Social Security: Issues Involving Benefit Equity for Working Women
(Chapter Report, 04/10/96, GAO/HEHS-96-55).

Pursuant to a congressional request, GAO examined whether married women
and widows were being treated fairly under the Social Security system,
focusing on: (1) how economic and demographic trends may affect women's
future benefits; and (2) past proposals that have addressed these
concerns.

GAO noted that: (1) the Social Security Act is gender neutral when it
comes to awarding individual retirement benefits; (2) women are more
likely to qualify for spousal and survivor benefits and experience dual
entitlement limits because of their lower lifetime earnings and longer
life expectancies; (3) the monthly benefits for a two-earner couple are
one-third less than those of a one-earner couple with the same
demographics and lifetime earnings; (4) the survivor of a one-earner
couple receives 67 percent of the total benefits received when both
spouses were alive; (5) the survivor of a two-earner couple can receive
between 50 percent and 67 percent of combined pre-death benefits; (6) a
retired married person is often eligible for larger benefits as a spouse
or survivor than as a retired worker; (7) an individual's retirement
contributions do not necessarily increase his retirement benefits, but
may provide disability coverage and family survivor coverage; (8) people
with lower lifetime earnings and fewer Social Security contributions
receive more Social Security benefits; (9) the issue of benefit fairness
is likely to become more widespread as women spend more time in the
workforce and increase their earnings; (10) the earnings sharing
proposal combines the covered earnings of a husband and wife for each
year that the couple is married and credits half of the combined total
to each spouse's Social Security earnings; (11) the double-decker plan
provides a flat-rate benefit for qualified beneficiaries and a
proportional plan based on beneficiaries' lifetime covered earnings; and
(12) SSA has not changed the way it administers retirement benefits
because of the potential program costs, inequity issues, and
administrative burdens.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-96-55
     TITLE:  Social Security: Issues Involving Benefit Equity for 
             Working Women
      DATE:  04/10/96
   SUBJECT:  Women
             Widowed persons
             Social security benefits
             Employee survivors benefits
             Retirement benefits
             Service credit
             Income statistics
             Employee retirement plans
             Income maintenance programs
IDENTIFIER:  Social Security Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed 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.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to the Honorable
William V.  Roth, Jr., U.S.  Senate

April 1996

SOCIAL SECURITY - ISSUES INVOLVING
BENEFIT EQUITY FOR WORKING WOMEN

GAO/HEHS-96-55

Benefit Equity for Working Women

(105678)


Abbreviations
=============================================================== ABBREV

  AIME - average indexed monthly earnings
  BLS - Bureau of Labor Statistics
  COLA - cost of living adjustment
  DRC - delayed retirement credit
  PIA - primary insurance amount
  SSA - Social Security Administration
  SSI - Supplemental Security Income

Letter
=============================================================== LETTER


B-259611

April 10, 1996

The Honorable William V.  Roth, Jr.
United States Senate

Dear Senator Roth: 

This report responds to your request that we review concerns about
the treatment of married women and widows under the Social Security
system, proposals that have been made to address these concerns, and
the likely impact of recent and projected economic and demographic
trends on women's future benefits. 

As agreed with your office, we are sending copies of this report to
the Commissioner of Social Security.  Copies will be sent to others
on request. 

This report was prepared under my direction.  Please contact Michael
D.  Packard, Senior Economist, at (202) 512-7250 if you have any
questions. 

Sincerely yours,

Jane L.  Ross
Director, Income Security Issues


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

Social Security is a large and complex program that protects most
workers and their families from income loss because of a worker's
retirement, disability, or death.  In 1994, about 95 percent of all
jobs in the United States were covered by Social Security, and almost
43 million people received $317 billion in benefits from the program. 

In the 1930s, when the program was established, less than 15 percent
of married women worked for pay outside of their homes, and, as a
result, most married couples were one-earner couples.  Today, about
60 percent of married women are paid workers, and most married
couples are two-earner couples.  However, Social Security's benefit
structure remains largely unchanged despite this and other
demographic changes, leading some to question the structure's
fairness. 

To better understand this matter, Senator William V.  Roth, Jr.,
asked GAO to (1) describe concerns about the fairness of the benefits
received by wage-earning women and their families in retirement, (2)
discuss how economic and demographic trends might affect the benefits
of such people in the future, and (3) discuss past proposals to
address these concerns. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

Policymakers in 1935 based the Social Security Act's (the act)
original benefit structure on the principle that only those who
contributed to Social Security should receive benefits and individual
benefit levels should relate to individual earnings levels.  This may
be termed the "individual equity" principle. 

In 1939, policymakers decided the program should also provide
benefits to the families of covered workers, and the Congress amended
the act to provide benefits to covered workers' wives, widows,
eligible children, and, in a few instances, parents.\1

These benefits are provided without additional contributions from the
worker.  By reflecting the notion that the worker and his family were
the unit to be protected, the 1939 amendments incorporated a "social
adequacy" rationale into the benefit structure. 

The 1939 expansion of benefit categories resulted in some people
being entitled to benefits both as a retired worker and as the spouse
or survivor of another worker.  These individuals are said to be
"dually entitled" to benefits.  However, dually entitled
beneficiaries do not receive the sum of these two benefits.  Rather,
they receive (1) the retired worker benefit they earned plus (2) the
difference, if any, between that benefit and the benefit they would
receive simply as a spouse or survivor.  In other words, because of
the dual entitlement provision (or dual entitlement limitation),
married (and widowed) beneficiaries receive a benefit equal, in
effect, to either their own earned retirement benefit or their
spousal (or survivor) benefit, whichever is higher. 


--------------------
\1 The wife's benefit is currently based on 50 percent of the
worker's benefit and the widow's on 100 percent (see app.  I).  These
benefits were made available to husbands and widowers in 1950. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

The provision of spousal and survivor benefits on social adequacy
grounds and the workings of the dual entitlement limitation laid the
groundwork for concerns about the fairness of the Social Security
benefit structure.  Some question whether benefits are equitably
distributed among different categories of beneficiaries and
contributors to the program.  For example, a two-earner couple will
receive lower combined benefits in retirement than an otherwise
identical one-earner couple.  And, a married woman who works and pays
Social Security taxes might not, because of the dual entitlement
limitation, receive higher benefits than if she had never worked and
received only a spousal benefit. 

The movement of women into the paid labor market is likely to
increase the concerns about the fairness of the Social Security
benefit structure.  As women's labor force participation continues to
grow, increasing numbers of women will earn benefits in their own
right.  This, in turn, will increase the number of two-earner
couples, heightening concerns about the fairness of their benefits
(and those of their survivors) vis-a-vis those of one-earner couples
with identical lifetime earnings.  It will also increase the
proportion of married women subject to the dual entitlement
limitation and may lead more women to question whether they receive
an adequate return on their contributions. 

Several proposals have been made over the years to address these
equity issues.  These include two broad proposals--"earnings sharing"
and a "double-decker" plan--and several more narrow proposals, such
as reducing spousal benefits.  However, none of the proposals has
been adopted, in part because they would either increase program
costs or reduce benefit amounts for some groups of beneficiaries. 
Their enactment could also impose a large administrative burden on
the Social Security Administration (SSA). 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4

The Social Security Act's benefits provisions are "gender neutral" in
that individuals, identical in all respects aside from gender, will
receive identical earned benefits from the program.  However, because
spousal and survivor benefits are not earned benefits, these
provisions result in benefit outcomes that raise certain fairness
concerns.  Because women tend to have lower lifetime earnings and
longer life expectancies than men, they are more likely than men to
qualify for spousal and survivor benefits and, thus, to be affected
by the dual entitlement limitation.  As a result, these concerns are
generally perceived as women's issues. 


      IMPLICATIONS OF THE BENEFIT
      PROVISIONS ON INDIVIDUAL
      EQUITY
-------------------------------------------------------- Chapter 0:4.1

The main equity concerns about the treatment of working women and
their families result from the characteristics of the spousal and
survivor benefits and how benefits are affected by the act's dual
entitlement limitation.  First, a two-earner couple receives a lower
monthly benefit in retirement than a one-earner couple with identical
lifetime earnings (and other relevant characteristics, such as age). 
Each couple receives the full retired worker benefits it earned.\2 In
addition, the one-earner couple receives a spousal benefit based on
the lifetime earnings of the worker.  The spousal benefit for the
two-earner couple, smaller than that for the one-earner couple
because it is based on the smaller lifetime earnings of the primary
earner, is further reduced, and often eliminated, by the dual
entitlement limitation.  As a result, the monthly benefits for a
two-earner couple can be as much as one-third less than those for a
one-earner couple having identical demographic characteristics and
identical lifetime earnings. 

Second, the disparity in monthly benefits can be even greater for the
survivors of these couples.  A survivor is entitled to a benefit that
is essentially equal to the larger of the pre-death retired worker
benefits received by the couple.  The survivor of a one-earner couple
receives 67 percent of the total benefit received when both were
alive.\3 The survivor of a two-earner couple, however, can receive as
much as 67 percent or as little as 50 percent of the couple's
combined pre-death benefits.\4 As a result, the survivor of a
two-earner couple can receive a benefit that is as little as one-half
that received by the survivor of a one-earner couple with identical
lifetime earnings. 

Third, a retired married person (usually the wife) might be eligible
for larger benefits as a spouse or survivor than as a retired worker
even though she (or he) made contributions to Social Security for
many years.  Because of the dual entitlement limitation, these
contributions do not always result in an increase in monthly
benefits.  This raises questions of whether married women receive a
fair return on the contributions they make.  Even though these
contributions may not increase their retirement benefits, they can
provide the worker with disability coverage and the worker's family
with survivor coverage--coverages not provided to those who make no
contributions. 

Finally, spousal and survivor benefits may provide some people with
higher benefits than other people who had higher lifetime earnings
and who made higher contributions to Social Security. 


--------------------
\2 The benefit formula is progressive (see app.  I) and a two-earner
couple usually has a larger portion of its average indexed monthly
earnings (AIME) replaced at the 90-percent tier of the benefit
formula and, therefore, receives larger combined retired worker
benefits than an otherwise identical one-earner couple. 

\3 This assumes both the husband and wife were aged 65 when they
first received benefits.  If not, the survivor could receive a
different percentage of the couple's pre-death benefit. 

\4 The larger reduction occurs when each partner receives equal
pre-death retired worker benefits, again assuming both were aged 65
at first benefit receipt.  At the death of one of the partners, one
of those benefit streams is eliminated, resulting in a 50-percent
reduction from the pre-death combined benefits.  The smaller
reduction occurs when the secondary earner is dually entitled. 


      TRENDS
-------------------------------------------------------- Chapter 0:4.2

Most of these concerns about benefit fairness are likely to become
more widespread in the future because of two trends.  First, women
are spending an increasing portion of their lives in the paid labor
force.  The labor force participation rates for women have doubled
since 1940, and those of working-aged women (those aged 25 to 54) are
expected to continue to increase in the future.  This stronger labor
force participation will result in more women receiving benefits
based, at least in part, on their own work records and may exacerbate
concerns about the return married women receive on their
contributions.  It will also increase the proportion of couples who
are two-earner couples, which may increase the concerns that
two-earner couples (and their survivors) receive lower benefits than
identical one-earner couples. 

Second, the earnings of women are rising relative to those of men. 
Because of the dual entitlement limitation, these higher lifetime
earnings will increase the proportion of married women whose earnings
will be too high for them to qualify for any spousal benefits, which
may reduce the concerns about returns on contributions.  However, it
may increase concerns that people who had lower lifetime earnings and
made smaller contributions to Social Security could receive higher
benefits from the program. 


      PROPOSALS FOR REFORM
-------------------------------------------------------- Chapter 0:4.3

Several proposals have been made during the past 25 years to address
individual equity concerns.  Two of the broader proposals--"earnings
sharing" and a "double-decker" plan--would address several of these
concerns at one time.  Earnings sharing would combine the covered
earnings\5 of the husband and wife for each year the couple was
married and credit half of this combined total to each spouse's
Social Security earnings record.  There would be no separate spousal
or survivor benefits, but, under most variations of earnings sharing,
survivors would inherit "earnings credits" from deceased partners for
the years of marriage. 

Because benefits would be based only on each spouse's shared earnings
credits, couples with the same total lifetime earnings would receive
equal retirement benefits, as would their survivors.\6 In addition,
because there would be no spousal benefits, a married woman would not
face the prospect of working and paying Social Security taxes only to
find that she is entitled to a spousal benefit larger than the
benefit she earned in her own right.\7

The double-decker plan, which is used in many countries, would also
equalize benefits in situations in which couples' lifetime earnings
were the same.  Each person who works in covered employment would,
upon retirement, receive benefits from two "decks." The first deck
would provide a flat-rate benefit to all qualified beneficiaries,
while the second deck would provide benefits proportional to the
beneficiary's lifetime covered earnings.  Spousal and survivor
benefits would be eliminated, but, under some versions of the plan,
survivors would inherit earnings credits from deceased partners. 

None of these proposals, broad or narrow, has been adopted.  One
obstacle has been costs.  Many of the proposals would entail
increasing the benefits for certain categories of beneficiaries
thought to be disadvantaged by the current system (for example,
two-earner couples), which would raise program costs if no offsetting
benefit reductions were made.  Financing such increases in a "cost
neutral" manner--that is, without the need for additional Social
Security payroll taxes--would require corresponding benefit decreases
for other categories of beneficiaries, which could raise certain
concerns about the adequacy of their benefits.  Also, implementing
any proposed reforms could put a large administrative burden on SSA
by forcing it to revise its reporting, recording, and operational
procedures. 


--------------------
\5 Covered earnings are earnings on which Social Security taxes are
paid. 

\6 This assumes that the couples are identical except for the
distribution of total earnings within the couples and that the
survivor of each couple could inherit all of the earnings credits of
his or her deceased partner. 

\7 Earnings sharing would not necessarily equalize benefits of
like-earning individuals who, under current law, have different
monthly benefits, if, for example, their spouses have different
earnings. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

GAO is making no recommendations. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:6

GAO provided a draft of the report to the Commissioner of Social
Security with a request for the agency's comments.  In commenting on
our report, the Commissioner stated that the report "adequately
addresses the objectives of the study." However, she noted a few
concerns. 

First, she indicated the report follows the premise that the authors
of the 1939 amendments did not envision a world in which most wives
would work.  The report has been modified to indicate the authors of
the 1939 amendments did envision such a world.  However, GAO's point
is that the increase in the number of women workers combined with the
current benefit structure has given rise to anomalies that lead to
the concerns discussed. 

Second, the Commissioner suggested GAO discuss what impact certain
proposals to address the equity concerns, such as eliminating spousal
and survivor benefits, would have on adequacy concerns.  Appendix II
covers adequacy concerns in some detail, and chapter 5 notes that
reforming the program to address equity concerns could create or
exacerbate existing adequacy concerns. 

Finally, the Commissioner noted that GAO's use of cross-sectional
data for the analyses in chapter 3 could provide biased estimates of
Social Security lifetime earnings and wealth, and she offered GAO the
use of an alternative data set created by SSA's Office of Research
and Statistics.  GAO did not use the recommended data set in its
analyses.  However, it included some published findings from studies
that used this data set. 


INTRODUCTION
============================================================ Chapter 1

When Social Security was enacted 60 years ago, relatively few women
worked in wage-paying jobs.  As a result, during the first decades of
the program, most women became eligible for benefits as either the
spouse or survivor of their Social Security-covered husbands.  Today,
a larger proportion of women work and earn Social Security benefits
in their own right.  However, the benefit structure has not
significantly changed to reflect this trend, leading to concerns
about its fairness for retired women and their families. 


   DEVELOPMENT OF SOCIAL
   SECURITY'S BENEFIT STRUCTURE
---------------------------------------------------------- Chapter 1:1

The Social Security Act (the act) was signed into law by President
Franklin D.  Roosevelt in August 1935.  It originally was designed to
provide retirement benefits only to people in commerce and industry
occupations.  Workers in these occupations could receive benefits at
age 65 or older if they had no earnings. 

Benefits under the original act had a strong "individual equity"
component--that is, benefits were to go only to retired individuals
who had contributed to Social Security during their working years,
and individual benefit levels were to be based on lifetime earnings. 
Benefits were not strictly proportional to earnings, however, because
there was (and still is) a larger increase in benefits for the first
increment of earnings than for later earnings increments.  Thus,
there was also a "social adequacy" component to the benefit
structure. 

The original act had no provision providing monthly benefits to
workers' dependents or survivors.\8 The first Advisory Council on
Social Security, believing that the family--not simply the
worker--should be the unit protected by the act, concluded in 1938
that this was a shortcoming.  In their view, the social adequacy
goals of the program needed strengthening.  The Council therefore
recommended that benefits be extended to the wives, widows, and
eligible minor children. 

The Congress in 1939 adopted the Council's recommendations and
provided "auxiliary" benefits to wives and widows aged 65 or older,
children under age 18 (but generally only if the covered worker was
male), and some dependent parents of deceased covered workers.\9
Today, spousal benefits are 50 percent of the worker's age 65 benefit
and the survivor benefits are 100 percent.  The auxiliary benefits
were (and are) funded, not by an additional tax on married workers or
a reduction in their own benefits, but from Social Security funds to
which all covered workers contribute.  Thus, auxiliary benefits are
not "earned" benefits. 

When the 1939 Congress provided auxiliary benefits to wives and
widows, it recognized that many married women would also be wage
earners entitled to Social Security retired worker benefits in their
own right.  The Congress did not want to deny such women the
opportunity to receive at least some spousal or survivor benefits
because doing so could result in lowered benefits for working wives. 
The auxiliary benefits of many women, particularly those with
low-paying jobs or few years of work, might be larger than the
benefits the women earned from their own jobs.  At the same time,
however, the Congress did not want working women who were not
dependent on their husbands to receive auxiliary benefits. 

The Congress' solution, seen in the act's dual entitlement provision
(or limitation), was to provide that a working woman in retirement
could receive benefits from two sources--(1) her own retired worker
benefit plus (2) any excess of her spousal or survivor benefit over
her retired worker benefit.  In effect, this meant that she would
receive not the sum of these two benefits but either her retired
worker benefit or her auxiliary benefit, whichever was higher.\10 \11

By providing benefits to people other than the workers themselves,
the 1939 amendments improved the "social adequacy" part of the
program but, in so doing, laid the groundwork for the "individual
equity" concerns.  Subsequent debates about the program have often
been over the appropriate balance between these two, often
conflicting goals.  It is not easy to move the program closer to one
goal without moving it away from the other. 

Since 1939, the act has been amended many times, with the changes
tending to strengthen the social adequacy features of the
program--for example, the 1956 amendments introduced disability
benefits.  Other changes included expanding the types of jobs
covered, adding beneficiary categories (such as benefits for disabled
workers and divorced spouses), increasing benefits, introducing
automatic cost of living adjustments (COLA), and changing the payroll
tax rate and level of earnings subject to the payroll tax.\12

Today, about 95 percent of all jobs in the United States are covered
by Social Security.  Workers in the program, provided they have made
the necessary contributions and met other requirements, may receive
retirement benefits upon reaching age 62 or disability benefits upon
becoming disabled.  In 1994, more than 138 million workers
contributed almost $345 billion to the program, and nearly 43 million
people received $317 billion in benefits. 


--------------------
\8 However, a deceased worker's survivor was eligible for a lump sum
payment equal to 3.5 percent of the worker's lifetime covered
earnings less the lifetime benefits received by the worker. 

\9 In 1950, the Congress made men eligible for spousal and survivor
benefits, albeit with more restrictions than applied to women. 
Today, the requirements to receive such benefits are identical for
men and women.  Only the marriage partner with the lower lifetime
earnings can qualify for auxiliary benefits. 

\10 For example, if a dependent married woman is entitled to a
retired worker benefit of $300 per month based upon her own work
record and a spousal benefit of $400 per month based upon her
husband's work record, she will receive a dual entitlement benefit of
$400--that is, her own retired worker benefit of $300 plus the $100
by which her spousal benefit exceeds her retired worker benefit.  In
effect, she gets the larger of the two, the amount of her spousal
benefit. 

\11 Although dually entitled people receive the same benefit amounts
they would have received had they never worked, their work provides
advantages not enjoyed by spouses and survivors who never worked. 
First, their earnings increase their family's income during the years
they work.  Second, these women are often eligible for disability
benefits.  Third, their families are eligible for survivor benefits. 
Finally, they can receive retired worker benefits at the time they
choose (after attaining age 62) and do not have to wait for their
spouses to retire or die to receive a Social Security benefit. 

\12 Workers whose earnings are subject to the Social Security payroll
tax (also called contributions) are called "covered" workers, and the
earnings on which payroll taxes are levied are known as "covered"
earnings. 


   HOW BENEFITS ARE CALCULATED
---------------------------------------------------------- Chapter 1:2

The calculation of Social Security benefits is essentially a
three-step process (see app.  I for a detailed description).  First,
a worker's lifetime covered earnings over his or her 35 highest
earnings years are calculated.  Social Security uses "average indexed
monthly earnings" (AIME) as its measure of lifetime covered earnings. 

Second, a progressive benefit formula is applied to these lifetime
earnings to determine the benefit that would be payable to the worker
at age 65.\13 This age-65 benefit is called the "primary insurance
amount" (PIA).  Finally, the benefit is adjusted for the age at which
the worker first receives the benefit.\14 If auxiliary benefits are
being calculated, a factor is applied to the PIA for the particular
type of benefit, and the resulting amount is adjusted for the
auxiliary beneficiary's age.\15

The benefits of retired and disabled workers are based on their own
PIAs.  The benefits of auxiliary beneficiaries, such as those
receiving spousal benefits, are normally based on the PIAs of their
current or former spouses.  The benefits for persons entitled to
auxiliary benefits and retired worker benefits in their own
right--"dually entitled" individuals--are based in part on their own
PIAs and in part on those of their current or former spouses. 


--------------------
\13 The benefit formula is progressive in that it replaces a
relatively larger portion of lifetime earnings for people with low
earnings than for people with high earnings. 

\14 Benefits are reduced if first taken before age 65 and increased
if first taken between ages 65 and 70. 

\15 A spousal benefit is 50 percent of the worker's PIA and a
survivor's benefit, 100 percent.  Each will be reduced if first
received by the auxiliary beneficiary before age 65, but not
increased if first taken at age 65 or older. 


   THE BENEFIT STRUCTURE AND WOMEN
---------------------------------------------------------- Chapter 1:3

Today, the act's benefit provisions are gender neutral.  At
retirement, men and women with identical Social Security earnings
histories and other relevant characteristics, such as age and marital
status, will receive identical earned benefits from the program. 
However, because women, on average, have lower lifetime earnings than
men, their average Social Security benefits are smaller.  In December
1994, benefits of female retired workers averaged $601 compared with
$785 for male retired workers.\16 The lower earnings for women
reflected their "weaker attachment" to the workplace in terms of
fewer years with earnings and lower wage rates for their years of
employment.  Also, women's benefits tend to be lower than men's
because women generally begin receiving benefits at younger ages than
men.  The younger the age at retirement, the larger the benefit
reduction. 

Because women tend to have lower earnings than men, their benefits,
relative to their lifetime earnings, are likely to be higher than
men's because of Social Security's progressive benefit formula. 
Also, because women have lower earnings than men, more women than men
receive auxiliary benefits as spouses or survivors,\17 and, thus,
more women will be subject to the dual entitlement limitation on
benefit amounts.  Finally, women have longer life expectancies than
men and will, therefore, receive Social Security benefits for a
longer period of time than men, on average.\18 As a result, the ratio
of the total benefits women and men receive over their lifetimes will
be higher, on average, than the ratios of their monthly benefits. 


--------------------
\16 Averages include the portion of spousal or survivor benefits
received by dually entitled persons. 

\17 For example, in December 1994, more than 3 million women, but
fewer than 40,000 men, were receiving spousal benefits. 

\18 A woman aged 65 today can expect to live another 19 years, but a
65-year-old man can expect to live only about 15 more years. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:4

To better understand the issues involved, Senator William V.  Roth,
Jr., asked us to (1) describe the concerns about the fairness of the
benefits received by job-holding women and their families in
retirement, (2) discuss how economic and demographic trends might
affect the benefits of such people in the future, and (3) discuss
past proposals to address these concerns. 

To address these objectives, we reviewed current and historical
Social Security documents to understand the rationale for the current
Social Security benefit structure; interviewed experts and
researchers who have studied Social Security's treatment of women;
reviewed the literature that described concerns about Social
Security's treatment of women during the past 25 years and that
proposed solutions to alleviate these concerns; and obtained
historical evidence and available projections of the changing labor
force participation and earnings levels of women. 

Most of the data we used to analyze economic and demographic trends
were published or provided by the Social Security Administration
(SSA), the Department of Labor's Bureau of Labor Statistics and
Women's Bureau, and the Bureau of the Census.  Because of a lack of
suitable and readily available longitudinal data to analyze these
trends, we relied heavily on cross-sectional data for our analysis. 
Using these data, we determined the likely effect of economic and
demographic trends on the Social Security benefits women will receive
in the future.  We conducted our work from November 1994 to December
1995 using generally accepted government auditing standards. 


      SOCIAL ADEQUACY ISSUES
-------------------------------------------------------- Chapter 1:4.1

In examining concerns about the fairness of the benefits structure
for working women and their families, our focus is on working women's
(and, by extension, couples') earnings and how they affect, or do not
affect, retirement benefits.  Insofar as the focus is on the link
between earnings and benefits, the report addresses mainly
"individual equity" issues. 

There are, as well, "social adequacy" concerns pertaining to the
Social Security benefits of women--for example, whether benefits for
aged widows should be increased or whether the program's treatment of
divorced women is fair.  Although these concerns were outside the
scope of this study, we have summarized some of the more common ones
in appendix II. 


IMPLICATIONS OF BENEFIT PROVISIONS
ON INDIVIDUAL EQUITY
============================================================ Chapter 2

Concerns about the fairness of the act's benefit provisions for women
and their families arise from two sources--the unearned nature of the
act's spousal and survivor benefit provisions and its dual
entitlement limitation.  Because women tend to have lower lifetime
earnings and longer life expectancies than men, they are more likely
than men to qualify for spousal and survivor benefits and to be
affected by the dual entitlement limitation.  As a result, the
following fairness concerns are generally viewed as women's issues,
although the concerns can apply to men as well. 


   COUPLE BENEFITS
---------------------------------------------------------- Chapter 2:1

One fairness concern is that a two-earner couple receives lower
combined retirement benefits than a one-earner couple with identical
lifetime earnings and other identical characteristics.\19 The reason
this occurs is that the two-earner couple will receive a smaller
spousal benefit than the one-earner couple and, in fact, may receive
no spousal benefit at all. 

Both couples receive all the retired worker benefit each partner
earned through his or her covered employment.  Because the benefit
formula is progressive, the two-earner couple will usually receive
larger total retired worker benefits than the one-earner couple. 

The one-earner couple also receives a spousal benefit (usually for
the wife) based on one-half the worker's benefit.  This spousal
benefit is not reduced by the dual entitlement limitation because the
spouse receives no retired worker benefits of her own.  Thus, the
one-earner couple's total benefit is 150 percent of the worker's
retired worker benefit (if both first received benefits at age 65). 

The unreduced spousal benefit for the two-earner couple will be
smaller than that for the one-earner couple because it is based on
the benefit of the couple's primary earner, whose lifetime earnings
are smaller than those of the worker in the one-earner couple.  The
dual entitlement limitation will further reduce, or perhaps
eliminate, this smaller spousal benefit by subtracting the secondary
worker's retired worker benefit from it.  The total benefit of the
two-earner couple will consist of the two retired worker benefits
plus, perhaps, a spousal benefit supplement.  The difference in
spousal benefits always exceeds the difference in total retired
worker benefits, so a two-earner couple always receives lower
benefits than an otherwise identical one-earner couple. 

A hypothetical example can illustrate how this occurs.  We assume
there are two couples who begin receiving benefits in 1995 and that
all four persons are aged 65.  We assume that the husband in the
one-earner couple has an AIME of $2,000 and that his wife had no
covered earnings.\20 In the two-earner couple, we assume each spouse
has an equal AIME of $1,000, for a total of $2,000. 

In this example, the worker in the one-earner couple would receive a
retired worker benefit of $939 per month and his wife, an unreduced
spousal benefit of $469 per month for a total benefit of $1,408. 
Each worker in the two-earner couple would receive a monthly retired
worker benefit of $591.  Neither would be eligible for a spousal
benefit because of the dual entitlement limitation.  Their total
benefit would be $1,182, $243 more than the retired worker benefit
received by the one-earner couple but $226 less than the one-earner
couple's total benefit. 

The total monthly benefit of the two-earner couple in this example is
about 16 percent less than that of the one-earner couple.  The
reduction can be greater or smaller depending on the relative
earnings of the husband and wife in the two-earner couple and the
couples' total AIME.  In some cases, a two-earner couple can receive
total benefits that are one-third less than an otherwise identical
one-earner couple's benefits. 

Figure 2.1 shows how varying distributions of earnings would affect
the combined benefits of couples with a combined AIME of $2,000
(assuming that benefits always start at age 65 in 1995).  At the left
end of the line, the wife's share of the couple's AIME is 0 percent
(the one-earner couple in our example), and the couple's total
monthly benefit is $1,408.  At the right end, the wife's share is 50
percent (the two-earner couple in the example), and the couple's
total monthly benefit is only $1,182.  The connecting line shows the
monthly benefit totals for two-earner couples with various
within-couple distributions of the $2,000 AIME. 

   Figure 2.1:  Total Couple
   Benefits for 65-Year-Old
   Couples With Various
   Within-Couple Distributions of
   a $2,000 AIME

   (See figure in printed
   edition.)

The declining segment of the line represents couples whose secondary
earner (the wife) is dually entitled to retired worker benefits and
spousal benefits.  The line declines because, as the wife's share of
the couple's AIME increases, the husband's share falls.  This reduces
both the husband's retired worker benefit and the spousal benefit
based on it.  Until the wife's retired worker benefit exceeds her
benefit as a spouse, the couples' total benefits will fall.  Once
both husband and wife are receiving retired worker benefits, the
couples' total benefit will either increase (if the wife's AIME puts
her on a higher tier of the benefit formula than her husband [see
app.  I]) or remain constant (if both partners are on the same
tier).\21


--------------------
\19 The benefit formula used to calculate benefits is adjusted each
calendar year and applies to all people attaining age 62 in that
year.  Also, benefits are adjusted depending on the age at which
benefits are first received.  Thus, assuming that these other
characteristics of couples are identical, especially the years of
birth and the ages at first benefit receipt, is important. 

\20 Recall that AIME stands for "average indexed monthly earnings,"
the measure of lifetime earnings SSA uses to calculate a person's
Social Security benefit. 

\21 For 65-year-old couples in 1995, the line will have a rising
segment when a couple's combined AIME level is less than about
$1,860. 


   SURVIVOR BENEFITS
---------------------------------------------------------- Chapter 2:2

A second concern is that the disparity in the monthly benefit
received by one-earner and two-earner couples with identical lifetime
earnings can be even greater for the survivors of these couples.  The
survivor of a couple is entitled to the higher of the two retired
worker benefits earned by the couple before the death of one
partner.\22 Because women tend to outlive their husbands, they are
the ones most often affected. 

The surviving member of a one-earner couple receives about 67 percent
of the couple's total monthly benefit when both were alive.\23 In
essence, the retired worker benefit continues, and the spousal
benefit ends. 

The surviving member of a two-earner couple can receive as much as 67
percent of the couple's total monthly benefit (if the secondary
earner was dually entitled) or as little as 50 percent (if each
earner had identical lifetime earnings).  Because two-earner couples
have lower couple benefits than identical one-earner couples, the
survivors of two-earner couples always receive lower survivor
benefits.  These survivors receive only the higher of the two retired
worker benefits the couple earned.  The retired worker benefit of the
secondary earner and any spousal supplement he or she might have
received as a dually entitled beneficiary are eliminated.  As a
result, the survivor of a two-earner couple can receive benefits that
are as low as one-half those received by the survivor of an identical
one-earner couple. 

Continuing our hypothetical example, the survivor of the one-earner
couple would receive $939 (two-thirds of the couple's $1,408 monthly
benefit) as shown in figure 2.2.  The survivor of the two-earner
couple in which the husband and wife had identical lifetime earnings
would receive only $591 (half the $1,182 the couple received while
both were alive).  The benefits of the survivor of our two-earner
couple are 37 percent less than those of the one-earner couple
survivor, even though both couples had identical lifetime earnings. 

   Figure 2.2:  Couple and
   Survivor Benefits for Different
   Within-Couple Distributions of
   the Couples' Total $2,000 AIME

   (See figure in printed
   edition.)

As seen in the figure, the survivor benefit of a two-earner couple
whose secondary earner was dually entitled while both partners were
alive will receive a benefit one-third smaller than the couple's
total benefit while both were alive.  Once the secondary earner is no
longer dually entitled, the percentage reduction will increase
because the secondary earner's retired worker benefit, which now
exceeds the spousal benefit based on the primary earner's benefit, is
eliminated. 


--------------------
\22 In a one-earner couple, the spouse has a retired worker benefit
of $0. 

\23 The exact percentage depends on the ages of the husband and wife
when benefits were first received.  In our example, the survivor
would receive two-thirds of the couple's benefit. 


   BENEFITS FOR MARRIED WORKING
   WOMEN
---------------------------------------------------------- Chapter 2:3

A third concern is that a married person (usually the wife) may work
and pay Social Security taxes only to discover at retirement that she
is entitled to a spousal benefit that is larger than the retired
worker benefit she earned in her own right.  Because the act's dual
entitlement limitation effectively limits a beneficiary to the larger
of these benefits, her contributions do not result in a larger
benefit at retirement.  It is sometimes argued that such a limitation
is unfair because it does not sufficiently reward a person for years
of contributions.\24

The first bar in figure 2.3 ("Working Wife") illustrates this point. 
In this figure, we assume that the wife had earned a retired worker
benefit of $250.  Her husband, however, earned a benefit of $1,000,
making her eligible for a $500 spousal benefit.  She receives a
dually entitled benefit of $500--her own $250 retired worker benefit
and the $250 difference between it and her full spousal benefit.  Had
she never worked, she still would have been eligible for a spousal
benefit of $500 (the second bar in the figure ["Nonworking Wife"]). 

   Figure 2.3:  Social Security
   Benefits of Two Married Women

   (See figure in printed
   edition.)


--------------------
\24 We note, however, that, while the dually entitled beneficiary's
benefit is not increased as a result of her contributions to the
program, her benefit is larger than the benefit that would be
received by a beneficiary who had identical lifetime earnings but who
is not dually entitled. 


   BENEFITS FOR WORKING AND
   NONWORKING INDIVIDUALS
---------------------------------------------------------- Chapter 2:4

Related concerns have also been raised about the relative benefits
that can be received by those who worked and contributed to the
program versus those who receive benefits solely as spouses or
survivors and about the benefits that can be received by those who
are dually entitled versus those who are not.  The first issue is
that workers, in spite of the taxes they pay to Social Security, may
not receive higher benefits in retirement than nonworkers who paid no
such taxes.  The second is that some workers may receive benefits
that are smaller than those received by people having lower lifetime
earnings and making lower contributions to the program.  These
concerns are most apparent when comparing benefits received by single
workers with those received by the lower earner in a married couple. 

These issues are illustrated in figure 2.4.  "Person 1" has a monthly
retired worker benefit of $500 and is not dually entitled.  "Person
2" has lower lifetime earnings and a lower retired worker benefit
($400 per month) but is a dually entitled spouse and receives a total
monthly benefit of $550.  "Person 3" never worked and receives no
retired worker benefit, but her spouse was a high earner.  Her
spousal benefit is $575 per month. 

   Figure 2.4:  Illustrative
   Benefits for Retired Worker
   Only, Dually Entitled, and
   Spouse Only Beneficiaries

   (See figure in printed
   edition.)

Persons 1 and 3 illustrate the first concern, that workers may
receive smaller benefits than nonworkers.  For this to occur, the
covered earnings of the nonworker's spouse must be more than twice as
large as the worker's covered earnings. 

Persons 1 and 2 illustrate the second concern.  Person 1 had larger
lifetime earnings and a larger retired worker benefit than person 2. 
However, he had a smaller total benefit because person 2 had a
relatively high-earning spouse.  As an extreme example, in December
1994, several thousand dually entitled spouses had monthly benefits
in excess of $900 per month.  This is a larger monthly benefit than
was received by more than three-quarters of all retired worker
beneficiaries in that month. 

These concerns can intensify when a couple's primary earner dies. 
The survivor then becomes eligible for the deceased's full retired
worker benefit.  Widows who were the couples' secondary earners are
even more likely than married spouse beneficiaries to have larger
benefits than workers who contributed more to the program than the
widows did. 


TRENDS IN WOMEN'S SOCIAL SECURITY
BENEFITS
============================================================ Chapter 3

The recent and projected trends in women's labor force participation
and earning levels are likely to heighten the concerns discussed in
chapter 2.  The type and amount of Social Security benefits women
receive depend upon several factors, including their "attachment" to
the labor force, their lifetime earnings (or AIMEs) relative to those
of current or former husbands, and their marital history. 

Women's labor force participation has been increasing, which should
increase the proportion of women qualifying for retired worker
benefits in the future.  This, in turn, will increase the proportion
of couples who are two-earner couples.  The earnings of women
relative to men's have also been increasing, which should increase
the proportion of women receiving only retired worker benefits (as
opposed to being dually entitled).  Finally, changes in marital
status have affected the types of benefits these women can receive. 


   WOMEN'S LABOR FORCE
   PARTICIPATION
---------------------------------------------------------- Chapter 3:1

There has been a long-term trend toward greater labor market
participation by women.  This is seen in table 3.1.  In 1940, only 28
percent of all women were in the labor force, and less than 15
percent of married women were working.\25 By 1993, almost 60 percent
of all women were in the labor force and married women were slightly
more likely than other women to be working. 

The growth in women's labor force participation is even more dramatic
for women in their prime earning years--ages 25 to 54.  The labor
force participation rate for these women increased from 42 percent in
1960 to 75 percent in 1993.  The rates for married women within this
age range increased even more rapidly. 



                         Table 3.1
          
          Women's Labor Force Participation Rates,
                          1940-93


               Aged 16    Aged 25-     Aged 16    Aged 25-
Year         and older        54\b   and older        54\b
----------  ----------  ----------  ----------  ----------
1940            27.9\c          NA      14.7\c          NA
1950            33.9\c          NA      23.8\c          NA
1960              37.7        41.6        31.9        34.6
1970              43.3        50.1        40.5        45.2
1980              51.5        64.3        49.9        59.3
1990              57.5        73.9        58.4        70.7
1993              57.9        74.5        59.4        72.3
----------------------------------------------------------
Note:  NA indicates data not available for this age breakdown. 

\a Married women with husbands present. 

\b As of March of the given year. 

\c Aged 14 and older. 

Sources:  Historical Statistics, Colonial Times to 1970, series D
29-41 and 49-62; Statistical Abstract of the United States 1994,
table 625; and data from the Bureau of Labor Statistics. 

This increased labor force participation reflects not only a higher
proportion of women working but also women staying in the labor force
for longer periods.  According to one study, recent generations of
married women have much stronger attachments to the labor force at a
given age than older generations did.\26 For example, married women
born from 1945 to 1949 were three times more likely to have worked 17
or more of the 20 years between ages 22 and 41 than were married
women born from 1930 to 1934 (34 percent versus 11 percent). 

The stronger attachment to the labor force of recent generations of
women is also demonstrated by the increasing percentage of mothers in
the labor force, especially those with young children (table 3.2),
and by the increasing percentage of women who have gained fully
insured status under Social Security (table 3.3).\27



                                         Table 3.2
                          
                            Labor Force Participation Rates for
                           Women, by Marital Status and Presence
                             and Age of Children, March 1960-93



                       Married\  Other\          Married\  Other\          Married  Other\
Year           Single         a       b  Single         a       b  Single       \a       b
------------  -------  --------  ------  ------  --------  ------  ------  -------  ------
1960             44.1      30.5    40.0      NA      18.6    40.5      NA     39.0    65.9
1970             53.0      40.8    39.1      NA      30.3    52.2      NA     49.2    66.9
1980             61.5      50.1    44.0    44.1      45.1    60.3    67.6     61.7    74.6
1990             66.4      58.2    46.8    48.7      58.9    63.6    69.7     73.6    79.7
1993             64.5      59.4    45.9    47.4      59.6    60.0    70.2     74.9    78.3
------------------------------------------------------------------------------------------
Note:  NA indicates data are not available. 

\a Husband present. 

\b Widowed, divorced, or separated. 

Source:  Statistical Abstract of the United States 1994, table 626. 



                         Table 3.3
          
            Percentage of Men and Women Who Were
                   Fully Insured, 1950-93


Year                                    Men          Women
----------------------------  -------------  -------------
1950                                   50.3           22.2
1955                                   75.5           44.7
1960                                   80.7           43.2
1965                                   85.6           54.3
1970                                   87.7           58.3
1975                                   88.9           63.8
1980                                   90.2           69.4
1985                                   89.2           72.3
1990                                   90.1           76.2
1993                                   90.1           77.5
----------------------------------------------------------
Source:  Social Security Administration, Office of the Actuary. 

The implication of this strong growth in women's participation rates
is that a higher percentage of women retiring in the future will be
fully insured and qualify for benefits based on their own work
records.  Also, they will have fewer years with no earnings in their
work histories when they apply for benefits than do women retiring
today.\28 A more complete work history, in turn, will increase the
AIMEs of women retiring in the future, and this should increase the
proportion of married women receiving benefits based solely on their
own work records. 


--------------------
\25 The labor force participation rates cited in this chapter include
all workers regardless of whether their employment is covered by
Social Security. 

\26 Howard M.  Iams, "Earnings of Couples:  A Cohort Analysis,"
Social Security Bulletin, Vol.  56, No.  3, pp.  22-32. 

\27 Being fully insured is necessary for entitlement to retired
worker benefits and for one's dependents to be entitled to spouse,
children, or survivor benefits.  To be fully insured, a person needs
1 quarter of coverage for each year after 1950, or the year age 21 is
attained, if later, and the year before the person reaches age 62,
dies, or becomes disabled.  In 1995, 1 quarter of coverage (up to a
maximum of 4) was credited for each $630 of covered earnings received
during a calendar year. 

\28 The median number of $0 earning years for insured workers
attaining age 62 in 1993 was 4 for men and 15 for women.  After
dropping the 5 years with lowest earnings, each year of $0 earnings
reduces the AIME by about 2.9 percent compared with what it would
have been had the earnings for that year equaled the worker's average
earnings for all years with work. 


      WOMEN'S FUTURE LABOR FORCE
      PARTICIPATION
-------------------------------------------------------- Chapter 3:1.1

Both the Office of the Actuary within SSA and the Department of
Labor's Bureau of Labor Statistics (BLS) estimate future labor force
participation rates for women.  As seen in figure 3.1, SSA estimates
women's overall labor force participation rates will peak at about
the year 2000 and then decline slowly as the baby boom generation of
women moves out of the labor force; while BLS estimates that women's
overall labor force participation rates will continue to increase
over at least the next 10 years. 

   Figure 3.1:  Past and Projected
   Labor Force Participation Rates
   for Women, Selected Years

   (See figure in printed
   edition.)

Sources:  Department of Labor, Bureau of Labor Statistics; and Social
Security Administration, Actuarial Study No.  108. 

The BLS estimates assume an increase in women's participation rates
at all ages, but especially for those aged 25 to 54 whose
participation rates are expected to increase by almost 4.5 percentage
points (from 74.7 percent in 1992 to 79.1 percent in 2005).  SSA
estimates assume more modest increases in women's labor force
participation rates within this age range during the same period.\29

Although these two estimates of women's future labor force
participation rates differ, the implications of the differences for
projections of Social Security benefits would take several years to
become evident because Social Security benefits are based on lifetime
employment, not employment in the last few years before retirement. 
Thus, the projections of the distribution of women's Social Security
benefits by type for the next few years would be similar regardless
of which estimate of future labor force participation rates is
used.\30


--------------------
\29 There are several reasons for the differences in the SSA and BLS
estimates.  Most result from different assumptions SSA and BLS used
in making their estimates, especially those for future disability
prevalence rates.  SSA assumed an increase in this rate, while BLS
did not.  This difference makes the SSA labor force participation
rate estimate lower than the BLS estimate, other things being equal. 

\30 The BLS estimates would project a slightly larger proportion of
women receiving benefits based on their own work records than would
the SSA estimates. 


   WOMEN'S EARNINGS PATTERNS
---------------------------------------------------------- Chapter 3:2

Women's benefits are not only influenced by their growing attachment
to the labor force but also by their lifetime earnings (see table
3.4).  Women's median year-round, full-time covered earnings were a
relatively constant 60 percent of men's earnings until about 1980. 
Since 1980, women's earnings have risen to about 70 percent of men's. 
This changing relationship will increase women's retired worker
benefits relative to those for men, but only slowly because benefits
are based on average earnings over the workers' entire lifetimes. 



                         Table 3.4
          
            Median Earnings of Year-Round, Full-
               Time Workers, by Sex, 1951-94


                                            Women's median
                                             earnings as a
                                             percentage of
Year                 Women           Men             men's
------------  ------------  ------------  ----------------
1951                $2,305        $3,605              63.9
1960                 3,293         5,417              60.8
1970                 5,323         8,966              59.4
1980                11,197        18,612              60.2
1983                13,902        21,854              63.6
1985                15,624        24,195              64.6
1990                19,822        27,678              71.6
1994                22,205        30,854              72.0
----------------------------------------------------------
Source:  Department of Labor, Women's Bureau. 

Social Security benefits are based on earnings in covered employment
rather than on all earnings.  Social Security covered earnings data
show that younger women have higher median covered earnings relative
to men their own age than do older women.  Relative covered earnings
fall as women age, reach a low in the women's early 30s when many
have left the labor force or reduced their hours of work to care for
children, and then rise slowly.  This "U-shaped" relative earnings
profile (relative earnings fall with age and then increase) is
flatter for more recent generations of women, reflecting their
tendency to have children later in life than older generations and to
return to work sooner after childbirth.  The higher relative lifetime
covered earnings of more recent generations of women will increase
their lifetime earnings and retired worker benefits relative to those
of men in their age group. 

In addition to the increase in women's relative median earnings, the
percentage of married women whose earnings exceed those of their
husbands' has also been increasing.  In families where both the
husband and wife have earnings (about 60 percent of all married
family households in 1992), the percentage in which the wife's
earnings exceed the husband's increased from 15.9 percent to 22.3
percent between 1981 and 1993.  It is difficult to tell from the data
whether these women consistently earn more than their husbands or
whether this is only an occasional occurrence.  The distinction is
important because Social Security benefits are based on lifetime
earnings. 

The ratio of a wife's lifetime earnings to those of her husband that
moves the wife from being dually entitled to receiving retired worker
benefits is not a constant.  The ratio will depend on the husband's
lifetime earnings, the difference in the ages of husband and wife,
the age when each begins to receive benefits, and the relative growth
in wages and prices from the time the older spouse is age 62. 

Figure 3.2 shows the AIMEs of a 65-year-old couple in 1995, both
receiving benefits for the first time at age 65, that just allow the
lower earning spouse (the wife) to receive only retired worker
benefits.\31 For comparative purposes, it also shows the AIME equal
to one-third of the husband's AIME (worker AIME/3).  To receive a
benefit based only on her own earnings, this wife needs an AIME as
high as 50 percent of her husband's (if his AIME is less than about
$387) or as low as about 26.6 percent (if his AIME is about $1,500). 
In general, her "break even" AIME will be between about 28 percent
and about 35 percent of her husband's AIME. 

   Figure 3.2:  Spouse AIME Needed
   to Receive Benefits as a
   Retired Worker, Both Husband
   and Wife Aged 65

   (See figure in printed
   edition.)

The recent increase in women's earnings relative to men's should
increase women's relative AIMEs for many years to come, regardless of
what happens to women's relative earnings in the near future.  The
higher married women's AIMEs are, relative to their husbands', the
more likely they are to receive retired worker benefits rather than
benefits as dually entitled spouses. 


--------------------
\31 That is, the AIME level at which her retired worker benefit
exceeds her benefit as a spouse. 


      FUTURE RELATIVE EARNINGS
-------------------------------------------------------- Chapter 3:2.1

We were unable to locate any quantitative estimates of the future
relative earnings of men and women.  We were also unable to obtain a
consensus on whether researchers think women's earnings will increase
or decrease relative to men's in the future.\32


--------------------
\32 The Women's Bureau at the Department of Labor believes women's
earnings will continue to increase relative to men's as the work
experience and educational attainment of women increase.  SSA
assumes, for the Social Security trustees' annual report, that the
real wage growth for men and women will be the same in the future. 
This implies that their relative earnings will be constant.  And an
economic demographer working with the 1994 Social Security Advisory
Council thinks women's earnings will fall somewhat relative to men's
in the future because (1) of the types of flat-earnings jobs women
have been taking over the past decade and (2) she does not think
men's earnings will continue to stagnate as they have during the past
10 years. 


   MARITAL STATUS
---------------------------------------------------------- Chapter 3:3

A woman's marital history at retirement determines if she is eligible
for spousal benefits based on her current or former husband's work
record, or for survivor benefits based on a deceased husband's work
record.  Eligibility for retired worker benefits is not dependent
upon one's marital status. 

As life expectancies for men increase, the percentage of women who
enter their retirement years either married or divorced should
increase, and the percentage entering as widows should decline.  As
seen in table 3.5, SSA projects that the proportion of women aged 65
to 69 who are married will remain relatively constant over the next
25 years, and that the proportion who are divorced will more than
double over this period. 



                         Table 3.5
          
            Projected Percentage Distribution of
            Women Aged 65-69, by Marital Status,
                         1995-2020


Year         Total    Single   Married   Widowed  Divorced
--------  --------  --------  --------  --------  --------
1995           100       4.5      58.6      29.6       7.3
2000           100       4.3      58.6      27.3       9.8
2005           100       4.4      59.8      24.0      11.9
2010           100       4.9      59.6      21.2      14.3
2015           100       5.4      58.9      19.4      16.4
2020           100       5.8      59.1      18.1      17.0
----------------------------------------------------------
Source:  Social Security Administration, Office of the Actuary. 

The divorce rate has more than doubled since Social Security was
enacted and the percentage of women aged 65 and older who are
divorced almost quadrupled between 1960 (1.5 percent) and 1993 (5.8
percent).  However, most divorced women do not qualify for divorced
spouse benefits because most marriages that end in divorce last less
than 10 years, the minimum marriage duration needed to qualify for
such benefits.  In addition, many divorced women who were married at
least 10 years do not receive divorced spouse benefits because they
either subsequently remarry or have retired worker benefits that
exceed their benefit as a divorced spouse. 

If the pattern of divorce after a relatively short marriage
continues, most women who divorce will not be eligible for divorced
spouse or surviving divorced spouse benefits in the future.  Lack of
eligibility for divorced spouse benefits may not pose a serious
problem for most divorced women retiring in the future because
women's increased labor force participation suggests that many will
be receiving adequate retired worker benefits based on their own work
records. 


   WOMEN'S BENEFIT TRENDS
---------------------------------------------------------- Chapter 3:4

The increasing labor force participation and earnings of women are
already reflected by an increase in the percentage of women who
receive benefits based, at least in part, on their own work records. 
As seen in table 3.6, in 1960, 43 percent of women aged 62 or older
were retired worker or dually entitled beneficiaries, while in 1993,
61 percent were.  As women with stronger attachments to the labor
force retire in the future, the percentage of women beneficiaries who
are receiving retired worker or dually entitled benefits should
increase and the percentage who are receiving spouse-only or
widow-only benefits should decrease. 



                         Table 3.6
          
              Percentage Distribution of Women
          Beneficiaries Aged 62 and Older, by Type
          of Benefit and Dual Entitlement Status,
                          1960-93


Type of
benefit     1960  1965  1970  1975  1980  1985  1990  1993
----------  ----  ----  ----  ----  ----  ----  ----  ----
Entitled    43.3  47.3  50.6  54.1  56.9  58.7  60.3  61.3
 as
 worker\a
Worker      38.7  40.7  42.1  42.3  41.0  38.6  36.9  36.2
 only
Dually       4.6   6.6   8.5  11.8  15.9  20.1  23.4  25.1
 entitled
Wife's       2.4   3.1   3.4   4.4   6.2   8.7  10.4  11.2
 benefit
Widow's      2.1   3.6   5.0   7.4   9.6  11.5  13.0  13.9
 benefit
Entitled    56.7  52.7  49.4  45.9  43.1  41.3  39.7  38.7
 as wife
 or widow
 only\b
Wife's      32.8  27.1  22.4  19.6  17.6  16.4  15.3  14.8
 benefit
Widow's     23.4  25.2  26.8  26.1  25.4  24.9  24.3  23.9
 benefit\c
==========================================================
Total       100.  100.  100.  100.  100.  100.  100.  100.
               0     0     0     0     0     0     0     0
----------------------------------------------------------
\a Includes disabled workers. 

\b Includes parents. 

\c Includes mothers and disabled workers. 

Source:  Barbara A.  Lingg, Social Security Bulletin (July 1990), pp. 
2-12; and Social Security Bulletin (Summer 1994), pp.  70-71. 

If women's labor force participation is increasing, more women should
be dually entitled and fewer entitled as a wife or widow only.  The
data support this expectation.  The percentage of women receiving
benefits solely on their own work records should also be expected to
increase, but this percentage turned downward after 1975. 

There are a number of reasons that, combined, probably explain the
downturn.  Women who might be eligible for retired worker benefits
include fully insured women who never married; those who were
divorced after being married less than 10 years (20 years before
1978); married women whose husbands had not yet taken Social Security
benefits; those whose husbands' jobs were not covered by Social
Security; and those women whose retired worker benefits exceeded
their benefits as a spouse or survivor. 

The percentage of women aged 65 and older who never married has been
falling since 1960.  As the percentage of working women who are
married increases, we would expect more to be dually entitled. 

The percentage of older women who are divorced has been rising.  In
addition, the dependency requirement divorced women faced was removed
in 1972,\33 and the years-of-marriage requirement was halved in 1977. 
The impact of these demographic and legislated changes has been a
large increase in the number of women receiving divorced spouse and
surviving divorced spouse benefits.  These rule changes also appear
to have allowed some divorced women to became dually entitled as
divorced spouses and surviving divorced spouses, whereas, without
these changes they would only have received retired worker benefits. 

As seen in table 3.7, the percentage of men whose benefits are
reduced because of early retirement is increasing.  Thus, fewer of
their wives will receive retired worker benefits while waiting for
larger dually entitled benefits their husbands' retirements can
bring. 



                         Table 3.7
          
              Percentage of New Retired Worker
          Beneficiaries Whose Benefits Are Reduced
           for Early Retirement, by Sex, 1960-93

Year                                    Men          Women
----------------------------  -------------  -------------
1960                                     NA           48.5
1965                                   30.2           48.4
1970                                   39.4           56.0
1975                                   48.9           62.3
1980                                   51.7           63.9
1985                                   65.7           75.2
1990                                   66.1           72.9
1993                                   67.8           73.0
----------------------------------------------------------
Notes:  Figures include dually entitled beneficiaries.
NA = reduced benefits not available. 

Source:  Annual Statistical Supplement, 1994 to the Social Security
Bulletin, p.  262. 

Finally, the 1983 amendments brought into covered employment many
workers who were not already in jobs covered by Social Security. 
This should increase the proportion of married women who are eligible
for dual entitlement benefits rather than retired worker benefits or
spousal benefits alone. 

Three other changes may also help explain the decline in the
proportion of women receiving benefits as retired workers.  First,
the number of years of earnings used to calculate benefits increased
by 1 year each year from 1960 until 1991 (with the exception for men
noted in the next paragraph) when it reached 35 years.  It is
scheduled to remain at 35 years.  Because women tend to have more
years without earnings than men, the lengthening computation period
often added years for which they had no earnings and, thus, reduced
their calculated AIMEs relative to men's. 

Second, the number of years of earnings used to compute men's retired
worker benefits was reduced by 3 years by the 1972 amendments.  This
had the effect of increasing men's AIMEs and PIAs slightly.  The
higher PIAs were reflected in men's higher retired worker benefits
and higher auxiliary benefits based on their work records.  This, in
turn, could reduce the number of women whose own benefits exceeded
their benefits as a spouse. 

Third, the 1977 amendments provided for ad hoc increases in the
maximum taxable earnings level for 1978 through 1981.  The maximum
taxable earnings level is now increasing faster than the growth in
median covered earnings (see fig.  3.3).\34 Because more men had
earnings above the maximum taxable level before the ad hoc increases,
this change increased the covered wages of men more than women,
resulting in higher AIMEs and PIAs for men than for women.  A higher
PIA for men increases both the auxiliary benefits based on this PIA
and the proportion of wives receiving benefits as dually entitled
spouses rather than as retired workers. 

   Figure 3.3:  Median Covered
   Earnings for Men and Women and
   the Maximum Taxable Earnings
   Level, 1940-95

   (See figure in printed
   edition.)

Source:  Annual Statistical Supplement, 1994 to the Social Security
Bulletin, pp.  169-70. 


--------------------
\33 Under this requirement, the divorced woman must have received at
least half her support from her former husband. 

\34 Maximum taxable earnings is the dollar amount ($62,700 in 1996)
above which earnings in covered employment are neither subject to
Social Security taxes nor creditable for benefit computation
purposes. 


PROPOSALS AIMED AT ADDRESSING
WOMEN'S EQUITY CONCERNS
============================================================ Chapter 4

Many proposals have been put forth over the past 25 years to address
the equity issues that arise because of existence of spousal and
survivor benefits and the workings of the dual entitlement
limitation.  Two of the broader proposals--"earnings sharing" and the
"double-decker" plan--while addressing the key concerns, also have
possible drawbacks or trade-offs associated with them. 

Although variations of the earnings sharing and double-decker
proposals are possible that would not be "cost neutral,"\35 our
discussion of these proposals assumes that the features have been
designed to be cost neutral in their impact on the Social Security
program.  That is, the benefit costs of the Social Security program
under either proposal would approximate the benefit costs of the
current program.  These costs do not reflect the potentially large
administrative burden on SSA that enactment of either of these
proposals would require. 

The other representative proposals discussed generally affect only
one or two of the equity concerns.  These proposals were usually not
designed to be cost neutral but could be made so, for example, by
adjusting general benefit levels. 


--------------------
\35 Under some variants of the double-decker plan, the flat amount
would be means tested.  These variations are generally referred to as
two-tier plans. 


   EARNINGS SHARING
---------------------------------------------------------- Chapter 4:1

Earnings sharing is a proposal that received a great deal of
attention in the late 1970s and early 1980s.\36 Under earnings
sharing, unlike under the current Social Security law, the partners
in a marriage would be treated as coequals.  The basic premise is
that both partners contribute equally to building a family and home,
even if only one has a wage-earning job. 

Under the proposal, for each year a couple is married, the annual
covered earnings of the husband and wife would be combined and
one-half of this total would be credited to each spouse's Social
Security earnings record.  Earnings sharing would cease if one spouse
becomes eligible for a disability benefit or a retired worker
benefit.  The benefit received by an individual would be based on his
or her own earnings for the years he or she was single and on his or
her half of the couple's combined earnings for the years of marriage. 

Auxiliary benefits for spouses and surviving spouses would be
eliminated under earnings sharing because each spouse would be
eligible for a benefit based on his or her own shared earnings
record.  However, under most versions of earnings sharing, a
surviving spouse would inherit the earnings credits of the deceased
spouse for the years the two were married.\37 Without this
inheritance feature, many survivors would receive lower benefits than
they do under current law. 

Earnings sharing essentially redistributes a portion of a couple's
covered earnings from the higher earner to the lower earner.  Because
women tend to have lower earnings than their husbands, earnings
sharing would increase the retired worker benefits of most married
women. 


--------------------
\36 Four reports with a major focus on earnings sharing were issued
during this period--U.S.  Department of Health, Education, and
Welfare, Social Security and the Changing Roles of Men and Women
(February 1979); U.S.  Department of Health and Human Services,
Report on Earnings Sharing Implementation Study (January 1985);
Center for Women Policy Studies, Earnings Sharing in Social Security: 
A Model for Reform (1988); and Congressional Budget Office, Earnings
Sharing Options for the Social Security System (January 1986). 

\37 The combined earnings would be capped at the annual maximum
taxable limit for each year. 


      SEVERAL CONCERNS ADDRESSED
      BY EARNINGS SHARING
-------------------------------------------------------- Chapter 4:1.1

Many of the concerns about the fairness of the Social Security
benefit structure would be addressed by earnings sharing.  It would
tend to equalize the benefits of identical one-earner and two-earner
couples with the same lifetime earnings by reducing the benefits of
one-earner couples.  Unlike under current law, there would be no
spousal benefit to increase the combined benefits of the one-earner
couple.  In fact, if the husbands and wives of two couples with equal
lifetime earnings were married for their entire working careers,
earnings sharing would credit each spouse with one-half the couple's
earnings and both couples would receive identical benefits. 

Earnings sharing also would tend to equalize the benefits of the
survivors of these one-earner and two-earner couples.  Unlike under
the current act, there would be no survivor benefits as such.  Under
most versions of earnings sharing, survivors would inherit the
earnings credits of their deceased spouses for the years they were
married.  However, a restricted inheritance provision would result in
some survivors receiving smaller benefits than they would under
current law. 

Also, a worker would not pay Social Security taxes only to find out
at retirement that the benefit she receives is the same as the
spousal benefit she would have received had she never worked. 
However, because a couple's earnings would be shared for the period
of marriage, an individual worker could still find that his or her
retired worker benefit was matched or exceeded by that of another
person who, as an individual, had lower lifetime earnings. 


      EARNINGS SHARING WOULD
      INVOLVE TRADE-OFFS
-------------------------------------------------------- Chapter 4:1.2

Earnings sharing would increase the benefits for certain categories
of beneficiaries and decrease them for others.  As noted above,
married women generally would receive higher benefits than they do
under current law as would many survivors.  However, the husbands of
these married women, one-earner couples, and two-earner couples with
dually entitled wives would receive lower benefits.  Other categories
of beneficiaries--such as married disabled men, women who qualify for
divorced spouse benefits under current law, and certain categories of
survivor beneficiaries--would also generally receive smaller benefits
under earnings sharing. 


      ADMINISTRATIVE BURDEN FOR
      THE SOCIAL SECURITY
      ADMINISTRATION
-------------------------------------------------------- Chapter 4:1.3

The Department of Health and Human Services reported that earnings
sharing would impose a large new administrative burden.  SSA would
have to obtain and maintain data on all marriages and divorces of all
individuals who work in covered employment.  Given data limitations,
SSA would find it difficult to implement earnings sharing
retroactively.  Prospective implementation would require a lengthy
transition or phase-in period if it were decided that the benefits of
those nearing retirement should be protected.  SSA would also have to
modify current, or establish new, operational procedures to ensure
the accuracy of initial and continued benefit determinations under
earnings sharing. 


      PROPOSALS SIMILAR TO THE
      BASIC EARNINGS SHARING MODEL
-------------------------------------------------------- Chapter 4:1.4

Several proposals have been made that are similar to the basic
earnings sharing "model" described above.  These proposals would
improve the fairness of benefits for one-earner and two-earner
couples that were otherwise identical and for their survivors, but
would not split earnings on a yearly basis.  The following are
examples: 

  Base a couple's benefit on its total lifetime earnings.  Under this
     proposal, the lifetime earnings of the husband and wife would be
     combined and used to calculate a benefit.  Each partner would
     receive a given percentage (usually 50 or 75 percent) of this
     "couple's" benefit. 

  Base survivor benefits on the couple's total benefits.  Benefits
     for a couple would be calculated as they are under current law. 
     However, when one spouse dies, the survivor would receive a
     given percentage (usually 67 percent) of the couple's total
     pre-death benefit. 


   THE DOUBLE-DECKER PLAN
---------------------------------------------------------- Chapter 4:2

Double-decker retirement systems are in use in a number of countries,
including Canada, Japan, Norway, Sweden, and the United Kingdom among
others.  These systems generally pay a flat amount to all qualified
beneficiaries (the adequacy deck) and then a proportional
earnings-based amount on top of that (the equity deck).  Thus, the
"social adequacy" component of benefits can be separated from the
"individual equity" component. 

In the double-decker plan, as in the earnings sharing plan, spousal
and survivor benefits would be eliminated.  In addition, some
variations allow the inheritance of the earnings credits of a
deceased spouse and the sharing of earnings credits at divorce. 

A double-decker system would redistribute benefits.  Those with very
small benefits under the current system would receive larger
benefits.  Those in the middle of the benefit distribution currently,
and possibly those at the high end, would receive lower benefits. 
The extent of the redistribution would be controlled by the relative
levels of the flat benefit and the generosity of the earnings-based
benefit.  (To keep program costs relatively constant, a larger flat
benefit would necessitate a less generous earnings-based benefit.) If
the earnings-based benefit is generous enough, those with very high
benefits under the current program could receive even higher benefits
under a double-decker system.  Whether women and couples with certain
characteristics will be absolutely better off or worse off under the
double-decker plan will depend in large measure on the extent of this
redistribution. 

Assuming that each member of retired couples receives the flat
benefit amount,\38 the equity deck payment--because it is
proportional to earnings--would equalize the benefits of two-earner
and one-earner couples with identical lifetime earnings.  The
survivors of such couples (assuming all earnings credits may be
inherited) would also have equal benefits.  However, if earnings
could only be inherited for the years of marriage, some of these
survivors could receive smaller benefits than others.  A survivor
from a shorter marriage could inherit less than a survivor from a
longer marriage. 

Also, a married working woman would not enter retirement and find
that her spousal benefit is larger than her retired worker benefit
because there would be no spousal benefit.  Married individuals would
have larger benefits than other married individuals with lower
lifetime earnings.  However, widows could receive smaller benefits
than other widows who had lower individual lifetime earnings but who
inherited more earnings from a deceased spouse. 

As under earnings sharing, there would be added administrative
burdens for SSA under a double-decker plan.  The entire process of
determining eligibility for benefits and computing benefit amounts
would change, and many procedures would also need to be changed.  If
earnings credits were shared at divorce or inherited at the death of
one spouse, SSA would also have to obtain and maintain detailed
marital history records for all covered workers. 


--------------------
\38 In some proposals, a minimum number of quarters with credited
covered earnings would be needed to receive the full flat benefit
amount. 


   OTHER REPRESENTATIVE PROPOSALS
---------------------------------------------------------- Chapter 4:3


      REDUCE SPOUSAL BENEFITS
-------------------------------------------------------- Chapter 4:3.1

One proposal would reduce spousal benefits from 50 percent of the
retired worker's benefit to 33.3 percent and increase retired worker
benefits by 12.5 percent.  This preserves the combined benefits of
one-earner couples.  At the same time, it increases the benefits of
all two-earner couples and single workers. 

Because retired worker benefits would be increased, benefits for
survivors would increase as well.  However, benefits for divorced
wives would decline by one-third.  The proposal would be costly
because the benefits of retired workers and survivors would increase
substantially. 

A second proposal would reduce couples' benefits by either reducing
the spousal benefit or modifying the benefit formula.  Benefits for
survivors would be a specified fraction (two-thirds or three-fourths)
of the couples' benefit.  Benefits for many survivors of two-earner
couples would be increased under this proposal. 


      CAP THE SPOUSAL BENEFIT
-------------------------------------------------------- Chapter 4:3.2

Spousal benefits were implemented because of concerns about the
adequacy of benefits for families of retired workers.  This proposal
would provide a limit on the maximum size of spousal benefits (and
possibly other auxiliary benefits), either by placing an upper limit,
or cap, on the spousal benefit or by limiting the dollar amount of
the retired worker's benefit that can be used as the base for spousal
benefits.  Such a cap would reduce, but not eliminate, the benefits
for spouses of workers with high AIMEs. 

A cap on spousal benefits was enacted as a part of the 1967
amendments to the Social Security Act.  However, this cap was
repealed by the 1969 amendments. 


      ELIMINATE THE SPOUSAL
      BENEFIT
-------------------------------------------------------- Chapter 4:3.3

Eliminating the spousal benefit would reduce benefits for one-earner
couples by one-third and for dually entitled two-earner couples by
less than one-third.  This proposal would not equalize benefits for
all couples with identical lifetime earnings because the progressive
benefit formula would be used to determine the benefit of each spouse
in two-earner couples, but only of the worker in a one-earner couple. 
As a result, the combined benefits for two-earner couples would be
larger than those for the one-earner couple with identical combined
AIMEs. 


      REDUCE THE DUAL ENTITLEMENT
      BENEFIT OFFSET
-------------------------------------------------------- Chapter 4:3.4

A person currently entitled to a spousal benefit has that benefit
reduced, or offset, by $1 for each $1 of retired worker benefits he
or she earned.  Benefits for such beneficiaries can be increased by
changing the dual entitlement benefit offset so that it reduces the
spousal benefit by less than $1 for each additional $1 of retired
worker benefits earned.  This would allow the beneficiary to receive
an incremental benefit increase for every additional dollar of
covered earnings he or she earns. 

The size of the increase would depend on the size of offset
established under the proposal.\39 Many two-earner couples would
receive an increase in benefits while one-earner couples and single
workers would receive the same benefits as they do under current law. 
Therefore, the current gap between the benefits received by
like-earning one-earner and two-earner couples would shrink under
this proposal. 


--------------------
\39 If the offset were small enough, both the husband and wife could
receive increased benefits, especially if their lifetime earnings
were nearly identical. 


      ALLOW SURVIVORS TO INHERIT
      THEIR SPOUSES' EARNINGS
      CREDITS
-------------------------------------------------------- Chapter 4:3.5

Allowing a survivor to inherit his or her spouse's earnings credits
has also been proposed.  These inherited credits would be added to
the survivor's own credits and a benefit would be calculated based on
these total lifetime earnings.  If all the deceased's earnings
credits could be inherited, survivor benefits for one-earner and
two-earner couples would be equalized.\40

However, if credits could only be inherited for the years of
marriage, benefits for survivors could fall if the couple had a
marriage of short duration. 


--------------------
\40 This assumes no restrictions on the total earnings credits the
survivor could have for any 1 year.  If yearly earnings credits were
limited to the annual maximum taxable amount, benefits for all
survivors might not be equalized. 


      REDUCE THE PERIOD USED IN
      COMPUTING BENEFITS
-------------------------------------------------------- Chapter 4:3.6

Workers currently becoming eligible for retired worker benefits have
their benefits based on their highest 35 years of indexed yearly
earnings.  Reducing the number of years used to compute benefits
would result in higher AIMEs and higher benefits for all workers, but
especially for women, who are more likely than men to have years with
no earnings (see ch.  3).  If shortening the benefit computation
period sufficiently increased wives' retired worker benefits relative
to those of their husbands, more women would have retired worker
benefits that exceeded their spousal benefits. 

This proposal would increase program costs, however, because the
shortened benefit computation period would increase retired worker
benefits and the related auxiliary benefits for all retired workers. 
In addition, this proposal raises concerns about the fair treatment
of workers with many years of covered work.  In this scenario,
long-term workers could receive the same benefits as short-term
workers even though the long-term worker would have paid more Social
Security taxes. 


CONCLUSIONS AND AGENCY COMMENTS
============================================================ Chapter 5


   CONCLUSIONS
---------------------------------------------------------- Chapter 5:1

Spousal and survivor benefits were incorporated into Social
Security's benefit structure in 1939.  To ensure these benefits went
only to persons who were dependent on another for their support, a
dual entitlement provision limiting the total benefit that could be
received was also enacted.  The introduction of these benefits
improved the social adequacy goal of the program and led to the
equity concerns we have discussed. 

These concerns are that (1) two-earner couples receive smaller
retirement benefits than identical one-earner couples; (2) survivors
of these two-earner couples can receive even smaller benefits
proportionally than survivors of one-earner couples; (3) dually
entitled beneficiaries make contributions to the program but receive
the same benefit they would have received had they never worked; and
(4) people who contribute to the program may receive smaller benefits
than people who either contributed less or contributed nothing at
all.  Because women receive spousal and survivor benefits more often
than men and, hence, are subject to the dual entitlement limitation
more often, these concerns are often viewed as women's issues. 

These concerns do not arise because Social Security reduces the
benefits of any group of beneficiaries.  Social Security is gender
neutral.  All beneficiaries receive a benefit at least as large as
the benefit to which their earnings histories entitle them.  The
concerns arise because some couples and survivors receive, for
adequacy reasons, benefits that exceed those based on their earnings
alone.  The size of this benefit supplement will vary according to
the characteristics of the couple or survivor. 

As the participation in the labor force of women--especially married
women--increases, we anticipate that some of these concerns could
intensify because the proportion of women who will be affected by one
or more of these concerns is expected to increase.  This will have an
impact on these equity concerns in several ways. 

First, as women's earnings and labor force commitments increase
relative to those of men, Social Security will be paying benefits to
more two-earner couples.  A larger proportion of couples will receive
benefits that are less than the total benefit received by a
one-earner couple with the same total lifetime earnings, which may
magnify the concerns relating to couples' benefits. 

Second, the increase in the number of two-earner couples will also
increase the concern about the equity of survivor benefits.  Because
(1) the survivor of a one-earner couple receives approximately
two-thirds of the couple's total benefits when both husband and wife
were alive, (2) the survivor of a two-earner couple can receive a
smaller portion, as low as one-half, and (3) two-earner couples have
smaller total benefits to begin with than one-earner couples with the
same total lifetime earnings, we anticipate that there could be
increased demands that benefits for survivors of like-earning couples
be equalized. 

Third, the anticipated growth in the percentage of women receiving
only retired worker benefits could reduce or increase the concern
that married women receive little or no incremental return for the
payroll taxes they pay.  These women will receive the same benefit
and the same return on their Social Security taxes as other retired
worker beneficiaries with the same earnings history.  This could
reduce the concern that they may not receive a fair return on their
payroll taxes.  However, the concern may intensify because more
married retired-worker women will find that their contributions to
the program result in only a small incremental increase to the
spousal benefits for which they were already qualified. 

Finally, the anticipated increase in the number of women receiving
only retired worker benefits may lessen the concern that some persons
receive larger benefits than other people who had higher lifetime
earnings and made larger contributions to the program.  However,
because there will always be some people who either never work for
pay (and receive pure spousal or survivor benefits) or who have
covered earnings sufficiently low that they receive dually entitled
spousal or survivor benefits, we do not expect this concern to
disappear entirely. 

A number of proposals have been made that would address concerns
about the fairness of the Social Security benefit structure for
working women and their families.  However, these proposals have not
been enacted, in part because they would either involve additional
costs to the program or require reductions in benefits to some
categories of beneficiaries.  This latter effect could create or
exacerbate benefit adequacy concerns for those categories of
beneficiaries.  In addition, several of the proposals, including
earnings sharing and the double-decker plan, would require
substantial changes in the program's operations and, thus, would
impose a large administrative burden on SSA. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 5:2

We provided a draft of the report to the Commissioner of Social
Security with a request for the agency's comments.  In commenting on
our report, the Commissioner stated that the report "adequately
addresses the objectives of the study" (see app.  III for the full
text of her letter).  However, she noted a few concerns. 

First, she indicated the report follows the premise that the authors
of the 1939 amendments did not envision a world in which most wives
would work.  The report has been modified to indicate the authors of
the 1939 amendments did envision such a world.  However, our point is
that equity concerns arise because the increase in the number of
women workers combined with the current benefit structure creates
certain anomalies.  In spite of these equity concerns, the benefit
structure has not been modified to eliminate these anomalies. 

Second, the Commissioner suggested we discuss how certain proposals
to address the equity concerns, such as eliminating spousal and
survivor benefits, would affect adequacy concerns.  Eliminating
spousal benefits would lower benefits received by all one-earner
couples and those two-earner couples in which the lower earning
spouse is currently dually entitled.  Eliminating the survivor
benefit would eliminate benefits for the nonworker in one-earner
couples and would lower benefits for all survivors who were the lower
earners in two-earner couples.  Appendix II covers adequacy concerns
in some detail and this chapter notes that reforming the program to
address equity concerns could create or exacerbate existing adequacy
concerns. 

Finally, the Commissioner noted that our use of cross-sectional data
for the analyses in chapter 3 could provide biased estimates of
Social Security lifetime earnings and wealth, and she offered us the
use of an alternative data set created by SSA's Office of Research
and Statistics.  We did not use the recommended data set in our
analyses.  However, we included some published findings from studies
that used this data set. 

SSA also provided us with a number of technical clarifications.  We
modified the report as appropriate. 


BENEFIT CALCULATIONS
=========================================================== Appendix I

This appendix describes how various types of Social Security benefits
are calculated.\41 Today, the calculation of a retired or disabled
worker's benefit is essentially a three-step process--(1) determining
a worker's average indexed monthly earnings (AIME, a measure of the
worker's lifetime earnings), (2) applying a benefit formula to the
AIME to determine the beneficiary's primary insurance amount (PIA,
the benefit payable at age 65 or at disability), and (3) adjusting
the PIA for the worker's age at first benefit receipt, if different
from age 65.  In addition, if auxiliary benefits are to be based on
this PIA, then the appropriate auxiliary benefit adjustment factor
must be applied.\42

The AIME calculation requires indexing the worker's covered earnings
for each year before he or she attains age 60.  Indexing is
accomplished by dividing the average wage in the national economy in
the year the beneficiary turned age 60 by the average wage for an
earlier year and multiplying the resulting amount by the
beneficiary's covered earnings for that earlier year.\43 This is done
for each year the beneficiary had earnings after 1950 and before age
60.  Earnings at age 60 or older are entered into the AIME
calculation at their nominal (unindexed) amount.  The AIME is
calculated by summing the 35 years of highest indexed earnings and
dividing by 420, the number of months in 35 years.\44

Once the AIME is determined, a benefit formula is applied to
determine the beneficiary's PIA.  The benefit formula has three
tiers--90 percent, 32 percent, and 15 percent.  The AIME levels at
which one moves from one tier to the next are known as bend points. 
The bend points are wage-adjusted annually and each year's bend
points apply only to those attaining age 62 (or becoming disabled or
dying before age 62) in a particular year.  In 1996, these bend
points are $437 and $2,635. 

Thus in 1996, the PIA for a 62-year-old worker would be 90 percent of
his or her AIME if the AIME was $437 or less; $393.30 (that is, 90
percent of $437) plus 32 percent of the AIME in excess of $437 if his
or her AIME was between $437 and $2,635; and $1,096.66 (that is, 90
percent of $437 plus 32 percent of $2,198 [= $2,635 - $437]) plus 15
percent of the AIME in excess of $2,635, if his or her AIME exceeded
$2,567.  For example, a 62-year-old worker with an AIME of $400 would
have the entire amount valued at the 90-percent level and his or her
PIA would be $360 (that is, 90 percent of $400).  A worker of similar
age with an AIME of $1,000 would have $437 valued at the 90-percent
level and the remaining $563 valued at the 32-percent level for a PIA
of $573 ($437 x 0.9 + $563 x 0.32 = $573) (see
fig.I.1). 

   Figure I.1:  Converting Average
   Indexed Monthly Earnings Into a
   Primary Insurance Amount for a
   Person Attaining Age 62 in 1996

   (See figure in printed
   edition.)

The PIA is increased by any cost of living adjustments (COLA) that
have occurred since the worker first became eligible for benefits
(this will be age 62 for retired worker beneficiaries).  Thus,
someone who waits until age 65 to claim retired worker benefits will
use the benefit formula in place when he or she was 62 and his or her
PIA will be increased by all COLAs provided beginning with the year
he or she attained age 62. 

The PIA is the figure from which almost all benefit amounts for the
worker and his or her dependents or survivors are derived.  Benefits
of retired workers and disabled workers are based on their own PIAs. 
Benefits of auxiliary beneficiaries are based on the PIAs of their
current or former spouses.\45 Benefits of dually entitled
beneficiaries are based in part on their own PIAs and in part on
those of their current or former spouses.\46

For a retired worker, the PIA will be adjusted for the age at first
benefit receipt, if other than the month the worker attains age 65
(the normal retirement age).\47 If retired worker benefits are first
taken before age 65, the monthly benefit will be the PIA reduced by
five-ninths of 1 percent for each month benefits are taken early.  If
benefits are first taken after age 65 but before age 70, the monthly
benefit usually will be the PIA increased by the delayed retirement
credit (DRC).\48 No DRC increases are given after age 70 is attained. 
Disabled worker benefits usually are equal to the worker's PIA and
are not adjusted for the age benefits are first received. 

To calculate an auxiliary benefit, the worker's PIA is multiplied by
a factor that varies according to the type of auxiliary benefit.  For
spouses and aged survivors, the benefit is then usually reduced if
the auxiliary beneficiary is under age 65 at first benefit receipt. 
No DRCs are given to auxiliary beneficiaries who first receive
benefits after age 65, although a survivor's benefit is increased by
any DRC the deceased worker would have received. 

Spouse (wife and husband) benefits and divorced spouse benefits are
available at age 62 and are equal to 50 percent of the worker's PIA. 
These benefits are reduced by 25/36ths of 1 percent for each month
the spouse first takes benefits before age 65.  Benefits for spouses
under age 65 who qualify because they have eligible children of
retired or disabled workers in their care (mothers and fathers) are
equal to 50 percent of the worker's PIA.  These benefits are not
reduced for age at first receipt. 

Survivor (widow or widower) benefits and surviving divorced spouse
benefits are available at age 60 and are equal to 100 percent of the
worker's PIA reduced by 19/40ths of 1 percent for each month the
survivor is under age 65 when benefits are first received.  However,
if the survivor benefit is first received after age 65, the survivor
benefit will not be reduced even if the survivor had first received
another type of benefit (for example, as a retired worker or spouse)
that was reduced because it was first received before the survivor
attained age 65.  Benefits of widowed mothers and fathers and
surviving divorced mothers and fathers, who qualify because they are
caring for an entitled child, are equal to 75 percent of the deceased
worker's PIA.  These benefits are not reduced for age at first
benefit receipt.  Disabled survivors and disabled surviving divorced
spouses are eligible for benefits as early as age 50.  If first
received before age 60, these benefits are equal to 71.5 percent of
the worker's PIA. 

Retired, disabled, and auxiliary beneficiaries will receive benefits
as calculated above.  Benefits of dually entitled beneficiaries
require additional steps.  First, the PIA of the current or former
spouse is multiplied by the appropriate factor to determine the
unreduced benefit for the auxiliary beneficiary.  Subsequent steps
differ for dually entitled spouses and dually entitled widows. 

For dually entitled spouses, the dually entitled beneficiary's
retired worker PIA is subtracted from the unreduced spouse
(auxiliary) benefit to determine the unreduced incremental auxiliary
benefit.  Then, the appropriate age-at-retirement adjustment factors
are independently applied to the beneficiary's retired worker PIA and
the unreduced incremental auxiliary benefit.  Finally, the resulting
figures are summed to determine the total dually entitled benefit. 

For dually entitled widows, the age-at-retirement adjustment factors
are independently applied to the beneficiary's retired worker PIA and
to the unreduced survivor benefit.  The beneficiary then receives the
larger of these two amounts. 


--------------------
\41 To be eligible for Social Security benefits on one's own work
record, one must meet certain age and insured status requirements. 
The insured status requirements are earned by having sufficient
numbers of quarters with recorded covered earnings. 

\42 If auxiliary benefits are paid on this PIA, a family maximum
limitation may be imposed if total benefits paid exceed 150 percent
of the PIA. 

\43 AIMEs of disabled and deceased workers are calculated by indexing
earnings to 2 years before the disability or death occurred. 

\44 The AIMEs of workers deceased or disabled before age 62 are
calculated using fewer than 35 years of indexed earnings as are those
for workers born before 1929.  Workers born before 1917 have benefits
calculated using earnings that are not indexed. 

\45 Parents' benefits are based on the PIAs of their deceased son or
daughter. 

\46 In a few infrequent instances, a person can be dually entitled to
benefits based on the work records of two other individuals, for
example, as the spouse of a current marriage partner and as the
survivor of a former marriage partner. 

\47 The normal retirement age is scheduled to increase gradually
beginning in 2000. 

\48 The DRC is being gradually increased.  For those attaining age 65
after 1970 and before 1982, the DRC was 1/12th of 1 percent per month
that benefit receipt was delayed between ages 65 and 72.  For those
attaining age 65 after 1981 and before 1984, the DRC was one-fourth
of 1 percent for each month that benefit receipt was delayed over
those ages.  For those attaining age 65 after 1983 and before 1990,
the DRC was one-fourth of 1 percent for each month that benefit
receipt was delayed between ages 65 and 70.  For those attaining age
65 after 1989 and before 2008, the DRC will be increased by 1/24th of
1 percent every other year and will apply between the normal
retirement age and age 70.  The DRC will be two-thirds of 1 percent
per month (8 percent per year) for those attaining age 65 in 2008 or
later. 


ADEQUACY CONCERNS
========================================================== Appendix II

The primary focus of this report is equity concerns relating to
Social Security and women, but discussions of Social Security and
women often include adequacy concerns as well.  These adequacy
concerns are based on the notion that Social Security should provide
at least a minimal standard of living for workers and their families
and survivors.  In this appendix, we discuss adequacy concerns for
three groups--elderly unmarried women, divorced women, and homemakers
and caregivers--and list some options that have been proposed for
improving the benefits of these groups.  In addition, we discuss a
proposal that was not specifically designed to address either equity
or adequacy concerns but that would affect the benefits for spouses
and survivors, if enacted. 


   ELDERLY UNMARRIED WOMEN
-------------------------------------------------------- Appendix II:1

The low incomes of elderly women, especially of unmarried women aged
75 and older, raise concerns about the adequacy of Social Security
benefits.\49 For example, in 1992 almost one in four unmarried women
aged 75 or older was poor (had an income below the poverty line) and
nearly two in five were poor or near poor (had incomes below 125
percent of the poverty line).\50 In contrast, only about one in four
unmarried men aged 75 or older was poor or near poor.  Elderly
married couples were less likely than unmarried men to be poor or
near poor. 

The problem these women have is one of low incomes, not necessarily
one of low Social Security benefits.  Most widows already receive
Social Security benefits based on their deceased husbands' work
records and these benefits are indexed to offset the effects of
inflation, thus maintaining the benefits' purchasing power over time. 
Other sources of income, such as private pensions and income from
assets, are rarely indexed and sometimes disappear upon the death of
the spouse who "owned" them. 

To improve income adequacy for these women, some advocate increasing
Social Security survivor benefits.  Increasing Social Security
benefits for widows (who account for more than 80 percent of the
unmarried women aged 65 and older living in poverty) and elderly
divorced spouses (the elderly group with the highest poverty rate) is
seen as a relatively direct and focused way of addressing this
problem.  Others believe that Social Security is doing the job it was
intended to do and that any income improvement for these women should
come from a different program such as the Supplemental Security
Income (SSI) program, which is designed to assist the aged, blind,
and disabled who have low incomes and low levels of assets. 

If it is decided that Social Security benefits should be increased to
improve the incomes of the very elderly, the proposals listed below
would accomplish this.  However, each of these proposals would
increase program costs and most would do a poor job of targeting
those elderly truly in need. 


--------------------
\49 "Unmarried" women refers to those who are widowed, divorced, or
never married. 

\50 These rates were significantly higher for black and hispanic
unmarried women than for white unmarried women. 


      INCREASE BENEFITS OF THE
      VERY ELDERLY
------------------------------------------------------ Appendix II:1.1

This option would provide a special benefit increase (for example, 10
percent) once a beneficiary attained a specified age (for example,
age 80 or age 85).  Such an increase would allow the benefits of
these elderly to catch up with improvements in the country's standard
of living that occurred since the beneficiary first began drawing
benefits.\51 This option is not well targeted to those in or near
poverty because it presumably would be given to everyone attaining
the specified age regardless of the adequacy of their incomes. 


--------------------
\51 Social Security benefits are indexed by a COLA to maintain the
benefits' purchasing power over time.  If all the beneficiary's
income were fully indexed, he or she would be able to maintain the
same standard of living he or she enjoyed when benefits were first
received.  However, real wage increases raise the standard of living
for the country as a whole.  Over time, the beneficiary's standard of
living will fall relative to the standard of living in the country as
a whole and what once might have been considered an adequate income
level might no longer be. 


      INCREASE COLAS FOR ELDERLY
------------------------------------------------------ Appendix II:1.2

A second option would give a special, one-time COLA to beneficiaries
once they reach a given age.  A related option would provide COLAs in
excess of the annual increase in prices for beneficiaries above a
specified age.  Both of these options would increase benefits for all
elderly beneficiaries, and, thus, do not target just the poor. 


      PROVIDE UNREDUCED BENEFITS
      FOR SURVIVORS
------------------------------------------------------ Appendix II:1.3

Currently, beneficiaries who first receive survivor benefits before
age 65 have their benefits reduced (by as much as 28.5 percent if
survivor benefits are first received at age 60).  Providing unreduced
survivor benefits regardless of the survivor's age at first receipt
would increase the benefits of many widows.  This option would
improve benefits only for those who become widowed before age 65. 


      ALLOW SURVIVORS TO INHERIT
      EARNINGS CREDITS OF THEIR
      DECEASED SPOUSES
------------------------------------------------------ Appendix II:1.4

This option would allow the survivor to inherit the earnings credits
of his or her deceased spouse, which would increase survivor benefits
for survivors of all two-earner couples.  However, it would not
improve the benefits of survivors of one-earner couples. 


      BASE THE SURVIVOR BENEFIT ON
      THE COUPLE'S TOTAL BENEFIT
------------------------------------------------------ Appendix II:1.5

Under this proposal, the survivor of a married couple would receive a
benefit equal to two-thirds the combined benefit of the couple.  This
would effectively increase the benefit for the survivor of a
two-earner couple to the level of the benefit received by the
survivor of a one-earner couple under current law. 


   DIVORCED WOMEN
-------------------------------------------------------- Appendix II:2

As shown in chapter 3, divorced spouse benefits are available if the
marriage lasted at least 10 years before the divorce.  However,
divorced spouse benefits are often considered to be inadequate.  For
married couples, the spousal benefit is intended to supplement the
worker's benefit and, if both spouses first collect benefits at age
65, the total benefit is 150 percent of the worker's benefit alone. 
The divorced wife's benefit, in contrast, is only 50 percent of her
former husband's benefit (again, if first taken at age 65), but this
benefit is not paid in addition to a higher benefit.  On a per person
basis, the married couple is receiving benefits that are 50 percent
larger than those received by the husband's divorced spouse (75
percent of the husband's PIA each for the married couple but only 50
percent of his PIA for the divorced spouse). 

The following options would improve benefits for divorced
beneficiaries. 


      REDUCE THE 10-YEAR MARRIAGE
      REQUIREMENT
------------------------------------------------------ Appendix II:2.1

Reducing the 10-year marriage requirement would make more individuals
eligible for divorced spouse benefits but would also increase the
probability that several former spouses would receive benefits on one
worker's earnings record.  This option raises the question of why a
person should be eligible for years of benefits based on a former
spouse's lifetime earnings history when the couple was married for
only a short time.  In addition, it does not address the primary
concern--the adequacy of benefits for divorced spouses. 


      SPLIT THE COUPLE'S EARNINGS
      CREDITS AT DIVORCE
------------------------------------------------------ Appendix II:2.2

Splitting the couple's earnings credits at divorce, a component of
the earnings sharing proposal, would equally divide the couple's
total covered earnings for each year of marriage.  Divorced women,
who were married less than 10 years and, thus, are not eligible for
divorced spouse benefits under current law, and currently eligible
divorced spouses, whose retired worker benefits exceed their divorced
spouse benefits, would generally benefit from this option. 

However, the proposal could reduce benefits for other divorced women
who are currently eligible for divorced spouse benefits or especially
surviving divorced spouse benefits because their benefits would no
longer be based on their former husbands' full lifetime earnings as
is presently the case.  This proposal would also reduce the benefits
available to most divorced men and to their subsequent wives and
families. 


   HOMEMAKERS AND CAREGIVERS
-------------------------------------------------------- Appendix II:3

Homemakers are only eligible for spouse and survivor benefits based
on their husbands' work records.  The contributions these women make
to their families are not compensated with paid wages and, therefore,
are not reflected with Social Security credits.  If a homemaker
becomes disabled or dies, the family will incur a cost to replace the
services she provided.  Social Security does not provide funds to
replace these services. 

A related group of women, caregivers, have some attachment to the
covered workforce, but spend a portion of their working lives out of
the labor force caring for children or other relatives.  Caregivers
can lose disability coverage when they drop out of the labor
market.\52 In addition, the time spent out of the labor force
performing caregiver activities is reflected in their Social Security
earnings histories as years with no earnings or years with low
earnings (if they worked only part of the year or part time) and
results in small retired worker benefits. 

The following options could improve benefits for homemakers and
caregivers. 


--------------------
\52 To be insured for disability coverage, a person generally needs
to have earned 20 quarters of coverage during the 40-quarter period
ending with the quarter in which the worker became disabled.  People
under age 31 need quarters of coverage in at least half the quarters
beginning with the quarter after they attained age 21 and ending with
the quarter of disability onset (with a minimum of 6 quarters of
coverage). 


      HOMEMAKER CREDITS
------------------------------------------------------ Appendix II:3.1

Retired worker, disability, and survivors coverage would be provided
to homemakers through credits that would be based on the imputed
value of the homemaker's unpaid services in the home.  Spousal
benefits would be eliminated.  Providing homemakers with earnings
credits would recognize the economic value to the family of their
work in the home.  Such homemaker credits would also benefit those
who drop out of the labor force to care for family members. 

However, the homemaker credit option poses several questions.  Should
women who hold full- or part-time jobs in addition to performing
unpaid homemaker services receive homemaker credits?  How is the
economic value of homemaker services determined?  Who pays for the
earnings credits provided to homemakers?  If they are paid for by the
homemaker and her spouse through a supplemental payroll tax, are such
payments voluntary or mandatory?  How would low-income families
afford their payments?  If the earnings credits are financed out of
general revenues, an equity problem would be created since covered
workers have to pay Social Security taxes to receive earnings
credits. 


      CAREGIVER CREDITS
------------------------------------------------------ Appendix II:3.2

The caregiver credits option would reduce the 35-year benefit
computation period by the number of years the caregiver was out of
the labor force providing caregiving activities.  Usually, the
proposal specifies a maximum number of caregiver credits that could
be earned by any one person (for example, 5 or 10 credits), and often
the caregiver would be required to have a preschool- aged child in
her or his care.  This proposal would reduce the number of years of
highest earnings used to calculate AIMEs and would result in
increased AIMEs and benefits for caregivers. 

However, this option may not accurately target the population of
concern because women are returning to work more quickly after the
births of their children.  If economic necessity causes mothers from
low-income families, but not high-income families, to return to work
while their children are young, then mothers from high-income
families would be the primary beneficiaries of this change, while
mothers from low-income families, presumably the primary target group
for this proposal, would forfeit caregiver credits for the years they
worked. 


      EARNINGS SHARING AND THE
      DOUBLE-DECKER PLAN
------------------------------------------------------ Appendix II:3.3

Under earnings sharing, homemakers would become eligible to receive
retired worker and disabled worker benefits and their dependents
would be insured for survivor benefits because earnings sharing would
provide the homemakers with an earnings history and a benefit
eligibility of their own.  Married women who reduce their attachment
to the paid labor force to care for children or other people would
also find that their earnings histories and benefit eligibility
generally improved from earnings sharing.  However, if they could not
inherit all their spouses' earnings credits, they could receive lower
benefits than under current law when they become widowed. 

Under the double-decker plan, homemakers and caregivers would each be
entitled to the flat-rate benefit designed to provide an adequate
floor-level benefit.  The caregiver would also receive the
earnings-related component based on her covered earnings; the
homemaker would not because she would have no covered earnings. 


   THE JOINT AND SURVIVOR PLAN
-------------------------------------------------------- Appendix II:4

This joint and survivor proposal is not directly related to adequacy
concerns, but its implementation would have consequences for the
benefits women receive as spouses and survivors. 

The rationale under which Social Security pays benefits to married
workers, their spouses, and their survivors is based in large part on
the adequacy goal of the system.  As currently structured, benefits
to auxiliary beneficiaries are not earned benefits.  Married workers
did not pay a higher payroll tax than unmarried workers to provide
the funding for auxiliary benefits.  Nor are their retired worker or
disabled worker benefits reduced to provide funding for such
benefits.  Instead, these auxiliary benefits are transfers to the
auxiliaries from the program and are paid for by a combination of
higher taxes on all workers than would be necessary and lower
benefits for all retired and disabled beneficiaries than could be
provided if these auxiliary benefits were not available. 

Another major source of retirement income, pensions, does not embody
an adequacy goal.  Like Social Security, pension plans very often pay
benefits in the form of a life annuity; that is, a benefit that is
paid to the retiree as long as he or she lives.  Spousal benefits are
rarely, if ever, provided by pension plans and survivor benefits are
usually provided only under a joint and survivor option. 

The joint and survivor option reduces a retired worker's benefit and,
in the event of his or her death, pays an annuity for the life of his
or her designated survivor.  The annuity value of the joint and
survivor option must be as great as the annuity value of the single
life option covering the retired worker alone.  However, many pension
plans allow the joint and survivor option to be waived, in which
case, the survivor would not receive an annuity upon the death of the
worker. 

If Social Security auxiliary benefits were restructured following the
joint and survivor model, spousal benefits would not be paid.  The
benefits of all married retired worker and disabled worker
beneficiaries would be reduced to provide survivor benefits to their
spouses in the event of the beneficiaries' deaths.  Thus, all married
couples, especially one-earner couples and two-earner couples in
which the lower earner is dually entitled, would receive lower
benefits than they would receive under current law.  However, the
cost savings from eliminating the spousal benefit and reducing the
transfers inherent for many in current survivor benefits could be
used to increase retired worker and disabled worker benefits. 

The survivor would receive his or her own retired worker benefit
(which also would have been reduced to provide a survivor benefit in
the event he or she died before his or her spouse) plus the survivor
benefit from his or her deceased spouse.  Survivors of one-earner
couples and many two-earner couples with uneven within-couple
distributions of earnings would receive lower benefits than under
current law, while survivors of two-earner couples with a more equal
distribution of earnings could receive increased benefits. 




(See figure in printed edition.)Appendix III
COMMENTS FROM THE SOCIAL SECURITY
ADMINISTRATION
========================================================== Appendix II


BIBLIOGRAPHY
============================================================ Chapter 1

Ball, Robert M.  Social Security Today and Tomorrow.  New York: 
Columbia University Press, 1978. 

Board of Trustees, Federal Old-Age and Survivors Insurance and
Disability Insurance Trust Funds.  1995 Annual Report of the Board of
Trustees of the Federal Old-Age and Survivors Insurance and
Disability Insurance Trust Funds.  Washington, D.C.:  Board of
Trustees, Federal Old-Age And Survivors Insurance and Disability
Insurance Trust Funds, Apr.  3, 1995. 

Bureau of Labor Statistics, U.S.  Department of Labor.  The American
Work Force:  1992-2005.  Washington, D.C.:  U.S.  Government Printing
Office, 1994. 

Burkhauser, Richard V., and Karen C.  Holden, eds.  A Challenge to
Social Security:  The Changing Roles of Men and Women in American
Society.  New York:  Academic Press, Inc., 1982. 

_____, and Timothy M.  Smeeding.  Social Security Reform:  A Budget
Neutral Approach to Reducing Older Women's Disproportionate Risk of
Poverty.  Syracuse, N.Y.:  Maxwell School, Center for Policy
Research, Syracuse University, 1994. 

Clarke, Sally C.  Advance Report on Final Divorce Statistics, 1989
and 1990.  Monthly Vital Statistics Report, Vol.  43, No.  9
Supplement.  Hyattsville, Md.:  National Center for Health
Statistics, 1995. 

Congressional Budget Office.  Earnings Sharing Options for the Social
Security System.  Washington, D.C.:  Congressional Budget Office,
1986. 

Ferber, Marianne A.  "Women's Employment and the Social Security
System." Social Security Bulletin, Vol.  56, No.  3 (1993), pp. 
33-55. 

Fierst, Edith U., and Nancy Duff Campbell, eds.  Earnings Sharing in
Social Security:  A Model for Reform.  Washington, D.C.:  Center for
Women Policy Studies, 1988. 

Grad, Susan.  Income of the Population 55 or Older, 1992. 
Washington, D.C.:  U.S.  Government Printing Office, 1994. 

Hughes, William J., and Sherwood L.  Boehlert, Chairman and Ranking
Republican of the Subcommittee on Retirement Income and Employment of
the Select Committee on Aging, House of Representatives.  How Well Do
Women Fare Under the Nation's Retirement Policies?  (Comm.  Pub.  No. 
102-879.) Washington, D.C.:  U.S.  Government Printing Office, 1992. 

Iams, Howard M.  "Earnings of Couples:  A Cohort Analysis." Social
Security Bulletin, Vol.  56, No.  3 (1993), pp.  22-32. 

_____, and Steven H.  Sandell.  "Changing Social Security Benefits to
Reflect Child-Care Years:  A Policy Proposal Whose Time Has Past?"
Social Security Bulletin, Vol.  57, No.  4 (1994), pp.  10-24. 

Lingg, Barbara A.  "Women Beneficiaries Aged 62 or Older, 1960-88."
Social Security Bulletin, Vol.  53, No.  7 (1990), pp.  2-12. 

_____.  "Social Security Benefits for Aged Women, December 1993."
Social Security Bulletin, Vol.  57, No.  2 (1994), pp.  70-71. 

Myers, Robert J.  Social Security, 4th ed.  Philadelphia:  Pension
Research Council, Wharton School of the University of Pennsylvania
and University of Pennsylvania Press, 1993. 

National Conference on Social Welfare.  Report of the Committee on
Economic Security of 1935 and Other Basic Documents Relating to the
Development of the Social Security Act, 50th anniversary ed. 
Washington, D.C.:  National Conference on Social Welfare, 1985. 

Ross, Jane L., and Melinda M.  Upp.  "Treatment of Women in the U.S. 
Social Security System, 1970-1988." Social Security Bulletin, Vol. 
56, No.  3 (1993), pp.  56-67. 

Social Security Administration.  Social Security Handbook, 1993, 11th
ed.  Washington, D.C.:  U.S.  Government Printing Office, 1993. 

_____.  Social Security Bulletin, Annual Statistical Supplement,
1994.  Washington, D.C.:  U.S.  Government Printing Office, 1994. 

Subcommittee on Retirement Income and Employment of the Select
Committee on Aging, House of Representatives.  How Will Today's Women
Fare in Yesterday's Traditional Retirement System?  (Comm.  Pub.  No. 
102-873.) Washington, D.C.:  U.S.  Government Printing Office, 1992. 

_____.  Congressional Symposium on Women and Retirement.  (Comm. 
Pub.  No.  102-897.) Washington, D.C.:  U.S.  Government Printing
Office, 1992. 

Subcommittee on Social Security of the Committee on Ways and Means,
House of Representatives.  Women and Social Security:  Families Are
Changing, The Workplace Is Changing, Should Social Security Change
Too.  (Serial 102-113.) Washington, D.C.:  U.S.  Government Printing
Office, 1992. 

U.S.  Department of Health, Education, and Welfare.  Report of the
HEW Task Force on the Treatment of Women Under Social Security. 
Washington, D.C.:  Department of Health, Education, and Welfare,
1978. 

_____.  Social Security and the Changing Roles of Men and Women. 
Washington, D.C.:  U.S.  Department of Health, Education, and
Welfare, Feb.  1979. 

U.S.  Department of Health and Human Services.  Report of Earnings
Sharing Implementation Study.  Washington, D.C.:  U.S.  Department of
Health and Human Services, Jan.  1985. 

Women's Bureau, U.S.  Department of Labor.  1993 Handbook on Women
Workers:  Trends and Issues.  Washington, D.C.:  Women's Bureau, U.S. 
Department of Labor, 1994. 


*** End of document. ***