To determine how certain elements of reform would affect disability and
dependent benefits,1 we first selected reforms by reviewing a number of
reform plans, current literature, and our past work to identify reform
elements that would have an impact on benefits. We then estimated the
percentage of disabled workers and dependents whose benefits would be
affected by certain reform elements by simulating the outcome for a sample
of workers born in 1985, using a microsimulation model.2 We also used data
from this model to analyze the difference in both median and total
lifetime benefits between currently scheduled benefits and those under
each reform element.3 Rather than consider the distribution of benefits
across recipients, we show the impact of reform and protections on median
benefits. This approach provides a snapshot of how the reforms would
affect individuals in each target group (disabled workers and dependents).
In addition, when we consider total lifetime benefits, we are able to
examine the extent to which individuals are affected. While scheduled
benefits are not attainable under current funding levels, they
nevertheless provide us with a point of comparison in examining the impact
of various reform elements. To identify options for protecting disabled
worker and dependent benefits, we reviewed the current literature and our
past work and interviewed relevant experts. To determine how these
protections would affect both disability and dependent benefits, we
analyzed the change in median lifetime benefits given various reform
elements and different protections, using the microsimulation model, as
well as some qualitative analysis. Finally, to determine the impact that
protecting benefits could have on the Social Security program itself, we
interviewed relevant experts, reviewed the literature, and conducted a
qualitative analysis of the issues involved. Throughout our analysis, we
assume that the Disability Insurance (DI) program maintains its current
operational structure in terms of disability determination, although GAO
has recommended certain measures that Social Security should take to
modernize the program and its administration. For more details on our
approach to this study and on the microsimulation model, please see
appendix I.

1Our approach to the analysis is based on the premise that, under benefit
reducing reforms, disabled workers and certain dependents would likely
have limited options for alternative sources of income to make up for the
loss in benefits. In particular, comparisons of the relative welfare of
disabled workers, dependents, and retirees are beyond the scope of this
job.

2We used the Genuine Microsimulation of Social Security (GEMINI)
microsimulation model under a license from the Policy Simulation Group, a
private contractor. GEMINI estimates individual effects of policy
scenarios for a representative sample of future beneficiaries. GEMINI can
simulate different reform features for their effects on the level and
distribution of benefits. See appendix I for more detail on the modeling
analysis, including a discussion of our assessment of the data reliability
of the model.

3We analyzed the impact of each reform element individually. We did not
simulate a scenario in which reform elements were combined and implemented
simultaneously. It is important to note that the individual reform
elements can interact with one another and that the impact of one reform
element, taken on its own, may change when combined with other reform
elements in a proposal. Moreover, each reform element has pluses and
minuses. As a result, Social Security reform proposals should be evaluated
as a package of reform options designed to meet certain stated objectives.

Throughout this report, the term "dependents" refers to those
beneficiaries who receive some or all of their Social Security benefits
based on a family member's earnings record. Our dependent category
includes survivors--such as widow(er)s and surviving children--as well as
spouses and children of both retirees and individuals receiving disability
benefits. Specifically, dependents need not demonstrate financial
dependence, but rather familial relationships such as those listed above.
Also, throughout the report, "disabled worker benefits" refers to the type
of benefit that an individual with disabilities who qualifies for the DI
program would receive based on his or her own earnings record. Further,
those individuals who receive disabled worker benefits are referred to as
"disabled workers." While we understand the sensitivity associated with
this terminology, our use of this term is consistent with the Social
Security Administration's terminology and aims to precisely represent such
beneficiaries and their benefits.

We conducted our work between October 2006 and October 2007 in accordance
with generally accepted government auditing standards.

Results in Brief

Most of the elements for reforming Social Security that we considered
would reduce future benefits from currently scheduled levels for Social
Security recipients, including the great majority of disabled workers and
dependents. These reform elements include longevity indexing (changing
benefits to reflect increased life expectancies), price indexing
(adjusting benefits so they grow at the rate of inflation rather than
wages), progressive price indexing (a graduated form of price indexing),
and changes to the cost-of-living adjustment. According to our
simulations, most such changes would reduce benefits for at least
three-quarters of disabled workers and dependent beneficiaries in the 1985
cohort. For example, progressive price indexing would affect more than 75
percent of these beneficiaries, while other reform elements we analyzed
would affect virtually all beneficiaries. In addition, most reform
elements would reduce the median lifetime benefits of disabled workers and
dependents to between 70 and 93 percent of currently scheduled levels,
with price indexing resulting in the largest reduction. These reform
elements would not affect individuals uniformly. For example, longevity
indexing would reduce the lifetime benefits of 86 percent of disabled
workers by between 10 to 25 percent; the other 14 percent would generally
face reductions of 10 percent or less.

There are many options for protecting both disabled worker and dependent
benefits. Some of these options can be targeted to particular groups of
beneficiaries. The options include, among others, a partial exemption--by
which the beneficiary receives currently scheduled benefits until
retirement age--and a full exemption--by which the beneficiary receives
benefits at the currently scheduled level throughout the remainder of his
or her life. Through our simulation of reform elements and various
protections, for example, we found that a one percentage point reduction
in the cost-of-living adjustment would reduce the median lifetime benefits
of disabled workers to about 89 percent of their currently scheduled
level, but a partial exemption could restore them to about 96 percent.
Options for protecting the benefits of disabled workers could also be
structured to mitigate benefit reductions that result from changes to the
initial benefit calculation. For example, disabled workers who receive
benefits for a prolonged period of time could be given a "super COLA" that
would allow their benefits to grow more rapidly. In terms of dependent
benefits, protections could cover a single group, such as widows, or
multiple groups. For example, increasing the maximum family benefits could
protect the benefits of child survivors, widowed parents, children and
spouses of disabled workers, and disabled adult children.

In general, although it may be desirable to protect the benefits of
certain disabled workers and dependents, such protections would come at a
cost to the program and could create unintended incentives that would
negatively affect Social Security's finances. While the reforms we
considered were aimed at improving the solvency of Social Security,
protecting the benefits of these populations would lessen the effect of
each of these reforms. In addition, certain protections may create
incentives for individuals who might not otherwise apply for Disability
Insurance to do so. For example, those nearing retirement may apply for
the Disability Insurance program if disability insurance benefits remained
stable while retirement benefits fell.

The Department of the Treasury provided technical comments. The Social
Security Administration (SSA) provided general and technical comments. We
incorporated the comments throughout our report as appropriate.

Background

Social Security is one the largest federal programs in the United States,
providing about $546 billion in benefits in 2006 to over 49 million
beneficiaries. Although the majority of Social Security benefits are paid
to retirees, Social Security does much more than provide retirement
income. Social Security Disability Insurance (DI) pays monthly cash
benefits to nearly 7 million workers who, due to a severe long-term
disability, can no longer remain in the workforce. Additionally, Social
Security provides benefits to over 11 million dependents,4 including
payments to widows and widowers as well as surviving parents and children
under Survivors' Insurance (SI), plus benefits to dependent spouses and
children of retired and disabled workers paid from the Old Age Insurance
(or Old Age) and DI trust funds. Social Security benefits often represent
a significant source of income for their recipients, providing an average
of $1,051 a month (as of July 2007) to retired workers, $995 a month to
widows and widowers, and $979 a month to disabled workers. Although
disabled workers and dependents receive slightly lower average monthly
benefits than retired workers, benefits could be particularly important to
these individuals. These beneficiaries may face considerable hardships;
for example, a disabling condition may make work and other activities of
daily living more difficult. As a result, these beneficiaries may have
financial difficulties planning and preparing for death or disability in
the way one might plan for retirement. Social Security was never intended
to provide an adequate income by itself, but instead serves as an income
base on which to build. In fact, the Social Security program balances the
goals of income adequacy with individual equity, i.e., that lower income
beneficiaries should receive higher benefits relative to wages than higher
income beneficiaries (adequacy), and beneficiaries with higher lifetime
income receive higher benefits in accordance with their income/lifetime
contributions (equity).

4For the purpose of this report, dependents will encompass benefits to
survivors as well as all dependent spouses and children.

Figure 1: Social Security Beneficiaries by Beneficiary Status, 2007

Notes: Some Social Security beneficiaries are entitled to more than one
type of benefit. If both benefits are financed from the same trust fund,
the beneficiary is usually counted only once in the statistics, as a
retired-worker or a disabled-worker beneficiary, and the benefit amount
recorded is the larger amount associated with the auxiliary benefit. If
the benefits are paid from different trust funds, the beneficiary is
counted twice and the respective benefit amounts are recorded for each
type of benefit. Accessed from
[52]http://www.ssa.gov/policy/docs/quickfacts/stat_snapshot/ on August 29,
2007.

History and Development of DI and Dependents' Benefits

Although Social Security had originally been envisioned to include
disability and survivors' insurance, the 1935 Social Security Act created
only a retirement program.5 Over the next 40 years, the program expanded
both the size and type of its benefits, introducing benefits for
dependents and disabled workers (fig. 2). The first new type of benefits
went to dependents, as the 1939 amendments offered payments to elderly
dependent wives and widows, as well as dependent children. (Some husbands
and widowers were allowed to receive these same benefits after 19506).
Creating these benefits was not only seen as socially desirable, but also
offered additional protections for workers and their families from risk
and spent down surpluses created by the system. Disability Insurance,
which had been recommended by the 1938 and 1948 advisory councils, was
established in 1956 to provide cash benefits to permanently disabled
workers over the age of 50. The DI program was later expanded to include
disabled workers under the age of 50 as well. In 1961, widows benefits
increased from 75 to 85 percent of their deceased spouse's benefits, and
then to 100 percent in 1972. In addition, eligibility was extended to
divorced spouses as well as to the spouses and children of disabled
workers. Furthermore, benefit levels for retirees, dependents, and
disabled worker beneficiaries grew during this time period. However,
facing solvency crises, legislative efforts to control the size of the
Social Security program were made in the mid 1970s and early 1980s.

5The Social Security Act of 1935 also included aid to states for programs
which are no longer considered to be part of Social Security.

6In order to receive these benefits, husbands and widowers needed to be
currently and fully insured at the time of their wives' retirement or
death. In addition, the husband (or widower) had to prove he was
financially dependent on his wife. In 1983, virtually all gender-based
distinctions were eliminated.

Figure 2: Disabled Worker and Dependent Benefits Timeline

In order to maintain trust fund solvency, major changes were enacted to
reduce the growth of Social Security benefit levels from the mid1970s to
the early 1980s. Additionally, a number of legislative changes to the DI
and dependents' programs eliminated or reduced certain benefits and
tightened the eligibility standards for receiving other benefits. However,
despite ongoing fiscal concerns, eligibility for a few dependents' and
disability benefits has been expanded since 1975, suggesting an interest
in protecting some vulnerable populations who may rely on Social Security
for a significant portion of their monthly income. Although recent reform
proposals have focused on elements intended to improve solvency, there
continues to be some interest in protecting some or all DI and dependents'
benefits from potential benefit reductions. Figure 3 shows how Old Age,
Survivors, and Disability Insurance (OASDI) has grown financially and in
terms of beneficiaries over time.

Figure 3: Evolution of the OASDI Program

Development of a Single Benefit Formula

Although the 1935 act did not provide for disability and dependents'
benefits, those benefits were later built upon the existing Social
Security structure, and today all benefits continue to be calculated from
a common formula. Dependents' benefit levels were set as fractions of the
benefits owed to the person upon whom beneficiaries depended. For example,
under the 1939 legislation, a widow would receive 75 percent of her
deceased husband's benefits, and dependent children or spouses would
receive 50 percent of the retired worker's benefits. When Congress created
the DI program in 1956, it provided a lower retirement age (50) for those
who were permanently and totally disabled. The same benefit formula used
in computing OASI benefits was adopted for disability benefits because the
original DI program treated disabled workers as being forced into
premature retirement. In 1960, when Congress expanded the DI program by
eliminating the requirement that disabled workers had to be 50 years old,
the same benefit formula applied. Because benefit types shared a common
formula, automatic indexing provisions implemented in 1972 and 1977
applied across the board.

The OASDI programs are tightly linked in other ways as well. These
programs are financed through a common mechanism--payroll taxes; receipts
from the payroll tax are deposited into the OASI and DI trust funds which,
like the two programs, are separate but often combined in discussion and
analysis of Social Security's solvency and sustainability.7 Furthermore,
beneficiaries can receive multiple types of benefits over their lifetimes,
moving into, out of, and among Social Security programs at different life
stages. When disabled workers reach the full retirement age (FRA), for
example, they begin to receive retirement benefits from the Old Age
program, in place of DI benefits; the common benefit formula keeps such
individuals' benefit levels stable.8 In another case, a recent widow(er)
might have her (or his) retirement or spousal benefits replaced with
survivors' benefits based on the relative earnings of the deceased spouse.
Because parents, children, or spouses may be eligible for dependents'
benefits through the Old Age, Survivors, and Disability Insurance
programs, a person can collect several types of Social Security benefits
over a lifetime, although generally not simultaneously.9 The many linked
pieces of Social Security could make developing a single, comprehensive
reform package challenging because such a package would need to take into
account all of these pieces.

7Payroll tax rates are specified separately for each program and receipts
deposited into two separate accounts in the United States Treasury.
However, over the years, there have been tax rate reallocations and loans
between the two trust funds.

8When a person is converted from the DI program to the Old Age programs,
there is no recalculation of his benefits, which continue to receive
annual COLAs; however, the benefits will be paid from the OASI trust fund
instead of the DI trust fund.

9Some beneficiaries may be dually entitled and receive benefits based on
their own record and on that of their spouse.

Benefits under Current Law

Under current law, Old Age benefits are generally calculated through a
four-step process in which a progressive yet earnings-based formula is
applied to an earnings history, and then updated annually through a
cost-of-living adjustment (COLA).10 For those who receive retirement
benefits, this earnings history is generally based on the 40 years in
which credited earnings were highest, with the 5 lowest-earning years
dropped out (leaving the highest 35 years of indexed earnings to be
included in the initial benefit calculation).11 Dependents' benefit levels
are determined as a given percentage of Old Age benefit levels. Eligible
children and spouses can receive up to 50 percent of a worker's benefit;
widow(er)s can be given up to 100 percent; and surviving parents or
children can collect up to 75 percent, subject to a family maximum.

DI benefits are calculated similarly to Old Age benefits, but are
generally based upon a shortened work history. (For more detail on how
benefits are calculated, refer to app. II.) To be eligible for benefits,
individuals must have a specified number of recent work credits under
Social Security when they first become disabled. Individuals must also
demonstrate the inability to engage in substantial gainful activity by
reason of a physical or mental impairment that has lasted or is expected
to last for twelve continuous months or to result in death. If not
eligible on medical grounds, SSA must also consider age, education and
past work history. In particular, medical eligibility criteria for DI are
less stringent for applicants over the age of 55.

Based on prior work, GAO has designated modernizing federal disability
programs (including the DI program) as a high risk area because of
challenges that continue today.12 For example, GAO found that federal
disability programs remain grounded in outmoded concepts that equate
medical conditions with work incapacity. While SSA has taken some actions
in response to prior GAO recommendations, GAO continues to believe that
SSA should continue to take a lead role in examining the fundamental
causes of program problems and seek the regulatory and legislative
solutions needed to modernize its programs so that they are aligned with
the current state of science, medicine, technology, and labor market
conditions. Moreover, SSA should continue to develop and implement
strategies to better manage the programs' accuracy, timeliness, and
consistency of decision making.

1042 U.S.C. S 415.

11The total number of years to be counted in the work history is the
number of elapsed years from the latter of 1950 or the year in which the
worker attains age 21 and before the worker becomes disabled, dies, or
attains age 62 (thus the 40 years in the text above)--whichever comes
first--excluding any years in which the worker is in a period of
disability.

12 See GAO, High Risk Series, An Update, [53]GAO-07-310 (Washington, D.C.:
January 2007).

Social Security's Financing

Social Security is currently financed primarily on a pay-as-you-go basis,
in which payroll tax contributions of current workers are used primarily
to pay for current benefits. Since the mid1980s, the Social Security
program has collected more in taxes than it has paid out in benefits.
However, because of the retirement of the baby boomers coupled with
increases in life expectancy, and decreases in the fertility rate, this
situation will soon reverse itself. According to the Social Security
Administration's 2007 intermediate assumptions,13 annual cash surpluses
are predicted to turn into ever-growing cash deficits beginning in 2017.
Absent changes to the program, these deficits are projected to deplete the
Social Security DI trust fund in 2026 and the OASI trust fund in 2042,
leaving the combined system unable to pay full benefits by 2041.
Reductions in benefits, increases in revenues, or a combination of both
will likely be needed to restore long-term solvency. A number of proposals
have been made to restore fiscal solvency to the program, and many include
revenue enhancements, benefit reductions, or structural changes such as
the introduction of individual accounts as a part of Social Security.
Because many reforms to the benefit side of the equation would reduce
benefits through changes in the benefit formula, they could affect DI and
dependents' benefits as well as Old Age benefits. Unless accompanied by
offsets or protections, these reforms might reduce the income of disabled
workers and dependents. This situation could be challenging for these
beneficiaries as they may have relatively low incomes or higher health
care costs and rely heavily on Social Security income. Many disabled
workers and dependents may also have trouble taking on additional work and
accumulating more savings and, thus, have difficulty preparing for Social
Security benefit reductions.

13Because the future is uncertain, the Trustees use three alternative sets
of assumptions to show a range of possible outcomes. The intermediate
assumptions represent the Social Security Administration's best estimate
of the trust funds' future financial outlook. The Trustees also present
estimates using low cost and high cost sets of assumptions. In addition,
the Trustees Report describes a range of possible outcomes using
stochastic modeling techniques.

Certain Benefit-Reducing Reform Elements Could Have a Substantial Impact on
Disabled Workers and Dependents

Many reform elements could have a substantial impact on the benefits of
Social Security recipients, including those of disabled workers and
dependents. We considered six such elements that have been included in
reform proposals to improve trust fund solvency. These reform elements
take a variety of forms and would change either the initial benefit
calculation or the growth of individual benefits over time. Our
projections indicated that most of these elements would reduce benefits
from currently scheduled levels14 for the majority of both disabled
workers and dependents. That is, most would reduce median lifetime
benefits for these beneficiary types--some more substantially than others.
Many of these beneficiaries would also experience a reduction in total
lifetime benefits; the extent of which would depend on the reform element
and individual.

Benefit-Reducing Reform Elements Take Different Forms

We considered six different reform elements that could help control costs
and improve Social Security solvency by reducing benefits.15 Five would
change how initial benefits are calculated, and one would limit the growth
of an individual's benefits over time.

We considered several ways to improve solvency:16

           o Longevity indexing would lower the amount of the initial benefit
           in order to reflect projected increases in life expectancy. Such
           indexing would maintain relatively comparable levels of lifetime
           benefits across birth years by proportionally reducing the
           replacement factors in the initial benefit formula.
           o Price indexing would maintain purchasing power while slowing the
           growth of initial benefits. This would be accomplished by indexing
           initial benefits to the growth in prices rather than wages, as
           wages tend to increase faster than prices.
           o Progressive price indexing, a form of price indexing, would
           control costs while protecting the benefits of those beneficiaries
           at the lowest earnings levels (in terms of career average
           earnings). It would continue to index initial benefit levels to
           wages for those below a certain earnings threshold and employ a
           graduated combination of price indexing and wage indexing for
           those above this threshold.
           o Increasing the number of years used in the benefit calculation
           would also control program costs. For example, initial benefits
           could be based on the highest 40, rather than 35, years of indexed
           earnings. This could be done either by eliminating the 5 years
           normally excluded from the calculation or by increasing the total
           number of years factored in from 40 to 45 years. In either of
           these cases, the initial Old Age benefit would be calculated using
           the highest 40 years of indexed earnings.17 (For more information
           on these reform elements and how we incorporated them into our
           microsimulation model, see app. I.)
           o Raising the age at which people are eligible for full retirement
           benefits could change the amount and/or the timing of initial
           benefits. Increasing the full retirement age would improve
           solvency by generally increasing the number of years worked,
           reducing the number of years benefits are received and increasing
           revenue to the system through payroll taxes in the additional
           years worked. Further, those who retire early would have their
           benefits actuarially reduced.
           o Though it would not generally affect initial benefit amounts, a
           change to Social Security's cost-of-living adjustment (COLA) could
           also control costs and improve solvency by limiting the growth of
           an individual's benefits over time. The COLA adjusts benefits to
           account for inflation by indexing benefits to price growth
           annually, using the Consumer Price Index (CPI).18 Setting the COLA
           below the CPI would limit the nominal growth of an individual's
           benefits over time,19 and as such those who receive benefits for a
           prolonged period of time would see the largest reductions.

14While currently scheduled benefits are not attainable under current
funding levels, they nonetheless provide a point of comparison in
examining the impact of various reform elements.

15These reform elements have been proposed as part of recent larger reform
proposals.

16See appendix II for a full explanation of how benefits are currently
calculated.

17Changing the number of drop-out years would have less of an effect (or
no effect) on disabled workers because their drop-out years are calculated
differently. For more on how initial benefits are calculated for disabled
workers, see appendix II.

18Specifically, Social Security's COLAs are based on the consumer price
index for urban wage earners and clerical workers (CPI-W), as opposed to
the CPI series for all urban consumers (CPI-U).

19We identified two options for modifying the COLA. The first option,
which we analyzed in this report, would reduce the COLA by 1 percentage
point to improve solvency. The second option, which we did not analyze,
would reduce the COLA by 0.2 to 0.4 percentage points in response to
methodological concerns that the CPI overstates the true rate of
inflation.

Most of the Social Security Reform Elements We Considered Would Reduce Benefits
for Virtually All Beneficiaries

According to our projections for the 1985 cohort, four of the five reform
elements that we analyzed would reduce total lifetime benefits for more
than three-quarters of disabled workers and dependents, relative to
currently scheduled benefits.20 Table 1 shows the proportions of disabled
workers and dependents affected by each of the reform elements. For three
of the elements--reducing the COLA by one percentage point, price indexing
and progressive price indexing--the percentage of disabled workers
affected is very similar to the percentage of dependents affected.
Moreover, for these three reform elements, more than 99 percent, or
virtually all, disabled workers and dependents would see their benefits
reduced. In contrast, progressive price indexing differs from other reform
elements in its impact: fewer beneficiaries are affected, and the
percentage of disabled workers affected varies from that of dependents.
While an estimated 87 percent of dependents would experience a reduction
in lifetime benefits under progressive price indexing, an estimated 77
percent of disabled workers would do so.

While the COLA reduction, longevity indexing and price indexing are all
designed in such a way that they affect virtually all beneficiaries, the
COLA, which has a greater impact on solvency than longevity indexing,21
affects relatively fewer disabled workers and dependents. This is because
the COLA reduction would first affect benefits one year after the initial
benefit payment was made, whereas both longevity indexing and price
indexing affect the initial benefit amount. Our simulations indicated that
1.11 percent of disabled workers died within the first year of receiving
benefits, while only 0.35 percent of dependents did so. Most such
beneficiaries would not have received a COLA.

20We did not analyze the simulated change in the FRA for both disabled
workers and dependents because of modeling constraints nor did we analyze
the increase in the number of computation years for disabled workers.

21A 1 percentage point decrease in the COLA may have a greater effect on
solvency than longevity indexing as we've modeled it. In part because
while the effects of both reforms are compounded over time, the COLA would
affect all beneficiaries once it is implemented and continue over a longer
horizon--until death. In contrast, longevity indexing would not affect
those who have already retired, nor would it have a great impact on those
who retire close to the implementation date. Its effects would be greater
on the benefits of those individuals who retire further out from the
implementation date.

Table 1: Reform Elements, Their Solvency Impacts, and the Percentage of
Disabled Workers and Dependents with Reduced Total Lifetime Benefit Levels
for the 1985 Cohort

                                                                Percentage of 
                                                              dependents with 
                              Solvency Percentage of disabled   reduced total 
                               impactb   workers with reduced        lifetime 
                            (actuarial         total lifetime  benefits (1985 
Elementa                  scoring)c benefits (1985 Cohort)         Cohort) 
COLA reduction -               1.49                  99.15           99.87 
1percentage pointd                                                         
Increase the number of         0.46                   n/ae           99.99 
computation years-- from                                                   
35 to 40d                                                                  
Longevity                      1.17                  99.89           99.98 
Indexing--reducing                                                         
formula factors by 0.5%f                                                   
Price indexingd                2.38                  99.95           99.97 
Progressive price              1.43                  76.78           86.90 
indexingd,g                                                                

Source: GAO analysis of GEMINI data and SSA OCACT.

aBecause of modeling constraints, we were unable to analyze the effects of
a change in the FRA. However, Social Security actuaries have estimated the
solvency impact of increasing the FRA to age 68. This reform would improve
actuarial balance by 0.62 percent of taxable payroll.

bSolvency impacts come from SSA's Office of the Chief Actuary (OCACT).
Taken individually, each of the reform elements would improve Social
Security solvency. For more on how the reform elements were scored for
solvency, see appendix I.

cActuarial balance as a percentage of taxable payroll.

dOCACT Score based on 2005 Trustees Report intermediate assumptions.

eWe were also unable to analyze the effects of increasing the number of
computation years for disabled workers.

fOCACT Score based on 2001 Trustees Report intermediate assumptions.

gIn the microsimulation model, approximately 80 percent of all
beneficiaries were affected by the progressive price indexing reform. Some
disabled workers also received benefits as dependents at some point in
their life, and if those workers are excluded then the percentage of
disabled workers affected would fall.

Note: Using the 1985 cohort, we also projected the percentage of workers
with reduced total lifetime benefits for Social Security beneficiaries who
were never disabled. 99.68% of Social Security beneficiaries who were
never disabled would see benefit reductions if the COLA were reduced by 1
percentage point, and 100% would see benefit reductions if the number of
computation years was increased from 35 to 40. Under longevity indexing,
price indexing, and progressive price indexing the percentage of
beneficiaries affected would be 99.99%, 99.99%, and 80.97%, respectively.

Reform Elements Reduce Median Lifetime Benefits for Disabled Workers and
Dependents to Varying Degrees

According to our simulations each of the reform elements we selected would
reduce median lifetime benefits for both disabled workers and dependents
relative to currently scheduled benefits (figs. 4 and 5).22 However, our
projections also indicated that these reductions would vary by reform
element. Price indexing would have the largest impact on disabled workers
and dependents, reducing median lifetime benefits by more than 25 percent.
Median lifetime benefits would fall from $473,960 to $343,350 for disabled
workers and from $351,910 to $244,745 for dependents. Progressive price
indexing, on the other hand, would create the smallest reduction in median
lifetime benefits, with median lifetime benefits falling by 7 percent for
disabled workers and 8 percent for dependents.23

22While this report focuses on the impact of certain reforms on disabled
worker and dependent benefits, the reforms would affect the benefits of
other Social Security recipients as well.

23Additionally, according to our simulations, disabled workers experienced
greater absolute reductions in median lifetime benefits than did
beneficiaries who were never disabled for each of the reform elements
considered. However, disabled workers also experienced a smaller
percentage change in median lifetime benefits than those who were never
disabled. Under current law, disabled workers would have higher lifetime
benefits (since they may receive benefits for a longer period of time);
therefore, the reduction in benefits represents a smaller share of current
law benefits.

Figure 4: Projected Median Lifetime Benefits for Disabled Workers (1985
Cohort)

Figure 5: Projected Median Lifetime Benefits for Those Who Have Ever
Received Dependent Benefits (1985 Cohort)

Additionally, increasing the full retirement age and increasing the number
of computation years would likely reduce median lifetime benefits for
dependents.24 Since dependent benefits are linked to those of the primary
worker, an increase in the full retirement age could shorten the period of
time over which they both receive benefits. Alternatively, some workers
may decide not to adjust their retirement plans in response to the
increase in the FRA. Those who maintain their original retirement plans,
retiring prior to the new FRA, will also receive reduced benefits relative
to current law. (See app. II for a discussion of how benefits are adjusted
for early retirement.) Thus, under both scenarios, total lifetime benefits
would be reduced, and so, too, would median lifetime benefits. A similar
outcome results from increasing the number of computation years by which
initial benefits are calculated. By increasing the number of computation
years, a worker's earnings history is expanded to include years of
possibly lower indexed earnings. As a result, total benefits for some
retired workers, and therefore, their dependents, would likely be reduced,
as would median lifetime benefits.

24Increasing the FRA would not affect disabled workers because the initial
benefits calculation for disabled workers does not involve the FRA.
However, it would change the age of conversion to retirement benefits.
Similarly, changing the number of drop-out years would have less of an
effect (or no effect) on disabled workers because drop-out years are
calculated differently for disabled workers. For more on how initial
benefits are calculated for disabled workers, see appendix II.

Disabled and Dependent Beneficiaries Are Not Uniformly Affected by Reform
Elements

Our projections suggest that, while lifetime benefits would be reduced for
virtually all disabled workers and dependents, such reductions would not
be uniform across individuals. Figures 6 and 7 compare beneficiaries'
total lifetime benefit reductions by each reform element, for disabled
workers and dependents, respectively. If the COLA were reduced by one
percentage point, our projections show that approximately 58 percent of
disabled workers experienced lifetime benefit reductions of 10 percent or
less,25 while about 42 percent of disabled workers experienced lifetime
benefits reduced by 10 to 25 percent. Almost no disabled workers would see
benefits fall by more than 25 percent.

25In particular, under the COLA reduction, less than 1 percent of disabled
workers had no change in lifetime benefits; about 2 percent of disabled
workers had a change of 1 percent or less; about 14 percent had a change
of more than 1 and up to 5 percent; and about 41 percent had lifetime
benefits change by more than 5 and up to 10 percent.

Figure 6: Percentage Range of Projected Reductions in Total Lifetime
Benefits for Disabled Workers (1985 Cohort)

Note: The above intervals include the upper endpoint. For example "Between
1 and 5 percent" includes 5, but not 1. No disabled workers had total
lifetime benefit reductions of more than 50 percent.

Figure 7: Percentage Range of Projected Reductions in Total Lifetime
Benefits for Individuals Who Have Ever Received Dependent Benefits (1985
Cohort)

Note: The above intervals include the upper endpoint. For example "Between
1 and 5 percent" includes 5, but not 1. For those who were ever classified
as a dependent, only a very few individuals had total lifetime benefit
reductions of more than 50 percent. Four individuals experienced a benefit
reduction of more than 50 percent under the COLA reduction, 21 individuals
under increasing the number of computation years, 13 under longevity
indexing, 34 under price indexing, and 5 under progressive price indexing.

Certain reform elements would create reductions in total lifetime benefits
for the vast majority of disabled workers and dependents. These reductions
may create new hardships for certain beneficiaries, such as disabled
workers, who may not be able to easily replace lost income. According to
our projections, price indexing would result in the greatest benefit
reductions for the largest percentage of beneficiaries, with decreases in
lifetime benefits of between 25 percent and 50 percent for almost 70
percent of disabled workers and about 90 percent of dependents. Both price
indexing and longevity indexing have a greater effect on initial benefit
amounts the longer the reform is in place. As such, people who leave the
workforce early may experience a smaller reduction in lifetime benefits
than those who leave at full retirement age.26 For example, as shown in
figures 6 and 7, longevity indexing could reduce lifetime benefits for
about 86 percent of disabled workers and about 96 percent of dependents by
10 to 25 percent.27

Progressive price indexing may have a more moderate effect on the benefits
of disabled workers and certain dependents because it is designed to
protect benefit levels for low earners and gradually apply benefit
reductions to beneficiaries with higher earnings. Because of shorter
earnings histories, some disabled workers would be in the low end of the
earnings distribution.28 Thus, under progressive price indexing, a greater
proportion of disabled workers would be likely to have benefits adjusted
by wage indexing. According to our projections, progressive price indexing
would reduce total lifetime benefits by 5 percent or less for 46 percent
of disabled workers and 35 percent of dependents.

Options Protecting the Benefits of Disabled Workers and Dependents Could
Mitigate the Effects of Benefit-Reducing Reform Elements

Various options are available to protect benefits in different ways,
including accelerating the growth of an individual's benefits, modifying
current constraints on benefit levels, and exempting certain populations
from reforms. Options can also target certain types of beneficiaries. We
analyzed some of these protections and found they could be structured to
mitigate the effects of benefit reductions for varying lengths of time. In
addition, we found that specific options to protect dependent benefits
could be targeted to certain vulnerable beneficiaries, such as widows and
dependent children.

26Many disabled workers leave the workforce in their fifties.

27The other 14 percent generally had reductions of 10 percent or less, and
very few disabled workers had reductions in lifetime benefits of more than
25 percent under longevity indexing.

28Many people have lower relative earnings in their first years of work.

Many Options Exist for Protecting Social Security Benefits, Including Full and
Partial Exemptions

We found a wide range of options exist for protecting disabled workers and
dependents from benefit-reducing reforms. Table 2 provides a summary of
the options.29 The protection options may be very specific in terms of
whom they protect and how, or broader in scope. For example, while two
protection options focus specifically on disabled adult children (DAC),
others, such as partial exemptions could apply to any vulnerable
population. In addition to each option having its own strengths and
weaknesses, the options could interact with each other and with the
various reform elements. When implementing a protection option, all of
these factors could influence its impact.

Table 2: Options for Protecting Disabled Worker and Dependent Benefits
from Social Security Reform

                           Given a   
                           specific  
                           reform,   
                             the     
                          protection 
                            option   
                            could    
                            affect   
                             the:    
                                                                      Conversion 
 Protection                                      Maximum                at the   
 option                    Initial   Growth of   family   Eligibility    full    
 available   Protection    benefit   individual  benefit      for     retirement 
 for         option         amount    benefits  available  benefits      age     
 Any         Full                                                                
 beneficiary exemption                                                           
 type        Partial                                                             
             exemption                                                           
             Minimum                                                             
             benefit                                                             
             Super COLA                                                          
             Age-indexed                                                         
             super COLA                                                          
 Children    Increase the                                                        
 and         percentage                                                          
 families of of the                                                              
 disabled    worker's                                                            
 workers     benefit that                                                        
             the                                                                 
             dependent                                                           
             family                                                              
             member                                                              
             receives as                                                         
             his/her                                                             
             benefit                                                             
             Increase the                                                        
             family                                                              
             maximum                                                             
             benefit                                                             
             level for DI                                                        
             Increase the                                                        
             percentage                                                          
             of the                                                              
             worker's                                                            
             benefit that                                                        
             a dependent                                                         
             child or a                                                          
             disabled                                                            
             adult child                                                         
             (DAC)                                                               
             receives as                                                         
             his/her                                                             
             benefit in                                                          
             combination                                                         
             with                                                                
             increasing                                                          
             the family                                                          
             maximum                                                             
             benefit                                                             
             Decouple DAC                                                        
             benefits                                                            
             from other                                                          
             family                                                              
             benefits                                                            
             Hold initial                                                        
             benefit                                                             
             amount                                                              
             harmless for                                                        
             family                                                              
             benefits                                                            
 Children    Expand                                                              
 and         eligibility                                                         
 families of rules for                                                           
 retired     divorced                                                            
 workers     spouses                                                             
             Increase the                                                        
             percentage                                                          
             of the                                                              
             worker's                                                            
             benefit that                                                        
             the                                                                 
             dependent                                                           
             family                                                              
             member                                                              
             receives as                                                         
             his/her                                                             
             benefit                                                             
             Increase the                                                        
             percentage                                                          
             of the                                                              
             worker's                                                            
             benefit that                                                        
             a dependent                                                         
             child or DAC                                                        
             receives as                                                         
             his/her                                                             
             benefit in                                                          
             combination                                                         
             with                                                                
             increasing                                                          
             the family                                                          
             maximum                                                             
             benefit                                                             
             Decouple DAC                                                        
             benefits                                                            
             from other                                                          
             family                                                              
             benefits                                                            
 Spouses     Increase the                                                        
             percentage                                                          
             of the                                                              
             worker's                                                            
             benefit that                                                        
             the spouse                                                          
             receives as                                                         
             his/her                                                             
             benefit                                                             
             Implement a                                                         
             child/family                                                        
             care credita                                                        
 Survivors   Hold initial                                                        
             benefit                                                             
             amount                                                              
             harmless for                                                        
             family                                                              
             benefits                                                            
             Hold early                                                          
             survivors                                                           
             (young                                                              
             children or                                                         
             young                                                               
             widow(er)s)                                                         
             harmless                                                            
             Increase the                                                        
             surviving                                                           
             spouse                                                              
             benefit to                                                          
             2/3 or 3/4                                                          
             of the                                                              
             combined                                                            
             couples'                                                            
             benefit                                                             
             Increase                                                            
             benefits for                                                        
             aged                                                                
             survivors                                                           
             Increase the                                                        
             early                                                               
             retirement                                                          
             ageb                                                                

Source: GAO.

aIf the spouse is eligible for retirement benefits based on his or her own
earnings record, Social Security will pay that amount first. However, if
the spouse benefit (based on his or her husband's earnings record) would
be a higher amount, Social Security will combine the benefits and pay the
higher amount. A spouse receiving such a benefit may also be eligible for
a care credit for his or her own earnings record. This care credit could
change the initial benefit amount.

bWhile an increase in the early retirement age could increase the initial
benefit amounts for those who retire at the new early retirement age,
others who may need to stop working at 62 may have no Social Security
payments for the elapsed time.

29This table results from our discussions with knowledgeable experts and
relevant officials as well as a review of the current literature on Social
Security reform. For more information on how these options could work, see
appendix I.

There are several protection options that could be applied to all disabled
workers and dependents. Under a full exemption, beneficiaries would not be
subject to a reform and their benefits would remain unchanged. Under a
partial exemption, beneficiaries would not be subject to a reform until a
certain point in time. For example, disabled workers could be exempt from
benefit changes until they are converted to the Old Age program at the
full retirement age. At this point, their benefit amount would be
recalculated to reflect the reform in proportion to the years they spent
working. In addition, a super COLA could help protect the benefits of
disabled workers and dependents. A super COLA would mitigate some of the
effects of a benefit-reducing reform by annually increasing benefits at a
rate above the consumer price index--which is currently used to index
benefits.

Some protection options could cover all dependents by increasing the
percentage of the worker's benefit that the dependent receives. (See app.
II for more detail on how dependent benefits are calculated.) For example,
a number of proposals have called for increasing the percentage of the
worker's benefit that widow(er)s receive. Another option that could
protect the benefits of a wide range of dependents would be to raise the
maximum benefit that families can receive based on one worker's earnings
record. Other protection options, such as caregiver credits, could focus
on protecting particular groups of dependents. Several reform proposals
have, in fact, called for providing caregiver credits to individuals who
spend time out of the workforce to care for their dependents or to those
with reduced or low earnings while attending to care-giving
responsibilities. Some proposals assign caregivers a specified level of
earnings for each year the caregiver received zero or low earnings
compared to prior years. Other proposals exclude zero-earning care years
from the initial benefit calculation.30 Another option specific to a
certain type of dependent would be to increase benefits for aged
survivors, since they are more likely to rely on Social Security to stay
out of poverty and could have fewer opportunities, such as returning to
work, to respond to benefit-reducing reforms.

Increasing the early retirement age could offer some protection for
survivors. If the early retirement age were raised--for example, from 62
to 64--then workers who take early retirement would receive actuarially
adjusted benefits for a shorter period of time under the new early
retirement age, and thus their monthly benefits would be relatively higher
than the monthly benefits they would have received if they had retired at
the current early retirement age.31 Since a dependent's benefit is linked
to the worker's initial benefit amount, an increase in the worker's
benefit would also increase the dependent's benefit, mitigating some of
the negative effects of other reforms. Similarly, raising the FRA coupled
with a partial exemption from a benefit reduction could offer some
additional protection for disabled worker benefits. With an increase in
the FRA, disabled workers would receive (exempted) DI benefits for a
longer period of time because the age at which their disability benefits
are converted to retiree benefits would rise with the new FRA.

30For more information on caregiver credits and other program
modifications, see GAO, Retirement Security: Women Face Challenges in
Ensuring Financial Security in Retirement, [54]GAO-08-105 (Washington,
D.C.: Oct. 11, 2007).

31 This would not be the case if at age 62 a worker could no longer remain
in the workforce. In such a case, a person may have to apply for DI or may
be unemployed (and have low income) for the 2 years prior to taking early
retirement.

Some Benefit Protections Could Restore Benefits to Levels near Those Scheduled
under Current Law

In general, the reform elements we examined reduce median lifetime
benefits for disabled workers and dependents. Because disabled workers may
not have the financial resources--especially earnings related income--to
adjust to benefit reductions, we explored the interaction of reform
elements and certain options to offset them.

According to our projections, protections from a reduction in the COLA
could restore benefits of disabled workers to levels close to those
scheduled under current law. Reducing the COLA by one percentage point
would result in about a 10 percent decrease in median lifetime benefits
for workers who become disabled before age 60. To offset such a decrease,
they could be partially or fully exempted. With a COLA reduction, a
partial exemption would mean that the Social Security Administration would
increase a disabled worker's benefits annually as scheduled under current
law (i.e., using the full COLA) until the worker reached the full
retirement age. At that point, the disabled worker's benefits would grow
annually by the reduced COLA (1 percentage point lower than what it would
be under current law). Our projections showed that a partial exemption as
described above would raise median lifetime benefits from their reduced
levels by 7 percent (up to 96 percent of scheduled levels under current
law). In contrast, a full exemption would allow annual COLA adjustments in
line with current law until death (fig. 8).32

32In figures 8 to 11, the benefits under a full exemption do not
completely return to pre-reform levels. This is because some individuals
receiving disabled worker benefits also received some benefits as
dependents from another worker's record. As those benefits would not be
subject to the same protections, median lifetime benefits for such
disabled workers may remain slightly lower than those under current law.

Figure 8: Projected Median Lifetime Benefits under a COLA Reduction of 1
Percentage Point (1985 Cohort)

Notes: In figures 8 to 11, the population labeled "Never disabled" refers
to all beneficiaries--dependent or retirees on their own record--who were
never disabled.

In figures 8 to 11, the benefits under a full exemption do not completely
return to pre-reform levels. This is because some individuals receiving
disabled worker benefits also received some benefits as dependents from
another worker's record at some point in their lifetime. As those benefits
would not be subject to the same protections, median lifetime benefits for
such disabled workers may remain slightly lower than those under current
law.

In addition to a decrease in the COLA, we analyzed options for protecting
the benefits of disabled workers under three reform elements that have an
impact on the initial benefit amount a disabled worker receives--price
indexing, longevity indexing, and progressive price indexing. There are
several protection options for mitigating the effects of these reform
elements, including full and partial exemptions. In the case of price
indexing initial benefits, we projected that the median lifetime benefits
of disabled workers would be about 75 percent of the median benefits under
current law (fig. 9). A full exemption for disabled workers would raise
the benefits of those disabled workers who exclusively receive DI benefits
to the currently scheduled levels. However, a partial exemption from price
indexing would restore the median lifetime benefit to 89 to 90 percent of
scheduled levels, depending on how the partial exemption is implemented.33
One type of partial exemption (Type I) uses price indexing to calculate
the portion of the benefits based on the years a person is out of the
workforce and receiving DI benefits. In contrast, the other type of
partial exemption (Type II) uses wage indexing to cover the same time
period. (For more details, please see app. I.) The difference between the
two partial exemptions becomes more substantial the earlier one becomes
disabled, as the difference between wages and prices increases over time.
While offering some protection from benefit reductions, both types of
partial exemptions involve a recalculation of benefits at the full
retirement age. This recalculation would result in lower benefits for the
DI recipient and could create a potential problem if that individual
relied on the prior benefit amount and had limited options for replacing
the lost income. (See figs. 10 and 11 for longevity indexing and
progressive price indexing, respectively.)

33We modeled two different ways of partially exempting disabled workers
from the various indexing reforms--one follows the methods outlined in the
Kolbe-Stenholm (2001) proposal, which we refer to as Type I . The other
partial exemption follows the methods outlined in the Graham (2003)
proposal, which we refer to as Type II. For more details on the
difference, please see appendix I.

Figure 9: Projected Median Lifetime Benefits under Price Indexing (1985
Cohort)

Notes: In figures 8 to 11, the population labeled "Never disabled" refers
to all beneficiaries--dependent or retirees on their own record--who were
never disabled.

In figures 8 to 11, the benefits under a full exemption do not completely
return to pre-reform levels. This is because some individuals receiving
disabled worker benefits also received some benefits as dependents from
another worker's record at some point in their lifetime. As those benefits
would not be subject to the same protections, median lifetime benefits for
such disabled workers may remain slightly lower than those under current
law.

Figure 10: Projected Median Lifetime Benefits under Longevity Indexing
(1985 Cohort)

Notes: In figures 8 to11, the population labeled "Never disabled" refers
to all beneficiaries--dependent or retirees on their own record--who were
never disabled.

In figures 8 to 11, the benefits under a full exemption do not completely
return to pre-reform levels. This is because some individuals receiving
disabled worker benefits also received some benefits as dependents from
another worker's record at some point in their lifetime. As those benefits
would not be subject to the same protections, median lifetime benefits for
such disabled workers may remain slightly lower than those under current
law.

Figure 11: Projected Median Lifetime Benefits under Progressive Price
Indexing (1985 Cohort)

Notes: In figures 8 to 11, the population labeled "Never disabled" refers
to all beneficiaries--dependent or retirees on their own record--who were
never disabled.

In figures 8 to 11, the benefits under a full exemption do not completely
return to pre-reform levels. This is because some individuals receiving
disabled worker benefits also received some benefits as dependents from
another worker's record at some point in their lifetime. As those benefits
would not be subject to the same protections, median lifetime benefits for
such disabled workers may remain slightly lower than those under current
law.

Another protection option would be to allow disability benefits to grow at
a greater rate than other benefits. For example, disabled workers could be
explicitly included in the scope of the reform, and receive reduced
initial benefits. However, instead of receiving annual increases based on
the current-law COLA, disabled workers could have their benefits increased
by a "super COLA"--one that is set above the Consumer Price Index. In this
case, benefits for the disabled would grow at a faster rate than they
would under current law and could approach or even exceed current law
levels. Variations on the super-COLA could include an "age-indexed super
COLA" which would be greater for those disabled at younger ages. For those
workers who become disabled near the full retirement age the COLA would be
closer to that used for retirees. These protections could be particularly
beneficial for disabled workers who receive benefits for a prolonged
period of time.

Protections for Dependent Benefits Could Be Targeted to Certain Vulnerable
Beneficiaries, Including Survivors

While protections for disabled workers would generally cover all such
beneficiaries, the options for protecting dependent benefits could be more
targeted to specific dependents and not necessarily applied to the full
range of dependents, which includes spouses, divorcees, widow(er)s, and
child survivors. The circumstances around which a person becomes a
dependent vary greatly, as does the role of Social Security benefits in
their lives. For some, Social Security may be the primary source of
support; for others, it may be only a small proportion of their income.

Protections could target children, who make up about 8 percent of Social
Security beneficiaries, receiving benefits as the survivors or dependents
of disabled or retired workers. Table 3 shows the number of children who
receive benefits in each category and the average monthly benefit for
these children. One way to protect the benefits of children would be to
exempt them from any reform, keeping their benefit calculation tied to
current law. Another way to protect their benefits to some degree would be
to raise the maximum benefit a family could receive on a single worker's
earnings record.34

34Rather than being an absolute amount, the family maximum is defined as a
percentage of the primary worker's benefit. Therefore, if a reform reduces
benefits, the maximum amount a family could receive would also decrease.
See appendix II for more details on the family maximum.

Table 3: Child Beneficiaries

                Number of                                                     
                 children Children as a                               Average 
                receiving percentage of     Total monthly benefit     monthly 
                 benefits       program      children receive (in benefit (in 
Program    (thousands) beneficiaries      millions of dollars)    dollars) 
All               3959           8.0                      1993         503 
Old Age            485           1.4                       253         522 
Survivors         1850          28.6                      1269         686 
Disability        1624          18.6                       471         290 

Source: GAO analysis. of Social Security data.

Note: Data accessed from
[55]http://www.ssa.gov/policy/docs/quickfacts/stat_snapshot/ on August 29,
2007. The children referred to in this chart receive benefits off someone
else's work record - generally a parent. For example, the 1,624,000
children receiving benefits from the DI program have a parent/guardian who
is disabled.

The majority of experts with whom we spoke told us that increasing the
maximum amount that a family could receive from one worker's earnings
record could help protect child and other dependent benefits. Such an
increase could help those dependents who are constrained by the family
maximum. A family may have several people receiving benefits based on one
worker's record. The sum of the family members' benefits may exceed the
specified maximum, which is calculated as a percentage of the worker's
benefit amount. Thus, any reform that would result in a decrease in the
primary benefit amount would also result in a decrease in the amount that
each eligible family member would receive and a corresponding decrease in
the total amount a family would receive.

Certain options, including increased allowable benefits for widows or
partial exemptions, could be designed to protect the benefits of
widow(er)s or others who may have fewer resources available to them. Under
current law, widows and widowers can collect 100 percent of their deceased
spouse's benefits (or their own benefit--whichever is greater); a "widow's
boost" would allow them to receive up to 75 percent of the couple's
combined benefits. Widow(er)s may rely on Social Security for a large
percentage of their retirement income, in part because they may live many
years beyond the exhaustion of other financial resources, may find it
difficult to work, or may incur large health expenses that deplete their
other resources.

A reduction in the COLA may be particularly detrimental to the lifetime
benefits of those who live long lives, because the effect of reducing the
COLA is compounded over time. As such, it may be desirable to protect
older widow(er)s--along with other individuals who receive benefits for a
prolonged period of time--from the effect of a COLA reduction. For
example, in our projections for the COLA reduction, we found that for the
group of widows who received some benefits and who died before age 75,
median lifetime benefits would be approximately 93 percent of those under
current law. In contrast for those who lived past age 95, median lifetime
benefits would be only 83 percent of currently scheduled levels.

Some Protection Options May Create New Costs and Unintended Incentives for the
Social Security Program

The options for protecting the benefits of disabled workers and those of
dependents come at a cost to the Social Security program in terms of its
solvency. In addition, some protections options may create incentives for
people to apply to the Disability Insurance program if DI benefits
increase while retirement benefits stay stable. Further, protection
options could provide disincentives for some to return to work.

Protecting the Benefits of Disabled Workers and Dependents Would Increase Costs
for the Social Security Program

The Social Security reform elements we examined were designed primarily to
improve program solvency. These reform elements would generally reduce
benefits from their currently scheduled but underfunded levels.35 While
protecting the benefits of disabled workers and dependents may be socially
desirable, such protection would come at some cost to the Social Security
program. In particular, the protections lessen the degree to which the
potential reforms could restore solvency. One could counter these costs
with further benefit reductions to beneficiaries considered less
vulnerable than those recipients whose benefits are specifically
protected. That is, reform packages with certain benefit protections for
vulnerable populations may necessitate further reductions in the benefits
of retired workers or increases in revenues to achieve the intended
solvency effect. In addition to the effects on solvency, some of the
protections discussed may also have administrative costs associated with
them.

35As mentioned earlier, solvency could also be improved by increasing
revenues (taxes).

Certain Protections May Create Incentives for Individuals to Apply for
Disability Benefits

Protecting the benefits of disabled workers may increase the number of
people who apply for disability benefits. This may also be relevant to
certain reform elements. An increase in the full retirement age coupled
with the reduction in benefits for early retirement could motivate some
individuals approaching the early retirement age to apply for disability
benefits, if they believed that they could qualify for the now greater DI
benefits.36 For example, before a change in the retirement age a worker
who is a year away from the full retirement age, and who would qualify for
DI but is unsure of that outcome, may choose to wait and only receive Old
Age benefits. Once the full retirement age is raised, this worker may
choose to apply for DI, rather than waiting to receive retirement
benefits. The greater the benefit disparity between the two programs, the
more likely it may be that DI applications and enrollment will increase.
Thus, the potential for an increase in DI program costs exists with any
reform elements that decrease the generosity of the Old Age component of
OASDI without a corresponding decrease in that of the DI component.37

Under current law, there may already be an incentive for older workers to
apply for DI rather than retire early. Using individual level data from
the simulation model, we analyzed the benefits of two similar individuals
under current law and under price indexing with and without full and
partial exemptions. Both of the simulated individuals had similar lifetime
earnings, close to the median for the simulated 1985 cohort, and both
would have received initial benefits at age 62. However, they differed in
two significant ways: one retired at age 62, while the other was disabled
at age 62, and the retiree had lower lifetime benefits under current
law.38 The retiree, who died at age 84, had lifetime benefits of about
$433,000, while the disabled worker, who died at age 82, had lifetime
benefits of about $505,000--about 16 percent higher than those of the
retired worker.

36Some research has been done on whether there is a link between the
increase in the FRA and the recent increase in the number and percentage
of people applying for DI. For example, Duggan, Singleton, and Song (2005)
find that the 1985 increase in the FRA had an important effect on the
percent of people applying for DI.

37Besides incentive-related costs, increasing the FRA would create other
costs for the DI program by expanding the size of the DI-eligible
population. Workers who remain in the workforce past the old FRA and
become injured prior to the new FRA may enroll in the DI program.

38In part, this may be because the retired individual chose to retire
early (at age 62) and thus his benefits were actuarially reduced. In
addition, while the present value of the individuals' lifetime earnings
were similar, the distribution of these earnings could have been different
and as such, the average indexed monthly earnings used as the basis for
determining benefit amounts could be different.

A full exemption for disabled workers from certain reform elements could
similarly create discrepancies between the two programs, resulting in
incentives to apply for the DI program. Under price indexing, the lifetime
benefits of both individuals would be reduced, but the relative difference
would remain at about 16 percent. However, if disabled workers were fully
exempted from price indexing, the simulated disabled worker's lifetime
benefits would be back to the initial amount of $505,000, or 72 percent
greater than those of the retired worker. This difference in potential
benefits would likely increase the incentive to apply for the DI program.
Figure 12 and table 4 show the total lifetime benefits and the average
monthly benefits of these two simulated individuals under current law,
price indexing, and with exemptions.

Figure 12: Simulated Total Lifetime Benefits for Two Similar Individuals
Who Began Receiving Benefits at Age 62, Exemptions Apply to Disabled
Workers (1985 Cohort)

Table 4: Total Lifetime and Average Monthly Benefits of Two Simulated
Individuals Who Began Receiving Benefits at Age 62, Exemptions Apply to
Disabled Workers (1985 Cohort)

                                              Average monthly
             Total lifetime benefits             benefits
                             Difference                          Difference
                                   as a                                as a
                             percentage                          percentage
                                     of                              of the
                             theretired                             retired
           Retired  Disabled   worker's       Retired  Disabled    worker's
            worker    worker   benefits        worker    worker    benefits
 Current   $433,000 $505,000     16.63%                $1,640      $2,104 28.29%
 law                                                                        
 Price     $293,000 $341,000     16.38%                $1,110      $1,421 28.02%
 indexing                                                                   
 Full      $293,000 $505,000     72.35%                $1,110      $2,104 89.59%
 exemption                                                                  
 Partial   $293,000 $389,380     32.89%                $1,110      $1,622 46.18%
 exemption                                                                  

Source: GAO analysis of GEMINI data.

However, partial rather than full exemptions or other protections, such as
an age-indexed super COLA, could provide benefit protections without
substantially increasing the disparity between the programs for people
approaching the early or full retirement ages. Under a partial exemption,
in which the disabled worker would be exempted from the reform until full
retirement age, the added incentive that could be created by a full
exemption would be reduced. Such a partial exemption for the disabled
worker in our example would result in lifetime benefits that are about 33
percent higher than those of the retired worker under price indexing.

Certain Protections May Affect the Work Decisions of Beneficiaries

The family maximum limits the amount that can be received off of a
worker's record. This limit is compatible with the incentive for
individuals to work. Changing such a limit could affect beneficiaries'
work decisions. For example, protecting benefits of dependents by
increasing the family maximum could affect an individual's work decisions.
Under the current family maximum with a benefit reduction in place, if a
person chooses to work 30 hours a week, an increase in the total amount a
family (or individual dependents) could receive might affect this decision
and decrease the person's time in the workforce. In such a case, the
individual may find that the increase in the benefits received would allow
for fewer weekly hours of work without a change in total income. In
addition, protections that increase the benefits of disabled workers, such
as the super COLA, can also create disincentives for such beneficiaries to
return to work. As such, some individuals may continue to rely on the DI
program, rather than finding a way to re-enter the workforce.

Conclusions

Social Security's financial challenges may result in program modifications
that may include benefit reductions. These benefit reductions will likely
affect all beneficiaries, including vulnerable individuals who may not be
able to adjust to these reductions or who rely on Social Security as their
primary source of income. While protecting the benefits of vulnerable
populations may be desirable, such action does come at a cost. Further
benefit reductions or revenue increases would be needed to achieve program
solvency. These offsets, in turn, may create new financial vulnerabilities
among other beneficiaries who would bear the burden of these protections.

Few reform proposals consider the impact that benefit reductions would
have on all beneficiary types, instead treating all beneficiaries
similarly. However, some special consideration should be given to the
effects of the reform on the benefits of the most vulnerable, especially
when these individuals are disproportionately affected. If the solution to
Social Security's financing problems includes benefit reductions, then the
equal treatment of all beneficiaries may need to be reconsidered, and the
complex interactions of benefit reductions, protections, and direct and
indirect costs to the system and to other retirees will need to be weighed
carefully. Benefit protections can be a part of a comprehensive reform
package and the reform debate should consider the design, inclusion, and
implications of such measures to assure income adequacy. Likewise, to the
extent that Social Security aligns the disability program with the current
state of science, medicine, technology, and labor market conditions, such
modernization should also be considered.

Matter for Congressional Consideration

Accordingly, in light of potential reform, Congress should consider the
potential implications of reform on disability and dependent
beneficiaries. Such a review might usefully be coordinated with any
modernization of the Social Security disability program.

Agency Comments

We provided a draft of this report to SSA and the Department of the
Treasury, which generally agreed with our findings. Both provided
technical comments, and SSA also provided general comments, which appear
in appendix III. We incorporated the comments throughout our report as
appropriate.

In general, SSA concurred with the methodology, overall findings, and

conclusions of the report. However, SSA felt that the report could benefit
from a more direct comparison of disabled beneficiaries and retired
beneficiaries (and a similar construct for dependents). While such a
comparison could be beneficial and give context to the reform discussion,
this report was premised on the notion that certain beneficiaries would be
less able to offset benefit reductions, rather than a comparison of
relative welfare.

Finally, GAO agrees that one could better assess the degree to which a
reform element or protection option support the program's goal of adequacy
if benefits were compared to a standard of adequacy; however such a
comparison was beyond the scope of the current study.

As agreed with your offices, unless you publicly announce its contents
earlier, we plan no further distribution until 30 days after the date of
this letter. At that time, we will send copies of this report to the
Social Security Administration and the Department of the Treasury, as well
as other interested parties. Copies will also be made available to others
upon request. In addition, the report will be available at no charge on
the GAO Web site at [56]http://www.gao.gov . Please contact me at (202)
512-7215, if you have any questions about this report. Other major
contributors include Michael Collins, Nagla'a El-Hodiri, Jennifer Gregory,
Joe Applebaum, Melinda Cordero, Mark Goldwein, Meaghan Mann, and Dan
Schwimer.

Barbara D. Bovbjerg
Director, Education, Workforce, and Income Security
Issues

Appendix I: Methodology

Microsimulation Model

To analyze the effects of individual reform elements and certain
protections from these reforms on Social Security benefit levels for
disabled workers and dependents, we simulated their benefits using the
Policy Simulation Group's (PSG) microsimulation models. We based our
analysis on projected lifetime benefits for a simulated 1985 birth cohort.
In order to have a point of comparison, we also used the microsimulation
models to simulate Social Security benefits of retirees who receive
benefits on their own record.

Description

For our simulations, we used the PSG's Social Security and Accounts
Simulator (SSASIM) and Genuine Microsimulation of Social Security Accounts
(GEMINI) simulation models. GEMINI simulates Social Security benefits and
taxes for large representative samples of people born in the same year.
GEMINI simulates all types of Social Security benefits including retired
workers', spouses', survivors', and disability benefits. It can be used to
model a variety of Social Security reforms. GEMINI uses inputs from
SSASIM, which has been used in numerous GAO reports,1 and from the Pension
Simulator (PENSIM), which was developed for the Department of Labor.
GEMINI relies on SSASIM for economic and demographic projections and
relies on PENSIM for simulated life histories of large representative
samples of people born in the same year and their spouses.2 Life histories
include educational attainment, labor force participation, earnings, job
mobility, marriage, disability, childbirth, retirement, and death. Life
histories are validated by PSG against data from the Survey of Income and
Program Participation, the Current Population Survey, Modeling Income in
the Near Term (MINT3),3 and the Panel Study of Income Dynamics.
Additionally, any projected statistics (such as life expectancy,
employment patterns, and marital status at age 60) are, where possible,
consistent with intermediate cost projections from Social Security
Administration's Office of the Chief Actuary (OCACT). At their best, such
models can provide only very rough estimates of future incomes. However,
these estimates may be useful for comparing future incomes across
alternative policy scenarios and over time.

1See [57]GAO-08-105 ; GAO, Social Security Reform: Implications of
Different Indexing Choices, [58]GAO-06-804 (Washington, D.C.: Sept. 14,
2005); and GAO, Distribution of Benefits and Taxes Relative to Earnings
Level, [59]GAO-04-747 (Washington, D.C.: June 15, 2004).

2While these models use sample data, our report, like others using these
models, does not address the issue of sampling errors. The results of the
analysis reflect outcomes for individuals in the simulated populations and
do not attempt to estimate outcomes for an actual population.

3MINT3 is a detailed microsimulation model developed jointly by the Social
Security Administration, the Brookings Institution, RAND, and the Urban
Institute to project the distribution of income in retirement for the 1931
to 1960 birth cohorts.

For this report, we used the Genuine Microsimulation of Social Security
Accounts (GEMINI) to estimate Social Security benefits for a sample of
approximately 2 percent of individuals born in the 1985 cohort. This
consisted of just over 97,000 individuals, with positive benefit levels
for just over 83,400 individuals.

For our baseline, benefits were simulated under current law. These
simulations are based on the Social Security Trustees' 2007 intermediate
economic and actuarial assumptions. While our simulations provide
projections of future retirement income, as promised under current law,
there is a considerable amount of uncertainty involved with these
estimates. Since these estimates could change significantly, depending on
assumptions used and behavior responses, they should not be considered
predictions. In addition, because simulations are sensitive to economic
and demographic assumptions, it is more appropriate to compare benefits
across the scenarios than to focus on the actual estimates themselves.

Evaluating Reform Proposals

In general, GAO has suggested that policy makers should consider three
basic criteria when evaluating reform proposals4

           o the extent to which the proposal achieves sustainable solvency
           and how the proposal would affect the economy and the federal
           budget;
           o the balance struck between the goals of individual equity5
           (rates of return on individual contributions) and income adequacy6
           (level and certainty of monthly benefits); and
           o how readily such changes could be implemented, administered, and
           explained to the public.

4See GAO, Social Security: Criteria for Evaluating Reform Proposals,
[60]GAO/T-HEHS-99-94 (Washington, D.C.: Mar. 25, 1999) and GAO, Social
Security: Evaluating Reform Proposals, [61]GAO/AIMD/HEHS-00-29
(Washington, D.C.: Nov. 4, 1999).

5For a discussion of individual equity issues, see GAO, Social Security:
Issues in Comparing Rates of Return with Market Investments,
[62]GAO/HEHS-99-110 (Washington, D.C.: Aug. 5, 1999).

6See GAO, Social Security: Program's Role in Helping Ensure Income
Adequacy, [63]GAO-02-62 (Washington, D.C.: Nov. 30, 2001).

Moreover, reform proposals should be evaluated as packages that strike a
balance among the individual elements of the proposal and the interactions
among these elements. The overall evaluation of any particular reform
proposal depends on the weight individual policy makers place on each of
the above criteria.

However, for the purposes of this study, we did evaluate individual reform
elements and individual protection options. We looked at specific elements
of solvency-improving reform proposals to analyze their effects on our
populations of interest--disabled workers and dependents. In particular,
we wanted to isolate each element in order to project the magnitude of
their effects on these populations and in order to model certain
protection options. Nevertheless, we recognize that there would be
important interactive effects with any set of reforms and maintain the
importance of considering all possible effects of any reform package as a
whole.

Assumptions and Limitations

Simulating retirement income almost 50 years into the future requires many
assumptions and simplifications and, consequently, our simulations have a
number of limitations. A primary limitation of our analysis is that our
simulations do not include important components of retirement income, such
as personal savings, earnings in retirement, health benefits, and other
public assistance programs such as SSI. These sources could also be used
to offset benefit reductions. In addition, the model is structured such
that changes in Social Security benefits have no behavioral consequences
in terms of decisions regarding work, disability, or retirement. As a
result, the individuals would not change their work decisions
(earnings/retirement) based on the reforms.

  2007 Social Security Trustees' Assumptions

The simulations are based on economic and demographic assumptions from the
2007 Social Security Trustees' report.7 We used Trustees' intermediate
assumptions for inflation, real wage growth, mortality decline,
immigration, labor force participation, and interest rates.

7The Board of Trustees, Federal Old-Age and Survivors Insurance and
Disability Insurance Trust Funds, The 2007 Annual Report of the Board of
Trustees of the Federal Old-Age and Survivors Insurance and Disability
Insurance Trust Funds (Washington, D.C.: Apr. 23, 2007).

Specification of Disabled Workers and Dependents

For the purpose of this analysis:

Disabled Workers consisted of those individuals who had a valid disability
onset age and who received a positive amount of benefits from Disability
Insurance. This includes any disabled workers who were dependents at some
point in time.

Dependents consisted of those individuals who received benefits based on
someone else's earnings record, at any point. Specifically, dependents
excluded those who only received benefits based on their own record as
retirees or as disabled workers.

Description of Social Security Reform Elements

Longevity Indexing

To simulate longevity indexing, which links the growth of initial benefits
to changes in life expectancy, we successively modified the PIA formula
replacement factors (90, 32, 15) beginning in 2009, reducing them annually
by multiplying them by 0.995. This specification mimics provision 1 of
Model 3 of the President's Commission to Strengthen Social Security
(CSSS).8 The CSSS solvency memorandum notes that the 0.995 successive
reductions "reduces monthly benefit levels by an amount equivalent to
increasing the normal retirement age (NRA) for retired workers by enough
to maintain a constant life expectancy at NRA, for any fixed age of
benefit entitlement."9 This provision as specified and scored--using the
intermediate assumptions of the 2001 Trustees' report--in the CSSS memo by
SSA's Office of the Chief Actuary would improve the long-range OASDI
actuarial balance (reduce the actuarial deficit) by an estimated 1.17
percent of taxable payroll.

8For more information on provision 1 or Model 3, see page 8 of the CSSS
proposal at http://www.ssa.gov/OACT/solvency/PresComm_20020131.pdf.

9We chose the CSSS specification because it was already scored and readily
available. Other constructions or interpretations of a longevity index are
certainly possible. For example, life expectancy at birth or some other
age could be used. Further, life expectancy could be defined as period or
cohort. A period life table represents the mortality conditions at a
specific point in time, whereas a cohort table depicts the mortality
conditions of a specific group of individuals born in the same year or
series of years.

Price Indexing

We also simulated the effects of price indexing, where initial benefits
would be indexed to the consumer price index (CPI) in order to limit the
growth of benefits. We successively modified the primary insurance amount
(PIA) formula replacement factors (90, 32, and 15) beginning in 2012,
reducing them successively by real wage growth in the second prior year.
This specification mimics provision B6 of the August 10, 2005, memorandum
to SSA's Chief Actuary regarding the provision requested by the Social
Security Advisory Board (SSAB), which is an update of provision 1 of Model
2 of the CSSS.10 As noted in the CSSS's solvency memorandum from SSA's
Chief Actuary, "[t]his provision would result in increasing benefit levels
for individuals with equivalent lifetime earnings across generations
(relative to the average wage level) at the rate of price growth (increase
in the CPI), rather than at the rate of growth in the average wage level
as in current law." This provision as specified and scored by OCACT in the
SSAB memo would improve the long-range OASDI actuarial balance (reduce the
actuarial deficit) by an estimated 2.38 percent of taxable payroll.

Progressive Price Indexing

To simulate the effects of implementing a progressive price index, we
mimicked provision B7 of the August 10, 2005, memorandum to SSA's Chief
Actuary.11 We created a new bend point at the 30th percentile of earnings,
beginning in 2012. We maintained current-law benefits for earners at the
30th percentile and below. We also maintained the lower two PIA formula
replacement factors (90 and 32). We reduced the upper two PIA formula
replacement factors (32 and 15) so that maximum worker benefits from one
generation to the next grew by inflation rather than the growth in average
wages. This provision as specified and scored by OCACT would improve the
long-range OASDI actuarial balance (reduce the actuarial deficit) by an
estimated 1.43 percent of taxable payroll.

10See page 3 of
http://www.ssab.gov/documents/advisoryboardmemo--2005tr--08102005.pdf for
description of the provision. For the original provision from the
President's Commission to Strengthen Social Security, see page 4 of
[64]http://www.ssa.gov/OACT/solvency/PresComm_20020131.pdf .

11See the August 10, 2005, memorandum to SSA's Chief Actuary, page 3 of
http://www.ssab.gov/documents/advisoryboardmemo--2005tr--08102005.pdf, for
a description of the provision.

Increase the Number of Computation Years Used in the Initial Benefit Calculation

In our modeling of this reform element, we gradually reduced the number of
drop out years from 5 to 0, thereby extending the number of computation
years from 35 to 40. The number of computation years would increase to 36
in 2007, 37 in 2008, 38 in 2010, 39 in 2012, and 40 in 2014. This
specification mimics provision B2 of the August 10, 2005 memorandum to
SSA's Chief Actuary.12 This provision as specified and scored by OCACT
would improve the long-range OASDI actuarial balance (reduce the actuarial
deficit) by an estimated 0.46 percent of taxable payroll.

Reduction in the Cost-of-Living Adjustment

We also simulated a reduction in the cost-of-living adjustment (COLA) of
one percentage point, beginning in 2012. This specification mimics
provision A2 of the August 10, 2005, memorandum to SSA's Chief Actuary.13
This provision as specified and scored by OCACT would improve the
long-range OASDI actuarial balance (reduce the actuarial deficit) by an
estimated 1.49 percent of taxable payroll. Some reform proposals have
called for reducing the COLA by about 0.2 percent to 0.4 percent, in
response to methodological concerns that the CPI for urban wage earners
and clerical workers, the current CPI measure used to adjust benefits,
overstates inflation. The intent of these proposals is to implement a COLA
that may more accurately reflect inflation.

12See the August 10, 2005, memorandum to SSA's Chief Actuary, page 3 of
http://www.ssab.gov/documents/advisoryboardmemo--2005tr--08102005.pdf, for
description of the provision.

13See page 3 of
http://www.ssab.gov/documents/advisoryboardmemo--2005tr--08102005.pdf for
description of the provision. We did make one adjustment to this
provision. We modeled the change beginning in 2012 so that the simulation
start date was consistent with the start date for price indexing and
progressive price indexing.

Description of Protection Options

Table 5: Options for Protecting Disabled Worker and Dependent Benefits
from Social Security Reform

Protection option                          How the protection option    
available for     Protection option        could work                   
Any beneficiary   Full exemption           The beneficiary would not be 
type                                       subject to reform. Benefits  
                                              would be unchanged.          
                     Partial exemption        Benefits would not be subject
                                              to reform until a certain point
                                              in time. At this point in time,
                                              benefits would be adjusted to
                                              reflect reform.              
                     Minimum benefit          Beneficiaries would be       
                                              guaranteed a specified minimum
                                              benefit level.               
                     Super COLA               A larger COLA would be applied
                                              to benefits to partially     
                                              mitigate reductions.         
                     Age-indexed super COLA   Works similarly to the super 
                                              COLA. However, this COLA would
                                              vary in size, with younger   
                                              beneficiaries receiving larger
                                              adjustments.                 
Children and      Increase the percentage  Children and spouses         
families of       of the worker's benefit  receiving benefits from a    
disabled workers  that the dependent       parent's or spouse's         
                     family member receives   earnings record receive a    
                     as his/her benefit       set percentage of their      
                                              parent's or spouse's         
                                              benefits as their benefit.   
                                              This option would increase   
                                              the size of this percentage. 
                     Increase the family      Families would be subject to a
                     maximum benefit level    higher family maximum benefit
                     for DI                   level if they receive benefits
                                              off the earnings record of a 
                                              disabled worker.             
                     Increase the percentage  This option would increase the
                     of the worker's benefit  percentage used to determine a
                     that a dependent child   dependent child or DAC's     
                     or a disabled adult      benefit level while also     
                     child (DAC) receives as  increasing the family maximum
                     his/her benefit in       benefit level.               
                     combination with                                      
                     increasing the family                                 
                     maximum benefit                                       
                     Decouple DAC benefits    This option would exclude DAC
                     from other family        benefits from the family     
                     benefits                 benefit calculation. While a 
                                              DAC would continue to receive
                                              benefits based from his/her  
                                              parent's earnings record, DAC
                                              benefits would not be included
                                              in the calculation of total  
                                              family benefits when         
                                              determining whether family   
                                              benefits have exceeded the   
                                              maximum family benefit.      
                     Hold initial benefit     If reform adjusts the initial
                     amount harmless for      benefit amount for a disabled
                     family benefits          worker, the amount used for  
                                              calculating the benefits for 
                                              any dependent of a disabled  
                                              worker would be unchanged.   
Children and      Expand eligibility rules This option would alter      
families of       for divorced spouses     eligibility rules for        
retired workers                            divorced spouses (e.g.       
                                              shorten the duration of      
                                              marriage requirement for the 
                                              divorced spouse to receive   
                                              benefits from his/her        
                                              spouse's earnings record).   
                     Increase the percentage  Children and spouses receiving
                     of the worker's benefit  benefits from a parent's or  
                     that the dependent       spouse's earnings record     
                     family member receives   receive a set percentage of  
                     as his/her benefit       their parent's or spouse's   
                                              benefits as their benefit. This
                                              option would increase the size
                                              of this percentage.          
                     Increase the percentage  This option would increase the
                     of the worker's benefit  percentage used to determine a
                     that a dependent child   dependent child or DAC's     
                     or DAC receives as       benefit level while also     
                     his/her benefit in       increasing the family maximum
                     combination with         benefit level.               
                     increasing the family                                 
                     maximum benefit                                       
                     Decouple DAC benefits    This option would exclude    
                     from other family        DAC benefits from the family 
                     benefits                 benefit calculation. While a 
                                              DAC would continue to        
                                              receive benefits based from  
                                              his/her parent's earnings    
                                              record, DAC benefits would   
                                              not be included in the       
                                              calculation of total family  
                                              benefits when determining    
                                              whether family benefits have 
                                              exceeded the maximum family  
                                              benefit.                     
Spouses           Increase the percentage  Husbands and wives receiving 
                     of the worker's benefit  benefits from their spouses' 
                     that the spouse receives earnings records receive a   
                     as his/her benefit       set percentage of their      
                                              spouses' benefits as their   
                                              benefit. This option would   
                                              increase the size of this    
                                              percentage.                  
                     Implement a child/family A child/family care credit   
                     care credit              could be work in one of three
                                              ways:                        
                                                                           
                                              Reduce the number of         
                                              computation years for        
                                              caregivers: A certain number of
                                              years would be "dropped" from
                                              the initial benefit level    
                                              calculation to credit years an
                                              individual spent caring for  
                                              children or other dependent  
                                              family members.              
                                                                           
                                              Credit caregivers a particular
                                              dollar amount for years spent
                                              out of the labor force       
                                              providing care to children:  
                                              Earnings used to calculate an
                                              individual's initial benefit 
                                              level would be adjusted upward
                                              for a particular number of   
                                              years.                       
                                                                           
                                              Provide caregivers with a    
                                              credit equal to 1/2 of median
                                              worker earnings for the years
                                              spent working part-time or   
                                              earning low incomes because  
                                              caring for children or other 
                                              family members: Replaces years
                                              with no or low earnings, up to
                                              a certain number of years.   
                                              Alternatively, instead of using
                                              1/2 of median worker earnings
                                              as the replacement, could use
                                              1/2 of the median earnings from
                                              a caregiver's remaining work 
                                              years.                       
Survivors         Hold initial benefit     If a reform adjusts the      
                     amount harmless for      initial benefit amount for a 
                     family benefits          primary beneficiary, the     
                                              amount for calculating the   
                                              benefits for any survivor of 
                                              a worker would be unchanged. 
                     Hold early survivors     Early survivors would not be 
                     (young children or young subject to any benefit change
                     widow(er)s) harmless     resulting from reform.       
                     Increase the surviving   This option would increase the
                     spouse benefit to 2/3 or size of the benefit a surviving
                     3/4 of the combined      spouse receives to 2/3 or 3/4
                     couples' benefit         of what the couples' combined
                                              benefit would have been.     
                     Increase benefits for    This option increases the size
                     aged survivors           of benefits older survivors  
                                              receive.                     
                     Increase the early       This option would increase the
                     retirement age           early retirement age.        
                                              Survivors' benefits are based
                                              on workers' benefit. If a    
                                              worker takes benefits during 
                                              the early retirement period, 
                                              then the monthly benefit level
                                              is reduced. By increasing the
                                              early retirement age, a worker
                                              wouldn't be able to receive as
                                              large a reduced monthly benefit
                                              or receive benefits for as long
                                              a period of time. Accordingly,
                                              any survivor's benefit would be
                                              larger.                      

Source: GAO.

Options Modeled to Protect Benefits of Disabled Workers

Full Exemption

To simulate the effects of fully exempting disabled workers from the
various reform elements, we modified the simulation to exclude the
benefits of disabled workers from the reform elements. As such, there
would be no recalculation of benefits when the exempted beneficiary
reached full retirement age.

Partial Exemptions

We defined partial exemptions for disabled workers to mean that their
benefit would be exempted from any simulated reform until the FRA and then
would be recalculated. For the COLA reduction, we simply started the one
percentage point reduction at the FRA for disabled workers. However, for
the reforms that involved a change in the initial benefit amount
(longevity indexing, price indexing, and progressive price indexing), we
simulated the recalculation of benefits at the FRA in two different ways.

Partial Exemption Type I--Kolbe-Stenholm

The first partial exemption, which we refer to as Partial Exemption Type
I, followed the Kolbe-Stenholm model of converting benefits at the FRA.
The Kolbe-Stenholm model reduces benefits in proportion to the difference
in the disabled-worker PIA and the retired-worker PIA at the DI-onset age.
This OASI benefit amount would be indexed by the COLA to for the years
between the disability onset age and age 62.

Partial Exemption Type II--Graham

The second partial exemption, or Partial Exemption Type II, followed the
Graham model of converting benefits at the FRA. The mechanism for
converting from DI to Old Age benefits is as follows:

(DIc)(Yd)/40 + (OASI62)(40-Yd)/40: 

where:

DIc is the promised DI benefit level under current law

Yd is the number of years (ages 21 to 62) that the disabled worker
received DI benefits

OASI62 is the OASI benefit level, calculated by computing the PIA under
the reform using the formula applicable for newly eligible retired workers
in the year the converting worker reached age 62. In this case, earnings
from the years prior to disability would be wage indexed. The disability
freeze years14 would apply in computing the AIME.

Data Reliability

To assess the reliability of simulated data from GEMINI, we reviewed PSG's
published validation checks and examined the data for reasonableness and
consistency.

PSG has published a number of validation checks of its simulated life
histories. For example, simulated life expectancy is compared with
projections from the Social Security Trustees; simulated benefits at age
62 are compared with administrative data from SSA; and simulated
educational attainment, labor force participation rates, and job tenure
are compared with values from the Current Population Survey. We found that
simulated statistics for the life histories were reasonably close to the
validation targets.

14The disability freeze is a special rule that helps increase retirement
or disability benefits. People who have a disability or are legally blind,
not on DI, and working may have lower earnings because of the disability
or blindness. In such a case, SSA can exclude those years when calculating
retirement or disability benefits. Because Social Security benefits are
based on average lifetime earnings, the exclusion of those years increases
benefits.

Appendix II: Calculating OASDI Benefits

Social Security offers a variety of types of benefits, and although they
are all based upon the same formula, they are calculated in different
ways. The methods for calculating the different types of benefits are
outlined below1.

Calculating Old-Age Benefits

Old Age benefits are calculated through a four-step process in order to
provide retirees with progressive yet wage-based cash payments (see fig.
13).

1To summarize the calculation of benefits, we relied on several Social
Security publications, including the Social Security Handbook and a
variety of resources available on SSA's website, www.ssa.gov.

Figure 13: Calculating Benefits for Retirees

Notes: Specifically, Social Security's COLAs are based on the consumer
price index for urban wage earners and clerical workers (CPI-W), as
opposed to the CPI series for all urban consumers (CPI-U).

Also the PIA numbers in step 2 refer to workers attaining age 62 in 2007.

First a worker's Average Indexed Monthly Earnings (AIME) is calculated by
indexing the worker's past earnings to changes in average wage levels over
the worker's lifetime and then averaging them. The AIME formula considers
all years in which a worker earned covered earnings.2 It then uses the
number of elapsed years from 1950 or attainment of age 21 through the age
of 62 (or death) and allows for 5 "drop-out years" so that the worker's
highest 35 years of covered indexed earnings are used in the calculation.3
Once the AIME is determined, a progressive formula is applied to the AIME
to yield a worker's Primary Insurance Amount (PIA). In 2007, the PIA
formula had the following bend points: 90 percent of the first $680 of
AIME, plus 32 percent of the next $3,420, and 15 percent of any earnings
above that level (fig. 13). For example, the PIA of a worker whose AIME
was $1000, the equivalent of at $12,000 annual salary, would be the sum of
$612 (90 percent of $680) and $102.40 (32 percent of $320), yielding a
total initial monthly benefit of around $715. Similarly, the PIA of a
worker with an $8,000 AIME (the equivalent of a $96,000 annual salary)
would be the sum of $612 (90 percent of $680), $1094.40 (32 percent of
$3420), and $585 (15 percent of $3,900), for a total of just under $2,292.
Because the formula is both wage-based and progressive, the second worker
receives a much higher actual benefit than the first worker ($2,292 versus
$715), but his benefits are a much lower proportion of his past earnings
than the first worker's benefits (28.6 percent versus 71.4 percent).

2Years with wages greater than those earned from 21 to 62 would also be
considered as long as the worker had "credited" earnings.

If a worker retires at the full retirement age, which is currently between
ages 65 and 66, and legislated to reach 67 in 2027, this PIA represents
the first year's benefit (although it is adjusted for inflation through a
cost-of-living adjustment (COLA)). However, workers can begin receiving
reduced benefits at 62; benefits are progressively larger for each month
workers postpone drawing them, up to age 70.4 In general, benefits are
actuarially neutral to the Social Security program; that is, the reduction
for starting benefits before full retirement age and the credit for
starting after full retirement age are such that the total value of
benefits received over one's lifetime is approximately equivalent for the
average individual.5 Those receiving benefits before the full retirement
age will also be subject to an earnings test. If earned income is above a
certain threshold, Social Security withholds one dollar of benefits for
every two dollars of earning above the threshold. Each year, benefits
receive a COLA to keep pace with inflation.

3For example, if a worker had some of his or her highest covered earnings
between the ages of 18 and 22, these would be counted. The 40 years
between age 22 and 62 simply give the formula the number of years to
consider when picking the top earning years.

4SSA reduces retired worker benefits by 5/9 of 1 percent per month for the
first 36 months and 5/12 of one percent for each additional month that a
worker elects to start benefits in advance of full retirement age.
Conversely, delayed retirement credits increase benefits for each month
the worker delays the start of benefits after full retirement age until
they reach age 70. The factor used to calculate these credits varies by
birth year. For workers born in1943 or later the increase is 2/3 of 1
percent each month (8 percent per year).

Calculating Disability Benefits

Similarly to Old Age benefits, disability benefits are determined by
calculating a worker's AIME, applying the progressive PIA formula to it,
and then adjusting benefit levels through yearly COLAs (fig. 14).

5This is the case if lifetime benefits are calculated on a present value
basis with a discount rate equal to the expected return for the Social
Security trust fund--that is, 2.9 percentage points above the rate of
inflation after 2015, according to the intermediate assumptions in the
Trustees' 2007 report--The Board of Trustees, Federal Old Age and
Survivors Insurance and Federal Disability Insurance Trust Funds, The 2007
Annual Report of the Board of Trustees of the Federal Old Age and
Survivors Insurance and Federal Disability Insurance Trust Funds
(Washington, D.C.: Apr. 23, 2007) 94.

Figure 14: Calculating Benefits for Disabled Workers

Note: The PIA numbers in step three refer to workers qualify for receiving
DI benefits in 2007.

However, because disabled workers are likely to have shorter work
histories, their benefits calculation relies on fewer years of earnings.
In general, the number of years of earnings used to calculate the AIME is
based on the total number of years between when a worker turns 21 and when
he applies for DI. If this number of years is 25 or more, a worker's 5
lowest (or zero) earnings years will be dropped from the calculation. The
number of drop-out years gradually declines as a worker applies for
disability earlier in life. If the disabled worker is 60 at the time of
application, for example, 38 years would have elapsed since age 21. He
will receive 5 drop out years, and his AIME will be calculated based upon
his 33 highest-earning years. In contrast, if a worker applies for DI at
32, he would have only had 10 elapsed years since age 21, and only be
eligible for 2 drop-out years; his AIME would be calculated based upon his
top 8 years. At the full retirement age, disabled workers begin receiving
retirement benefits, instead of disability benefits; however, benefit
levels remain the same and continue to grow through annual COLAs.

Calculating Benefits for Dependents of Retired and Disabled Workers

           o Spouses: In addition to being eligible to receive retirement
           benefits on their own earnings records as early as age 62,
           individuals can also receive dependents' benefits at age 62, based
           on their spouse's benefit amount or, in some cases, that of an
           ex-spouse (table 5). These individuals can collect these benefits
           regardless of whether their spouse is concurrently receiving
           retired or disabled worker benefits. If collection begins at full
           retirement age, these individuals are eligible for either one-half
           of their spouse's benefit amount, or the benefits based on their
           own earnings record; whichever is greater. As with Old Age
           benefits, adjustments are made if these individuals chooses to
           take early retirement.

           o Dependent Children: Dependent children may also qualify for
           one-half of their retired or disabled parent's benefit amount.
           This benefit is available for disabled adult children who are not
           working on a regular basis, children under age 18, or children
           still in high school and under age 19.

           Like other benefits, dependents' benefits receive annual COLAs.
           Dependent benefits are subject to a family maximum, whereby a
           family is limited in the total amount of benefits that can be
           received from a single individual's earnings record. The size of
           the family maximum is currently between 150 percent and 188
           percent of the primary beneficiary's benefit.6

6The formula used to compute the OASI family maximum is similar to that
used to compute the PIA. It involves computing the sum of four separate
percentages of portions of the worker's PIA. For 2007, these are 150
percent of the first $869, 272 percent of the amount between $869 and
$1,255, 134 percent of the amount between $1,255 and $1,636, and 175
percent of the amount over $1,636. The disability family maximum is equal
to 85 percent of the disabled worker's AIME, but cannot be less than his
or her PIA, nor more than 150 percent of his or her PIA.

Table 6: Dependent Benefits

                       Worker retires or                                      
                       becomes disabled       Worker dies                     
Child of worker     Can collect 50% of     Can collect 75% of worker's     
                       worker's benefits if   benefits if under 18, in high   
                       under 18, in high      school, or disabled             
                       school, or disabled                                    
Spouse of worker    Can collect 50% of     Can collect 100% of worker's    
                       worker's benefits at   benefit at FRA or reduced       
                       FRA or reduced amount  amount at age 60                
                       at age 62                                              
Spouse taking care  Can collect 50% of     Can collect 75% of worker's     
of young or         worker's benefits at   benefits if not eligible for    
disabled child      any age                more through other provisions   
Disabled spouse     No special provision   Can collect 100% of worker's    
                                              benefits at age 50              
Ex-Spouse of worker Can collect normal     Can collect normal spousal      
                       spousal benefits if    benefits if marriage lasted     
                       marriage lasted longer longer than 10 years and        
                       than 10 years and      survivor is unmarried           
                       dependent is unmarried                                 
Parent of worker    No special provision   Can collect 75% or 82.5% (if    
                                              only one parent is entitled) of 
                                              worker's benefits if parent is  
                                              age 62 or above, and dependent  
                                              on worker for over half of      
                                              income                          

Source: GAO.

Note: All benefits, except those for divorcees, are subject to the family
maximum. A spouse can collect benefits on his own earnings record, if this
amount is greater than the corresponding dependent benefits.

Calculating Survivors' Benefits

Widow(er)s may be eligible to receive a one-time death benefit of $255. In
addition, widow(er)s, surviving parents, children under the age of 18 (19
if the child is still in school) and disabled adult children can collect
benefits off of the deceased person's earnings record. A widow(er) at full
retirement age will receive 100 percent of his or her spouse's benefits,
unless his or her own benefit is higher. Younger widow(er)s (those between
age 60 and the full retirement age) can receive between 71 and 99 percent
of their deceased spouses' benefits depending on how close they are to the
full retirement age.7 Furthermore, regardless of age, a widow(er) with
young children, can receive 75 percent of the deceased spouse's benefit.
Surviving parents and children can also collect up to 75 percent of their
deceased family members' benefits. All of these benefits receive annual
COLA adjustments and are subject to the family maximum.

7Widow(er)s who begin collecting benefits at age 60 receive 71 percent of
their deceased spouses' benefit amount, and this percentage increases for
every year they delay collecting survivors benefit, reaching 100 percent
at the FRA.

Appendix III: Comments from the Social Security Administration

Related GAO Products

Retirement Security: Women Face Challenges in Ensuring Financial Security
in Retirement. [65]GAO-08-105 . Washington, D.C.: October 11, 2007.

Retirement Decisions: Federal Policies Offer Mixed Signals about When to
Retire. [66]GAO-07-753 . Washington, D.C.: July 11, 2007.

Social Security Reform: Implications of Different Indexing Choice.
[67]GAO-06-804 . Washington, D.C.: Sept ember 14, 2006.

Social Security: Societal Changes Add Challenges to Program Protections.
[68]GAO-05-706T . Washington, D.C.: May 17, 2005.

Options for Social Security Reform. [69]GAO-05-649R . Washington, D.C.:
May 6, 2005.

Social Security Reform: Answers to Key Questions. [70]GAO-05-193SP .
Washington, D.C.: May 2, 2005.

Social Security Reform: Early Action Would Be Prudent. [71]GAO-05-397T .
Washington, D.C.: March 9, 2005.

Long Term Fiscal Issues: The Need for Social Security Reform.
[72]GAO-05-318T . Washington, D.C. February 9, 2005.

Social Security: Distribution of Benefits and Taxes Relative to Earnings
Level. [73]GAO-04-747 . Washington, D.C.: June 15, 2004.

Social Security: Program's Role in Helping Ensure Income Adequacy.
[74]GAO-02-62 . Washington, D.C.: November 30, 2001.

Social Security Reform: Potential Effects on SSA's Disability Programs and
Beneficiaries. [75]GAO-01-35 . Washington, D.C.: January 24, 2001.

Social Security: Evaluating Reform Proposals. [76]GAO/AIMD/HEHS-00-29 .
Washington, D.C.: November 4, 1999.

Social Security Reform: Implications of Raising the Retirement Age.
GAO/HEHS-99-112. Washington, D.C.: August 27, 1999.

Social Security: Issues in Comparing Rates of Return with Market
Investments. [77]GAO/HEHS-99-110 . Washington, D.C.: August 5, 1999.

Social Security: Criteria for Evaluating Social Security Reform Proposals.
GA0/T-HEHS-99-94. Washington, D.C.: March 25, 1999.

(130615)

GAO's Mission

The Government Accountability Office, the audit, evaluation, and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( [78]www.gao.gov ). Each weekday, GAO posts
newly released reports, testimony, and correspondence on its Web site. To
have GAO e-mail you a list of newly posted products every afternoon, go to
[79]www.gao.gov and select "E-mail Updates."

Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
DC 20548

To order by Phone: Voice: (202) 512-6000
TDD: (202) 512-2537
Fax: (202) 512-6061

To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

Web site: [80]www.gao.gov/fraudnet/fraudnet.htm
E-mail: [81][email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional Relations

Gloria Jarmon, Managing Director, [82][email protected] , (202) 512-4400
U.S. Government Accountability Office, 441 G Street NW, Room 7125
Washington, DC 20548

Public Affairs

Susan Becker, Acting Manager, [83][email protected] , (202) 512-4800 U.S.
Government Accountability Office, 441 G Street NW, Room 7149 Washington,
DC 20548

To view the full product, including the scope
and methodology, click on [84]GAO-08-26 .

For more information, contact Barbara Bovbjerg at (202) 512-7215 or
[email protected].

Highlights of [85]GAO-08-26 , a report to congressional requesters

October 2007

SOCIAL SECURITY REFORM

Issues for Disability and Dependent Benefits

Many recent Social Security reform proposals to improve program solvency
include elements that would reduce benefits currently scheduled for future
recipients. To date, debate has focused primarily on the potential impact
on retirees, with less attention to the effects on other Social Security
recipients, such as disabled workers and dependents. As these
beneficiaries may have fewer alternative sources of income than
traditional retirees, there has been interest in considering various
options to protect the benefits of disabled workers and certain
dependents.

This report examines (1) how certain elements of Social Security reform
proposals could affect disability and dependent benefits, (2) options for
protecting these benefits and how they might affect disabled workers and
dependents, and (3) how protecting benefits could affect the Social
Security program. To conduct this study, GAO used a microsimulation model
to simulate benefits under various reform scenarios. GAO also interviewed
experts and reviewed various reform plans, current literature, and GAO's
past work.

[86]What GAO Recommends

Congress should consider the potential implications of reform on
disability and dependent beneficiaries.

GAO received general and technical comments from SSA and the Department of
the Treasury, which were incorporated as appropriate.

We considered several reform elements that could improve Social Security
Trust Fund solvency by reducing the initial benefits received or the
growth of individual benefits over time. According to our simulations,
these reform elements would reduce median lifetime benefits for disabled
workers by up to 27 percent (see graph) and dependents by up to 30 percent
of currently scheduled levels. While the size of the benefit reduction
could vary across individuals, it could be substantial for the vast
majority of these beneficiaries, depending upon the reform element.

Simulated Median Lifetime Benefits for Disabled Workers (1985 Cohort)

Note: Currently scheduled benefits are not attainable under current
funding levels.

Options for protecting the benefits of disabled workers and dependents
from the impact of reform elements include, among others, a partial
exemption, whereby currently scheduled benefits are maintained until
retirement age. For example, while simulations showed that one reform
element could decrease median lifetime benefits of disabled workers to
about 89 percent of currently scheduled levels, a partial exemption could
restore them to about 96 percent. Further, these protections could be more
targeted. For example, a larger cost of living adjustment would result in
more rapid benefit growth for those disabled workers who receive benefits
for a prolonged period of time. Some protections for dependent benefits
could be targeted to a single group of dependents, such as widows, while
others could affect multiple groups. For example, increasing the maximum
benefit a family can receive could protect a wider group of beneficiaries,
including children and spouses of disabled workers, and disabled adult
children.

While it may be desirable to protect the benefits of disabled workers and
certain dependents, such protections would come at a cost to Social
Security. Protecting benefits could lessen the impact that a reform
element would have on solvency. In addition, such protections could create
incentives to apply for Disability Insurance, if disability benefits
remained stable while retirement benefits were reduced.

References

Visible links
  52. http://www.ssa.gov/policy/docs/quickfacts/stat_snapshot/
  53. http://www.gao.gov/cgi-bin/getrpt?GAO-07-310
  54. http://www.gao.gov/cgi-bin/getrpt?GAO-08-105
  55. http://www.ssa.gov/policy/docs/quickfacts/stat_snapshot/
  56. http://www.gao.gov/
  57. http://www.gao.gov/cgi-bin/getrpt?GAO-08-105
  58. http://www.gao.gov/cgi-bin/getrpt?GAO-06-804
  59. http://www.gao.gov/cgi-bin/getrpt?GAO-04-747
  60. http://www.gao.gov/cgi-bin/getrpt?GAO/T-HEHS-99-94
  61. http://www.gao.gov/cgi-bin/getrpt?GAO/AIMD/HEHS-00-29
  62. http://www.gao.gov/cgi-bin/getrpt?GAO/HEHS-99-110
  63. http://www.gao.gov/cgi-bin/getrpt?GAO-02-62
  64. http://www.ssa.gov/OACT/solvency/PresComm_20020131.pdf
  65. http://www.gao.gov/cgi-bin/getrpt?GAO-08-105
  66. http://www.gao.gov/cgi-bin/getrpt?GAO-07-753
  67. http://www.gao.gov/cgi-bin/getrpt?GAO-06-804
  68. http://www.gao.gov/cgi-bin/getrpt?GAO-05-706T
  69. http://www.gao.gov/cgi-bin/getrpt?GAO-05-649R
  70. http://www.gao.gov/cgi-bin/getrpt?GAO-05-193SP
  71. http://www.gao.gov/cgi-bin/getrpt?GAO-05-397T
  72. http://www.gao.gov/cgi-bin/getrpt?GAO-05-318T
  73. http://www.gao.gov/cgi-bin/getrpt?GAO-04-747
  74. http://www.gao.gov/cgi-bin/getrpt?GAO-02-62
  75. http://www.gao.gov/cgi-bin/getrpt?GAO-01-35
  76. http://www.gao.gov/cgi-bin/getrpt?GAO/AIMD/HEHS-00-29
  77. http://www.gao.gov/cgi-bin/getrpt?GAO/HEHS-99-110
  78. http://www.gao.gov/
  79. http://www.gao.gov/
  80. http://www.gao.gov/fraudnet/fraudnet.htm
  81. mailto:[email protected]
  82. mailto:[email protected]
  83. mailto:[email protected]
  84. http://www.gao.gov/cgi-bin/getrpt?GAO-08-26
  85. http://www.gao.gov/cgi-bin/getrpt?GAO-08-26
*** End of document. ***