The Social Security Earnings Limit Provision

January 9, 1995 • Testimony

Mr. Chairman, I am honored to have the opportunity to discuss the issue of reforming the Social Security earnings test.

The Social Security earnings test provision is anti‐​work, anti‐​senior citizen, and anti‐​tax fairness. The earnings test should not be reformed; it should be repealed. Its major effect is to impose punitive tax rates on senior citizens thus driving them out of the labor force altogether, or to reduce their hours worked.

If the Social Security earnings test cannot be repealed, then we should minimize the damage of this policy by raising the income threshold as in the Republican contract, and by lowering the amount of lost benefits for each dollar earned. Currently, the earnings test leads to a 33 cents in lost benefits for each additional dollar earned; that should be lowered to 20 cents.

Given the trend of an aging workforce in America, such anti‐​work policies are becoming increasingly economically unaffordable.

Indeed, although reforms in the Social Security earnings test are opposed by some on the grounds that these tax changes would injure the financial status of the Social Security system, the truth is that unless ways are found to keep healthy and productive senior citizens in the workforce and off of this and other benefit programs, Social Security will be financially insolvent when the baby boom generation reaches retirement age. In any case, even if one assumes that the Congressional Budget Office static scoring is correct–which I believe it is not–and the GOP earnings test provision loses $7 billion over five years, that could be paid for easily with reforms in Social Security that must be made in any case.

One of the most troubling economic trends in America is that American workers have been retiring from the labor force at earlier and earlier ages. This trend has occurred over the past 30 years and is shown in the attached Table 1. It shows that in 1963 only 20 percent of American men aged 62 or over were retired; today that percentage is nearly 50 percent. At age 65, less than half of American men were retired in 1963, today, three‐​quarters of all men are retired.

It is true that these earlier retirement ages are in part a function of an increasingly wealthy society where individual workers with higher living standards begin to prefer more leisure over more work. The trend is also a function of private pension policies and private sector retirement rules. But there is no question that Social Security has played a substantial role in encouraging retirement from the workforce. The big spikes in retirement by age occur when people reach of 62, i.e. when they become eligible for early retirement under Social Security, and at age 65, when they become eligible for full benefits under Social Security.

One reason this trend is troubling from an economic standpoint is that over this same time period, life expectancy has risen substantially. In 1960 life expectancy was 70. Today, it is 76. Life expectancy could easily reach 80 over the next thirty years. Moreover, life expectancy for those who reach the age of 65 today, is about 83. In other words, the average person who goes on Social Security collects government benefits for 18 — 20 years.

This combination of earlier retirement, longer life spans, and the aging of the baby boom cohort, is leading America head‐​on toward a financial trainwreck of spectacular proportions. The ratio of workers to retirees in America has dropped from 15 to 1 in 1950; to 3 to 1 today; to 2 to 1 by 2025. (See Table 2) No nation can prosper under such demographic conditions. The wagon will become too heavy to pull. To the extent we can adopt policies that can entice healthy and productive senior citizens to prolong their work years, we in a sense take that person out of the wagon and create an additional hand to help pull.

Skeptics of repealing or reforming the earnings test contend that the current policy does not deter senior citizens from working. Evidence from tax rate reductions in the 1960s and 1980s suggest otherwise. Both the Kennedy and Reagan supply side tax rate reductions had a significantly positive impact on work effort. Much of the literature on this subject is covered in Larry Lindsey’s book The Growth Experiment. What is especially interesting and relevant about the economic literature on labor market responses to changes in tax rates is that it is women and the elderly who are among the most responsive to changes in tax policy.

* An Urban Institute study by Robert Haveman found that “persons 62 and over” had “high responsiveness to changes in work incentives” after the Reagan tax cuts. (See: Robert Haveman, “How Much Have the Reagan Administration’s Tax and Spending Policies Increased Work Effort?” in Charles L. Hulten and Isabel Sawhill, The Legacy of Reaganomics, Urban Institute, 1984)

* A Brookings Institute study on retirement and earnings data from 1969–79 concluded: “The elimination of the [earnings] test is estimated to raise the work effort of average retirees over age 62 by 30 to 40 percent.” (See: Gary Burtless and Robert A. Moffitt, “The Effect of Social Security Benefits on the Labor Supply of the Aged,” in Henry Aaron and Gary Burtless, Retirement and Economic Behavior, Brookings Institution, 1984)

* Economists Stanley Masters and Irwin Garfinkel find that the elderly’s work patterns are influenced by government benefit levels. (See: Stanley Masters and Irwin Garfinkel, Estimating the Labor Supply of Income Maintenance Programs, New York, Academic Press, 1977)

Perhaps the most compelling evidence that senior citizens’ work effort changes in response to Social Security policies is that this was one of the intentions of the program in the first place. An explicit goal of Social Security when it was originally conceived by Franklin Roosevelt in the early 1930s was to drive the elderly out of the workforce. Social Security was passed in 1935 during an era of high unemployment and widespread Keynesian economic illiteracy in Washington where policymakers believed that the way to create more jobs and prosperity was to have less people working. Senator Robert Wagner, the original chief sponsor of the Social Security Act advised his colleagues to voter the bill because “the incentive to the retirement of elderly workers will improve efficiency standards, [and] make new places for the strong and the eager.”

The Social Security earnings test was invented to serve as a financial punishment to senior citizens who chose to continue to work beyond the age of 65 rather than freeing up jobs for younger workers. The very first income limit was set at $15 per year.

The United States can no longer afford policies based on economic illiteracy. Today, we need to adopt policies that encourage senior citizens to work, or at least, which do not push them involuntarily into retirement. And we need to do so with a sense of urgency. Along these lines, Congress should:

1) Allow workers to gradually opt‐​out of Social Security through a Super‐​IRA approach and corresponding reduction in payroll taxes. This would allow workers to manage their retirement earnings and pensions to match their individual choices.

2) Gradually raise the retirement age and early retirement age for Social Security by four months per year until early retirement age reaches 66 and normal retirement 70.

3) Reduce benefits for those choosing early retirement.

4) Repeal the 1993 increase in the tax on Social Security benefits;

5) Alter the delayed retirement credit to encourage later retirement and allow the credit to be expanded past the age of 70;

6) Repeal the earnings test.

Even if the earnings test cannot be repealed, the reform proposed in the Republican contract is not, in my opinion, the best way to reform this onerous law. Rather than, or in addition to, raising the income threshold, the best way to reduce the perverse incentives of the Social Security earnings test would be to reduce its marginal impact on all working–or potentially working–senior citizens. This could be accomplished by changing the income test formula. Currently, that formula is $1 of lost benefits for every $3 of earnings. This should be changed to $1 of lost benefits for every $5 of earnings. This reduces the increased marginal tax rate on senior citizens as a result of the earnings test from 33 to 20 percent.

In 1989 then‐​Senator Lloyd Bentsen proposed changing the formula to $1 of lost benefit for every $4 of earnings. Certainly this new reformist Congress can be at least as bold as Bill Clinton’s first Treasury Secretary.

Paying for these changes should not deter immediate action. The earnings test changes under consideration could be paid for by raising of the retirement age by four months per year as recommended above. This is a change that Congress should do as a matter of essential Social Security reform in any case.

The elderly have been called the people that the supply side revolution forgot. Today, we impose unfair–even punitive–tax rates on our senior citizens today to discourage them from working. The earnings test is particularly offensive because it is only a tax on work–not on unearned income. When all of the existing federal and state income taxes are combined with the special earnings taxes imposed on the elderly today, marginal tax rates for these Americans often reach 60, 70 and even 80 percent. Once upon a time, the nation could afford such economic folly. That time has long passed.

                           TABLE 1           Percentage of Men Aged 55-70 in Retirement            1963         1970      1985  60       12           16        29  61       16           19        34   62       20           26        49  63       24           31        55   64       28           36        58  65       46           50        70  66       57           55        74  67       61           61        76  68       67           62        80  69       67           66        80  70       73           70        84    Source: Rita Ricardo-Campbell and Edward P. Lazear, Issues in  Contemporary Retirement (Stanford: Hoover Institution, 1989.   Based on Bureau of Labor Statistics data.                           TABLE 2    The Growing Dependency Crisis of Social Security                 Workers Per Social Security Recipient        1950          16  1993           3  2025 est.      2  2040 est.    1.8  2070 est.    1.3    Source: Social Security Trustees Report, 1993.  
About the Author