Abstract
One of the main retirement income concerns is adequacy, which is generally measured as the percentage of preretirement income replaced by the retirement benefit. For defined-benefit pension plans the replacement rate or ratio is measured against some measure of final income. For Social Security (more accurately OASDI), however, the Social Security Administration uses a complex measure of career-average earnings. Since Social Security is the main, and often the only, source of retirement income for many millions of Americans, the question of adequacy is paramount. The benefit formula is heavily skewed in favor of lifetime low-income earners. In addition, Supplemental Security Income (SSI) and state augmentations add to the retirement income of low-income claimants. The combination of OASDI and SSI benefits result in retirement income that, while not overly generous, is adequate. This article examines the origins and development of OASDI and SSI with a focus on benefit adequacy.
When most people refer to “Social Security” they mean the Old-Age, Survivors and Disability Insurance (OASDI) program. OASDI is only one of the many programs contained in the Social Security Act of 1935, as amended. Others include Supplemental Security Income (SSI), Medicare, Medicaid, and the Children’s Health Insurance Program (CHIP).
OASDI is the foundation of the social insurance program of the United States. It provides income replacement to the elderly and disabled and their dependents and survivors. Almost all of us will one day become (or are) elderly and many will become disabled. And we all have family and friends who are OASDI benefit recipients. OASDI is an extremely important program.
This article examines income replacement ratios and benefit adequacy of the OASDI program. To a lesser extent it touches upon replacement rates of employer-sponsored pension plans and other retirement savings arrangements.
Origins
The Social Security Act (Pub. L. 74-271) was signed into law in 1935 as one of the main elements in the New Deal response to the Great Depression. It was amended in 1939 by the Federal Insurance Contributions Act (Pub. L. 86-272; FICA). Social Security initially applied to covered workers in the private sector. They and their employers make equal mandatory contributions as a percentage of earnings to the Old-Age and Survivors Insurance (OASI) Trust Fund. OASI is essentially a huge defined-benefit (DB) retirement income program that pays benefits for life.
Expansion
OASI coverage was expanded in 1954 to farmers and clergy. Also in 1954, Congress passed the Self-Employment Contributions Act (Pub. L. 114-38; SECA), the counterpart to the FICA. In 1956, Social Security was extended to active-duty military and naval personnel. Disability Insurance (DI) were added in 1957 and expanded upon in 1960. Thus, the OASI became the OASDI program. DI (usually SSDI to distinguish it from other types of disability insurance) is essentially an insurance arrangement that protects workers against loss of income due to a physical or mental ailment that precludes gainful employment. SSDI benefits convert to retirement benefits at full retirement age (FRA). Early retirement at age 62 with reduced monthly benefits was added for women in 1956 and for men in 1961. 1
In the 1950s and 1960s, OASDI was extended to some, but not all, state and local government employees on a voluntary (to the states) basis through what are called Section 218 Agreements. Today, about 25% of state and local government employees remain outside of Social Security (mainly teachers in 15 states, police officers and fire fighters). Nationwide, 40% of public school teachers and two thirds of police officers, fire fighters and other first responders are not covered by Social Security. 2 New federal employees were covered beginning 1984.
Coverage
It would be hard to overstate the importance of the OASDI program to the U.S. economy and society. As reported in Table 1, there were over 60.9 million OASDI beneficiaries at the end of 2016 (18.7% of the population). They included 44.3 million retired workers and their dependents, 10.6 million disabled workers and their dependents and 6.0 million survivors, mainly widows and widowers of deceased workers.
Number of Beneficiaries, Total Benefits Paid and Average Monthly Benefit as of December 31, 2016.
Source. Social Security Administration. Social Security Program Fact Sheet. December 31, 2016. Retrieved from www.ssa.gov/OACT/FACTS/index.html.
Table 1 also reports the total monthly benefit payments. For all OASDI beneficiaries it is $75.9 billion per month. That includes $58.2 billion for retired workers and their dependent, $11.0 billion for disabled workers and their dependents and $6.8 billion for survivors.
Retirement Benefits
OASI retirement benefits are not welfare. They are a statutory right that have been paid for by required employee and employer contributions. Benefits are available to any claimant who has attained the minimum retirement age of 62 or over and has earned 40 credits (quarters of covered employment). In 2017, a worker must have earned at least $1,300 in a quarter for it to count as a credit. That amount is indexed to the national average wage index (AWI).
The primary insurance amount (PIA) is the unmodified retirement benefit to which a claimant is entitled at full retirement age, currently 66. The PIA is calculated by indexing annual earnings to the AWI to age 60. The highest 35 of those years and un-indexed years over age 60 are then converted to the average indexed monthly earnings (AIME).
The AIME is subjected to a progressive benefit formula derived from the two “bend points” in the three-segment (kinked) AIME curve. For 2017, the first bend point is $885 and the second is $5,336. AIME up to $885 is multiplied by 90%. AIME from $885 to $5,336 is multiplied by 32%. AIME above $5,336 is multiplied by 15%. The sum of these products is the claimant’s PIA.
This arrangement results in a higher percentage of earnings replacement for those with lower AIMEs. The bend points segments increase with the year of eligibility (the year in which the worker attains the age of 62 or becomes disabled before the age of 62). 3
Minimum OASDI Retirement Benefits
Social Security has had a minimum retirement benefit provision since 1939. It was originally $10 per month but was increased several times after that.
The 1972 Amendments to the Social Security Act (Pub. L. 92-603) created the Special Minimum PIA. Effective 1973, the law set the benefit at $8.50 per year of covered employment. That was increased it to $9.00 in 1974. There was no provision to automatically increase (index) the minimum benefit.
In 1977, the Social Security Financing Amendments (Pub. L. 95-216) increased the minimum benefit to $11.50 per year of coverage. It also indexed the minimum benefit to prices (Consumer Price Index), not wages (AWI) effective 1979. That was a mistake. Retirement and disability benefits are indexed to the AWI (wages).
Wages increase more rapidly than prices due to improvements in labor productivity. 4 Thus, over time, fewer and fewer claimants were eligible for the minimum benefit. By 2013, only 35,000 of the 54 million OASDI beneficiaries were eligible for the Special Minimum PIA and it added an average of $46 to their monthly benefit.
Since 1999, the minimum-benefit provision has largely applied to new claimants whose PIA had been reduced by the Windfall Elimination Provision (WEP). The WEP reduces the retirement benefit of claimants entitled to a government pension plan.
The Social Security Administration projects that the Special Minimum PIA will have no effect on claimants turning 62 in 2019 or later.
Disability Benefits
To be eligible for SSDI benefits, the disability must be a medically determined physical or mental condition that is permanent and total. “Permanent” means that the condition is expected to last for at least 1 year or result in death. “Total” means that the claimant is unable to engage in substantial gainful activity (SGA). In 2017, the threshold for SGA was not being able to earn $1,170 per month if non-blind or $1,959 if blind. The disability benefit amount is based on the worker’s PIA at the time of disability.
Eligibility for SSDI benefits is also affected by age and the number earned credits. A disabled individual who has 40 credits is “fully insured” and eligible for SSDI benefits. If not fully insured, the number of credits needed varies with age. Before age 24, the requirement is six credits earned in the previous 3 years. Between age 24 and 31, it is half of the quarters between age 21 and the time of the disability. If the claimant is age 31 to 42, he or she needs at least 20 credits prior to the disability. The requirement then increases until it is 40 credits at age 62.
Replacement Rates
The average monthly retirement benefit is $1,360 for retired workers and $1,171 for disabled workers (Table 1). Whether or not these benefits are adequate is debatable. It depends largely on one’s political and economic philosophy and, to some extent, how one does the math.
The term “replacement rate” or “replacement ratio” is a generally accepted measure of benefit adequacy. However, the term is used quite differently for traditional defined-benefit pension plans and for OASDI benefits. This may be the source of confusion.
Pension Plans
In a single-employer DB pension plan, the benefit amount is typically based on the participant’s years of service, an accrual factor or multiplier (say 1.5) and some measure of final income, typically the average of the final or highest three consecutive years (usually the same). Since the average of the high-3-year salary is multiplied by all years of participation in the plan, such plans disproportionately reward (and encourage) long service. They are said to be “back loaded.” In designing such a pension plan, the employer may choose an accrual factor that provides the desired income replacement ratio for long-service employees (however defined).
In this context, the replacement ratio is calculated as a percentage of final income (pension benefit ÷ final income). That is the appropriate measures if the aim is to allow employees to maintain their current standard of living in retirement. Under this approach, the replacement ratio for long-service employees can be quite high.
Under a defined-contribution arrangement, such as a Section 401(k) plan, that pays a benefit as a lump-sum distribution, the replacement rate is meaningless unless the account balance is used to purchase a group or individual annuity contract from an insurance company. In that case, it would be the monthly annuity payment divided by pre-retirement monthly earnings.
In the private sector, collectively bargained multiemployer pension plans typically pay a dollar amount per year of participation. Replacement ratios as a percentage of final income are usually pretty small and are seldom discussed. In the public sector they are more generous.
OASDI
The Social Security Administration (SSA) replacement-rate calculation for the OASDI program is not based on the final or highest years of income. Rather, it uses a complex measure of career average earnings. The numerator is the Social Security benefit. The denominator is the average of the highest 35 years of wage-indexed earnings for hypothetical workers to the year before retirement. This is almost the same as the AIME calculations discussed above except that the AIME calculations use wage-adjusted earnings to age 60 rather than to the year before retirement. 5
From 1989 through 2013, the Annual Trustees Reports included replacement rates for several hypothetical workers. 6 That was discontinued in 2014. However, the replacement rates remain available elsewhere at Actuarial Note Number 2016.9. Presumably, they will continue to be available.
Initially, the hypothetical workers were steady workers earning the same wages throughout their careers. More recently, “scaled” earnings patterns were developed by the SSA’s Office of the Chief Actuary to reflect more accurate earnings pattern by age. 7
It had been found that assuming constant earnings throughout the work life overrepresent earnings at the beginning and end of careers and underrepresent those in the prime earnings years. Most males who are retired or who are now retiring spent two or more years in the service at relatively low pay after high school or college. And many OASDI claimants have no or significantly reduced earnings in the years before age 62 due to medical or other reasons.
Table 2 depicts career average earnings, wage-indexed retirement benefits in 2016 dollars and replacement rates for five categories of hypothetical workers born in 1950 who retired in 2016. The first four categories—very-low earners (with career average earnings in 2015 of $11,933), low earners, median earners and high earners—are scaled beginning at age 21. The fifth hypothetical worker has earned the maximum FICA wage base from age 22 until retirement age. Table 2 reports data for retirement at age 66 (current full retirement age).
Social Security Benefit Replacement Rates for Hypothetical Workers at Different Earnings Levels (Born 1950, Attained FRA 2016 at Age 66.
Note. FRA = full retirement age.
Source. Social Security Administration, Office of the Chief Actuary. (2016, June). Actuarial Note 2016.9. Table C, pp. 7-8. Prepared by Michaael Clingman, Kyle Burkhaulter and Chris Chapman. Retrieved from www.ssa.gov/OACT/NOTES/ran9/an2016-9.pdf.
Note in Table 2 how the replacement rates at FRA are significantly higher for very low earners (76.3%) and low earners (55.4%) compared to high earners (34.1%) and steady maximum earners (27.3%). That is due to the progressive benefit formula based on the bend points of the AIME curve discussed above. It reflects a massive transfer of resources from high earners to low earners. Few would object to this arrangement.
Claiming early retirement benefits at age 62, rather than working until age 66, reduces the retirement benefit. When the FRA was 65, a claimant who retired at age 62 had his or her full retirement benefit reduced by 20%. When it was exact integral age 66, it was 25%. When the FRA will be age 67, for those born in 1960 or later and retiring at age 62 (in 2022 or after), the reduction will be 30%. This is important. Many people claim early retirement benefits.
Table 3 relates the number, average age at retirement and the percentage of men and women who retire at age 62 and at FRA for selected years 1960 through 2015. The percent retiring at FRA excludes those converting from disability to retirement benefits at full retirement age.
Number, Average Age and Percent Retired at Age 62 and at Full Retirement Age by Sex, Selected Years, 1960 to 2015.
Note. FRA = full retirement age.
Excludes those converting from disability benefits to retirement benefits at FRA.
Early retirement benefits did not become available for men until 1961.
Source. Social Security Administration. (2017, May). Annual Statistical Supplement to the Social Security Bulletin, 2016. Table 6.B5, pp. 6.17-6.18. Retrieved from www.ssa.gov/policy/docs/statcomps/supplement/2016/index.html.
There are many reasons for the large number of retirements at 62. They include poor health or physical condition, being unemployed and unable to find another job because of age or health, wanting to join an older spouse in retirement, not expecting to live a long life based on health or family history and not understanding the extent to which the retirement benefit will increase with additional years of work.
In 2015, 31.9% of men and 37.4% of women claimed retirement benefits at age 62. Only 17.6% of men and 12.3% of women worked until full retirement age. That is up from 11.2% for men and 7.4% for women in 2005. Most of the remainder claimed between age 62 and the FRA.
Observe how early retirement at age 62 for men grew markedly from 19.0% in 1970 to 50.1% in 2005 and then has declined to 31.9% by 2015. The percentage of men continuing to work and contribute until FRA declined from 42.3 in 1970 to 11.2 in 2005. It subsequently increased to 17.6% in 2015.
The experience of women is similar. Those retiring at age 62 increased from 27.1% in 1960 to 55.9% in 1990. That was 17.6 percentage points higher than the male rate and 15 years earlier. Women became eligible for early retirement in 1956. Men had to wait until 1961. The percentage of women who retired at FRA bottomed in 2005 at 7.4 (as it did for men at 11.2). Women on average retire earlier than men due in part to the fact that most women marry older men.
This was due in part to the average age at retirement. For men, it declined from 66.8 in 1960 to 63.6 in 2005. For women, from 65.2 to 63.6 over the same period. It then increased to 64.5 for men and 64.2 for women in 2015.
Adequacy of OASDI Replacement Ratios
For those at the upper end of the income distribution who have made prudent decisions about saving and investing, OASDI benefits may be a welcomed addition to their total retirement income. For many others, it is essential.
Social Security benefits represent about 34% of the income of the elderly in general. Among elderly recipients, 48% of married couples and 71% of unmarried individuals receive 50% or more of their income from Social Security. More striking, 21% of married couples and 43% of unmarried individuals rely on Social Security for 90% or more of their retirement income. 8
Using the AIME career average as the denominator in calculating the OASDI replacement ratios may understate them. However, individuals and couples receiving low OASDI benefit amounts are usually entitled to federal SSI benefits that are often augmented by additional state benefits. If they are included in the calculations, the replacement ratio for those most in need improves significantly.
Supplemental Security Income
The SSA also administers a needs-based benefit program financed from the General Fund (not OASDI payroll contributions) called Supplemental Security Income. SSI was also created by the 1972 Amendments. It became effective January 1, 1974, and is Title XVI of the Social Security Act. SSI replaced, federalized and made uniform a number if preexisting need-based benefit programs run by the states. It provides income for the blind, disabled and elderly (age 65 and over) individuals and couples with limited income and limited resources.
“Income” in this context includes money earned from work, payments from Social Security, workers compensation, unemployment insurance, veterans’ benefits and even free food and shelter. “Resources” are limited to $2,000 for an individual and $3,000 for a couple exclusive of a home and vehicle ownership.
In 2017, the maximum federal SSI payment was $735 for an individual and $1,103 for a couple. Those numbers are wage indexed on the same basis as Social Security: third quarter of current year compared to same quarter of previous year determines the benefit amounts for next year.
This is not to imply that an individual receiving SSI will receive an additional $735 per month. Rather, the SSI benefit is reduced by what the SSA calls “countable income.”
Countable income includes “unearned” income (such as OASDI and pension benefits), “earned” income (from work) and “deemed” income (of a spouse or parent with whom one resides).
The first $20 of unearned income and the first $65 of plus one-half of earned income above $65 is excluded from countable income.
The computation of the actual SSI payment is situational and can get complicated. A simple example would be an elderly recipient who receives $300 per month in OASDI retirement benefits would have $280 in countable unearned income ($300 − $20 = $280). That amount is subtracted from the $735 maximum SSI benefit to give him or her SSI benefit of $455. 9
If the same individual had employment paying $500 per month, the SSI benefit would be further reduced. After excluding the first $65 and half of earnings above $65, an additional $217.50 would be added to countable income. When that is added to the $280 from the above example, it gives $497.50. And when that is subtracted from the $735 maximum, it leaves $237.50 as the SSI benefit on top of the claimant’s OASDI benefit (author’s calculations).
Table 4 reports the number and percentage of total claimants and those with reduced benefits due to early retirement by their monthly benefit amount as of the end of 2015. Note the large number and percentage of total and reduced-benefit claimants with monthly benefits of $700 or less. That makes them eligible for SSI payments.
Number and Percentage of Retired Workers by Selected Monthly Benefit Amounts, December 2015.
Note. Selections and calculations by author.
Source. Social Security Administration. (2017, May). Annual Statistical Supplement to the Social Security Bulletin, 2016. Table 5.B6. Retrieved from www.ssa.gov/policy/docs/statcomps/supplement/2016/index.html.
It is also interesting to observe that 81.8% of those with benefits of less than $700, and 88.8% with benefits of less than $1,000 claimed benefits early compared to only 32.0% of those with benefits of $2,000 or more. Clearly there is an association between low lifetime earnings (and AIME) and reduced benefits due to early claiming.
State Augmentations to Federal SSI
Replacement ratios are further complicated by state augmentations of the federal SSI benefit. Five states do not augment the federal SSI payments to their resident recipients. They are Arizona, Mississippi, Northern Mariana Islands, West Virginia and North Dakota. The Northern Marianas and Washington, D.C. (but not Puerto Rico) are considered states in this context.
Eleven states have their supplemental augmentations administered by the Social Security Administration: California, Delaware, District of Columbia, Hawaii, Iowa, Montana, Nevada, New Jersey, Pennsylvania, Rhode Island and Vermont. The remaining 37 states administer their own supplemental payments programs. 10
The specifics of the state augmentations to the federal SSI payments lie beyond the scope of this article. However, the SSA recently published information on California’s SSI augmentation for 2017. In California, the combined federal and state SSI benefit for an aged or disabled individual living alone was $895 and for couples $1,510. For a blind individual it was $952 and for a couple $1,661. 11 Thus, the state adds $160 for an individual and $407 for a couple to the SSI maximums.
Conclusion
OASDI is the centerpiece of the social insurance program of the United States. Many millions of elderly and disabled Americans depend on it for much or all of their income. The matter of benefit adequacy is important to them personally, to the nation and to policy makers at all levels. The replacement ratio is generally accepted as a measure of adequacy.
Defined-benefit pension plans in the private and public sectors calculate replacement ratios as the percentage of preretirement income covered by the pension benefit. That makes sense. Most people expect to continue at their same standard of living in retirement. However, most participants in DB pension plans or annuitized defined-contribution arrangements continue working until they retire.
A large number of OASDI claimants do not work until retirement. Many have zero or curtailed earnings in the years before attaining early retirement eligibility at age 62. A large percentage of Social Security claimants retire at that age. In 2015, it was 31.9% for men (down from 50.1% in 2005). For women it was 37.4% (down from 54.3% in 2005).
Because of this, the SSA measures the benefit replacement ratio against wage-indexed career average earnings. While measuring adequacy against career average rather than final earnings may overstate the appearance of adequacy, individuals and couples at the lower end of the benefit distribution are helped by federal SSI and in most cases by state augmentations to SSI. The result is that the combined OASDI and SSI benefit is generally acceptable. While not overly generous, it is adequate from a public policy perspective.
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.