|Tax Policy Outlook logo 130|
No. 3, July 2008
Most seniors view the Social Security earnings test as a "tax" that reduces their Social Security benefits by fifty cents for each dollar they earn above a modest limit. In fact, the earnings test is not a tax at all: at a person's full retirement age, Social Security increases benefits to account for any lost to the earnings test in earlier years. Over the typical retiree's lifetime, total benefits are almost exactly the same. Most retirees are unaware of this because the Social Security Administration (SSA) and financial advisers fail to inform them of how the earnings test works. Retirees need better information--and policymakers should consider whether the earnings test makes sense at all.
Readers of financial columns know that the Social Security earnings test reduces benefits for retirees aged sixty-two to sixty-six who continue to work. Most seniors view the earnings test as a 50 percent tax on top of their income and payroll taxes, which discourages them from working and receiving benefits at the same time.
But the conventional wisdom regarding the Social Security earnings test is wrong. At the full retirement age (sixty-six for those born in 1943-54), Social Security increases benefits for individuals who had been subject to the earnings test, which completely offsets benefits lost earlier. Most seniors do not know this, since government agencies and financial advisers rarely tell them. Until very recently, SSA's own publications on the earnings test did not provide details on the benefit adjustment that takes place at the full retirement age. As a result, many retirees work less at the very time when continued work could benefit them most.
How the Earnings Test Works
The earnings test has been in existence since Social Security was founded in 1935. In early years, there was a hard earnings test, under which individuals were ineligible for Social Security benefits if they had even a dollar of earned income. This provision was designed to push older Americans out of the workforce in the mistaken belief that having fewer older workers would create more job opportunities for younger Americans.
In later years, the earnings test was amended so that it steeply reduced--but did not eliminate--benefits. Until the year 2000, the earnings test applied to all Social Security beneficiaries under the age of seventy. In that year, however, Congress repealed the earnings test for beneficiaries over the full retirement age, keeping it in place only for early claimants.
Here is how the retirement earnings test works today: for individuals claiming benefits who are between age sixty-two and the full retirement age, Social Security withholds $1 of benefits for every $2 in earnings above $13,560. For instance, a typical retiree receiving a $1,000 monthly benefit would see his benefits eliminated if he earned more than $37,560. As noted above, after the full retirement age, the earnings test no longer applies. Over the next fifteen years, however, the retirement age will gradually increase from sixty-six to sixty-seven, which will increase the number of working seniors subject to the earnings test.
While most people claiming early Social Security benefits stop working, about three-quarters of those who work earn enough to trigger the earnings test. The misinformation makes the system seem outrageous. Consider, say, a sixty-three-year-old who receives a Social Security benefit while continuing to work at an annual salary of $35,000. This worker is likely subject to a marginal federal income tax rate of 15 percent, a Social Security tax of 6.2 percent, and a Medicare tax of 1.45 percent. State income taxes might add another three percentage points, creating a total marginal tax rate of 25.65 percent. On top of this, however, the Social Security earnings test reduces benefits by fifty cents for each dollar by which earnings exceeded the $13,560 threshold. This can amount to a perceived fifty- percentage-point increase in the marginal tax rate on earnings. Put simply, workers perceive that they are taxed at 75.65¢ on the dollar, leaving them less than 25¢ in take-home pay for each dollar they earn.
Anger is the least worrisome of retirees' reactions to the earnings test. Worse is that many workers reduce their earnings or leave the workforce to avoid this "tax." Data shows a "bunching" of earnings just below the $13,560 threshold, indicating that seniors recognize the earnings limit and take steps to avoid triggering it.
Working less is exactly what most seniors should not do. With life spans rising, traditional pensions declining, and Social Security benefits in need of adjustment, working longer may be the best way for individuals to improve their retirement security. Researchers at the Urban Institute estimate that an additional year of work would increase the typical retiree's income by 9 percent; a five-year delay in retirement would increase total income by 56 percent.
But here is the good news that financial columns neglect: the Social Security earnings test is not really a tax. At the full retirement age, Social Security not only stops withholding benefits but increases monthly benefits to replace those taken by the earnings test. The benefit increase at the full retirement age works through the "reduction factors" that Social Security generally uses to reduce benefits for early retirees.
Put in simple terms, the process works as follows: Individuals who claim benefits prior to the full retirement age have their full benefits multiplied by a reduction factor. This factor is generally equal to five-ninths of 1 percent (0.56 percent) for each month the person claimed prior to the full retirement age. This amount is deducted from his benefit. If the person loses a month of benefits due to the earnings test, however, the reduction factor used equals the number of months of early claiming minus the months of lost benefits.
For example, imagine a person who claimed benefits twenty-four months prior to the full retirement age of sixty-six and did not work. This person's benefits would be subject to a reduction factor of 13.3 percent. This reduction in benefits for early claiming would be permanent. But now imagine that this person continued to work, such that he lost six months of benefits due to the earnings test. At the full retirement age, the reduction factor applied to his benefits would be recalculated to 10 percent.
In effect, Social Security treats months in which retirees lose benefits due to the earnings test as months in which they delayed claiming benefits and applies the delayed claiming formula to increase their monthly benefits.
The formula for adjusting benefits for delayed retirement is roughly actuarially fair, meaning that the present discounted value of expected lifetime benefits is the same, regardless of the age that a person retires. Because they use the same formula, the benefit adjustments for the earnings test are also roughly fair. Average total lifetime benefits received are about the same--with or without the earnings test. For that reason, the earnings test should not be viewed as a tax, and individuals should not take extraordinary measures to avoid it.
To test this, I analyzed a hypothetical elimination of the Social Security earnings test using the Policy Simulation Group's GEMINI microsimulation model. The model estimates that for only one in one hundred retirees does the earnings test increase or decrease the present value of total lifetime benefits by more than 1 percent. For the vast majority of retirees, the earnings test does not reduce benefits; it merely delays them. The test has almost no effect on the long-term solvency of the Social Security program, consistent with the fact that average lifetime benefits are unchanged by the earnings test. To the degree that removing the earnings test encourages more Americans to work, Social Security's finances would improve.
Social Security Secrecy
Why is this good news such a secret? One reason is lack of publicity by SSA. Agency publications, which many financial advisers rely on, explain the reductions in current benefits in great detail but rarely mention the increase in future benefits. While visitors to Social Security field offices are given extensive explanations of how the earnings test reduces current benefits, they are generally not informed of the recalculation of benefits that takes place at the full retirement age--perhaps the single most important piece of information for most people to know.
When I worked at SSA and was quoted in the Wall Street Journal explaining the benefit adjustment at the full retirement age, at least one reader called the agency's toll-free number to inquire about this statement; he was told it was untrue and that there was no adjustment to benefits at the full retirement age.
Financial columns are not much better. For instance, a recent personal finance column in the Wall Street Journal answered the question of how Social Security applies the earnings test with a detailed explanation of how benefits are reduced but no mention of how benefits would later be increased. Likewise, the National Committee to Preserve Social Security and Medicare has a Social Security advice column containing a number of detailed questions and answers about the earnings test, but none notifies questioners that benefits lost to the earnings test will be returned beginning at the full retirement age. This information renders all the complex questions regarding the earnings test practically moot, as individuals ask those questions principally because they are under the impression that benefits lost to the earnings test are lost forever.
Awareness of the true state of affairs is changing. SSA is updating its publications about the earnings test as part of a broader modernization initiative by Commissioner Michael Astrue. The updated version of the SSA publication "How Work Affects Your Benefits" contains this new language:
If some of your retirement benefits are withheld because of your earnings, your benefits will be increased starting at your full retirement age to take into account those months in which ben-efits were withheld. As an example, let us say you claim retirement benefits upon turning 62 in 2007 and your payment is $750 per month. Then, you return to work and have 12 months of benefits withheld. We would recalculate your benefit at your full retirement age of 66 and pay you $800 a month (in today's dollars). Or, maybe you earn so much between the ages of 62 and 66 that all benefits in those years are withheld. In that case, we would pay you $1,000 a month starting at age 66.
This important new information eventually will filter through to the public. In the meantime, however, financial advisers and journalists should better inform older Americans about how the earnings test actually works.
Testing the Earnings Test
Given all the confusion caused by the earnings test and SSA's failure to explain how it works, why continue it? After all, it provides all the work disincentives of a steep tax without actually raising any money, it is hard to explain, and its elimination would have little effect on Social Security's long-run solvency. Here are some of the arguments.
Ending the earnings test would increase the number of retirees who work. One recent study found that when the earnings test was lifted on individuals over the full retirement age in 2000, men's hours worked jumped by 12-17 percent. Even larger increases occurred among men with high school diplomas, whose retirement income would especially benefit from additional time in the labor force.
Repealing the earnings test may cause some individuals to claim benefits earlier, however, since some people avoid the test altogether by delaying claiming benefits until the full retirement age. Claiming earlier causes a permanent reduction in benefits for such individuals, potentially increasing the share of retirees in poverty in old age. But one study suggests that individuals who claim earlier would disproportionately be higher-income workers who face little risk of poverty in old age.
A more general point in favor of the earnings test is that once people understand how the earnings test actually works, it may serve a useful social insurance purpose: it limits benefits when retirees have other earnings and concentrates them at the ages when retirees need Social Security the most. The earnings test increases benefits later in life, then, and it can help reduce poverty among the oldest Americans.
The misperceptions about the earnings test offer lessons for three different groups. Seniors should not stop working because of the test. They will not lose benefits, and the extra time on the job could lead to big increases in income when they do retire. Government agencies and financial planners must ensure that the public understands the earnings test; otherwise, they risk pushing people out of the workforce and hurting their retirement income security. Finally, Congress can consider whether the earnings test should be retained as part of a reformed Social Security program.
Andrew G. Biggs (email@example.com) is a resident scholar at AEI. From 2007 to 2008, he was the principal deputy commissioner of the Social Security Administration.
AEI research assistant Adam Paul worked with Mr. Biggs to produce this Tax Policy Outlook.
1. This earnings threshold is increased annually at the rate of average wage growth.
2. In addition, in the year in which a person reaches the full retirement age, the earnings threshold is increased and the test reduced from $1 for each $2 in earnings to $1 for each $3. For individuals reaching the full retirement age in 2008, the annual exempt amount is $36,120. More information on the earnings test is available at www.ssa.gov/OACT/COLA/rtea.html.
3. See Leora Friedberg, "The Labor Supply Effects of the Social Security Earnings Test" (Working Paper 7200, National Bureau of Economic Research, Cambridge, MA, June 1999), available at www.nber.org/papers/w7200 (accessed July 14, 2008). See also David Loughran and Steven J. Haider, "Do the Elderly Respond to Taxes on Earnings? Evidence from the Social Security Retirement Earnings Test" (Working Paper 223, RAND Corporation, Santa Monica, CA, January 2005), available at www.rand.org/pubs/working_papers/WR223/index.html (accessed July 14, 2008).
4. Barbara A. Butrica, Karen E. Smith, and C. Eugene Steuerle, "Working for a Good Retirement" (Retirement Project Discussion Paper 06-03, Urban Institute, Washington, DC, May 23, 2006), available at www.urban.org/Publications/311333.html (accessed July 14, 2008).
5. More information on the process is available in Social Security Administration (SSA), Office of Policy, "Social Security (Old-Age, Survivors, and Disability Insurance) Program Description and Legislative History," Annual Statistical Supplement, 2007 (Baltimore: SSA, 2008), available at www.ssa.gov/policy/docs/statcomps/supplement/2007/oasdi.html (accessed July 14, 2008).
6. For more information, see the Policy Simulation Group's website at www.polsim.com.
7. The principal SSA publication dealing with the earnings test is "How Work Affects Your Benefits." An earlier version neglected to mention the future increased benefits. See SSA, "How Work Affects Your Benefits" (No. 05-10069, February 2004), available at www.adap.net/socialsecurity/How%20Work%20Affects%20Your%20Benefits.pdf (accessed July 14, 2008).
8. Glenn Ruffenach, "The Baby Boomer's Guide to Social Security," Wall Street Journal, November 28, 2007.
9. Kelly Greene, "How Does Social Security Apply Earnings Test?" Wall Street Journal, June 28, 2008.
10. National Committee to Preserve Social Security and Medicare, "Ask Mary Jane," available at www.ncpssm.org/contact/ask/earlyretirearchive/ (accessed July 14, 2008).
11. SSA, "How Work Affects Your Benefits" (No. 05-10069, January 2008), available at www.ssa.gov/pubs/10069.pdf (accessed July 14, 2008). Earlier versions, such as the 2004 edition cited in note 7, lacked this explanation.
12. Gary V. Engelhardt and Anil Kumar, "The Repeal of the Retirement Earnings Test and the Labor Supply of Older Men" (Working Paper 2007-1, Center for Retirement Research, Boston College, Chestnut Hill, MA, December 2006), available at http://crr.bc.edu/working_papers/the_repeal_of_the_retirement_earnings_test_and_the_labor_supply_of_olde.html (accessed July 14, 2008).
13. See, for instance, Jonathan Gruber and Peter R. Orszag, "Does the Social Security Earnings Test Affect Labor Supply and Benefits Receipt?" (Working Paper 7923, National Bureau of Economic Research, Cambridge, MA, September 2000), available at www.nber.org/papers/w7923 (accessed July 14, 2008).
14. See David Loughran and Steven J. Haider, "Do the Elderly Respond to Taxes on Earnings? Evidence from the Social Security Retirement Earnings Test."