The earnings limit raises several distributional questions. For example, do Social Security recipients who are subject to the earnings limit have more, or less, income from sources such as dividends, interest, and pensions? Does the earnings limit penalize low-income workers who need to continue working? This Data Digest tries to answer some of these questions by examining data from the Current Population Survey (CPS) during the three-year sample period from 1996 to 1998.
Description of the Earnings Limit
The earnings limit reduces Social Security benefits when a beneficiary's labor income exceeds a certain threshold, called the "exempt amount." The earnings limit applies only to wage and salary income above the exempt amount; it does not apply to pension, dividend, or interest income that may exceed the exempt amount, because the FICA tax is paid only on work. The exempt amount is indexed to wage growth. In recent years, the following wage or salary amounts were exempt from the earnings limit:
Until the law was changed in 2000, the earnings limit reduced the Social Security benefit by one of two calculations, depending on the age of the beneficiary.
- For ages 62-64, the Social Security Administration withheld $1 in benefits for every $2 in wages earned above the exempt amount.
- For ages 65-69, $1 was withheld for every $3 in wages above the exempt amount.
The 50-cents-on-the-dollar benefit reduction for ages 62-64 is still in place, while the earnings limit has been lifted for age 65 and over starting in 2000.
Although it is not well understood, individuals who receive reduced Social Security benefits because of the earnings limit will eventually recapture all of their lost benefits after they reach the normal eligibility age for full benefits, and if they live long enough. The period required for full recapture of benefits is computed actuarially. Thus, the earnings limit resembles a program of enforced saving more than it does a tax, although whether this forced saving is efficient can be debated. The recapture feature may be inequitable to certain racial or income groups that tend to have shorter lifespans. (Gruber and Orszag, 1999)
The analyses below use data from the Current Population Survey (CPS) conducted during the years 1996, 1997, and 1998. Three CPS survey years were pooled to amass sufficient observations to make the results statistically meaningful. The CPS dataset is well known and widely used by researchers, but its income measures have some shortcomings. Income is self-reported, and it is known that many people under-report their non-wage income, such as interest and dividend income, to the CPS and to other surveys. This underreporting affects data for those age 62 and over, who have more capital income than younger groups.
It was necessary to drop certain observations from the dataset. Following the methodology of Gruber and Orzag (2000), people aged 62 or 65 were dropped from the dataset because of ambiguity regarding respondents' age during the period under survey. Additionally, people who reported that they retired during the survey year were also dropped from the dataset. This was because the presence of mid-year retirees could result in a large number of observations with wages below the earnings limit exempt amount, and muddle the behavioral interpretation of these lower wages. While some people may time their retirement so as to stay under the earnings limit exempt amount, it is likely that many more choose to retire at a particular point in the year because of poor health, the need to care for an ill spouse, or some other personally important event. State and local government workers were also dropped because some state and local governments do not participate in the Social Security system. Finally, anyone with missing values for any of the relevant variables was dropped from the dataset.
In the analyses that follow, individuals age 63-64 were analyzed separately from individuals age 66-69. This is because people who claim Social Security below the normal eligibility age for full benefits face an additional benefits reduction that can be as high as 20.8 percent under current law. The normal eligibility age is currently 65 and two months; for the period from 1996-98 that is under study here, it was age 65. Additionally, Congress recently decided to lift the earnings limit for ages 65 and older, but retain it for the younger age group, increasing the importance of separating out any behavioral differences between the two age groups. The population aged 62 and older averaged 38 million during the three-year sample period. After the various adjustments to the dataset, the remaining weighted observations numbered 9.6 million individuals in the combined 63-64 and 66-69 age groups, about 354,000 of whom were subject to the earnings limit.
Earnings Limit Incidence
Table 2 examines the incidence of the earnings limit among workers within each wage/salary bracket. The table shows that workers in the wage/salary bracket just above the exempt amount (the first shaded rows) were more likely to be subject to the earnings limit, because they received Social Security, than workers in higher wage brackets.
This first wage group above the exempt amount may be of concern to policymakers. These are the people who may need to supplement their wage income with Social Security because of high medical or other expenditures. Table 1 also shows that workers aged 66-69 have a much higher rate of collecting Social Security benefits than workers in the younger group. This may be partially because workers over age 65 no longer face an actuarial reduction for taking benefits below the normal retirement age. This may also reflect a practice of encouraging workers to apply for benefits at age 65 when they register for Medicare benefits (Gruber and Orszag, 1999).
Cutting the numbers another way, Table 3 examines the incidence of the earnings limit across wage/salary groups for the subset of workers who collect Social Security. While Table 2 sums across the rows to 100 percent, in Table 3 the columns sum to 100 percent. In other words, within each age group, the cells show the incidence of the earnings limit by wage. Focusing again on the workers in the wage bracket immediately above the exempt amount (the first shaded row), the percentage of workers in these cells is sizeable. The balance of the affected workers would appear to have wage/salary incomes that lie more than $10,000 over the exempt amount. (The earnings limit is adjusted for wage inflation every year, so the top wage bracket of 63 to 64-year-olds had more than $18,280 in wages in 1986, and more than $19,120 in wages in 1998. The top wage bracket of 66 to 69-year-olds had over $22,500 in wage income in 1996, and over $24,500 in wage income in 1998.)
(Click here to see Chart 1 Comparison of Income Sources: Social Security Recipients and Non- Recipients Ages 63-64.)
Total Income Sources of Individuals Who Are Subject to the Earnings Limit
The next set of analyses asks whether workers who are subject to the earnings limit have different amounts, and sources, of income than individuals in the same age group who also continue to work but do not yet receive Social Security. Chart 1 looks at the 63-64 age group. Chart 2 looks at the 66-69 age group.
Workers were divided into four wage/salary brackets. The first bracket consists of workers who had wage or salary earnings that were below the earnings limit. The second bracket consists of workers who had wage or salary income of one dollar to $5,000 above the earnings limit. The third bracket consists of workers who had wage or salary income between $5,001 and $10,000 above the earnings limit. The final bracket consists of workers who had wage or salary income that exceeded the earnings limit exempt amount by more than $10,000.
There are two bars within each wage/salary bracket relative to the exempt amount. The bar on the left of each pair represents workers who do not receive Social Security and the bar on the right represents workers who do receive Social Security. The percent of Social Security beneficiaries within each wage bracket is given below the horizontal axis. It is noteworthy that as wages increase, the percent of workers collecting Social Security falls dramatically.
The height of the bars represents total non-Social Security income. The shaded portion at the bottom of the bars (and the associated dollar figure) represents the mean wage or salary of workers. The solid portion at the top of the bars (and the associated dollar figure) represents the mean income from all other sources, such as pensions, dividends, and interest for the same persons. The figures and bars do not include Social Security income.
The first pair of bars on the left of Chart 1 represents workers aged 63-64 who had wage or salary income below the exempt amount. Thus, even though some of these workers received Social Security, they were not subject to the earnings limit. Within this lowest wage bracket, those who did collect Social Security had higher mean wage/salary earnings ($4,957) than those who did not collect Social Security ($4,725). The total income of the Social Security recipients was lower than that of non-recipients, however. This is because the non-recipients had much higher income from other sources, such as interest, dividends, and pensions, than non-recipients. Of the workers in this lowest wage bracket, 71% chose to receive Social Security.
(Click here to see Chart 2 Comparison of Income Sources: Social Security Recipients and Non-Recipients Ages 66-69.)
The earnings limit applies to all Social Security beneficiaries that have labor earnings above the earnings limit exempt amount (Chart 1, bars 3-8). Within the three higher wage brackets, the workers who were affected by the earnings limit (because they collected Social Security) had lower mean wage/salary income than workers in the same wage/salary brackets who do not yet receive Social Security. The earnings limit group frequently had higher income from non-Social Security sources such as pensions, dividends, and interest, however.
For workers with wage/salary income between one dollar and $10,000 above the exempt amount, this income from other sources was enough to raise total (non-Social Security) income above the level of workers in the same wage bracket who did not yet receive Social Security.
Distributional Impact of the Earnings Limit for People Aged 66-69
Turning to the 66-69 age group, Chart 2 shows that at all but the highest income levels ($10,001 or more above the earnings limit exempt amount), workers affected by the earnings limit had slightly higher mean wage or salary income than workers who did not yet receive Social Security. Workers affected by the earnings limit also had higher total (non-Social Security) incomes, because they had higher mean income from other sources than workers who do not yet receive Social Security, again with the exception of the highest wage bracket. In the highest wage/salary bracket, Social Security recipients had lower mean wage/salary income, and lower mean income from other sources, than non-recipients.
Policymakers have expressed concern that the Social Security earnings limit affects workers who have wages immediately above the exempt amount, and who may need to supplement their wages with Social Security to meet everyday or extraordinary expenses. The data presented here lead to mixed conclusions about the impact of the earnings limit on lower-wage workers. Tables 2 and 3 showed that a sizeable number of workers with wages immediately above the exempt amount were affected by the earnings limit. Yet Charts 1 and 2 showed that workers who were affected by the earnings limit generally had higher total income than workers within the same income bracket who did not receive Social Security. The earnings limit group aged 66-69 often had higher mean wage or salary income, and both age groups often had higher mean income from other sources such as interest, dividends, or pensions, than non-beneficiaries in the same wage bracket. Unfortunately, the data do not provide any insight into expenses faced by people who are subject to the earnings limit, such as medical or other expenses.
Gruber, Jonathan and Peter Orszag, "Does the Social Security Earnings Test Affect Labor Supply and Benefits Receipt?" September, 2000.
Gruber, Jonathan and Peter Orszag, "What to do About the Social Security Earnings Test," Center for Retirement Research, Boston College, July 1999, number 1.
Written by Alison Shelton, AARP Public Policy Institute
May be copied only for noncommercial purposes and with attribution; permission required for all other purposes.
Public Policy Institute, Public Affairs, AARP, 601 E Street, NW, Washington, DC 20049
Discounts & Benefits
Next ArticleRead This