Social Security Earnings Reduce Work by the Elderly

08/01/1999
Summary of working paper 7200
Featured in print Digest

...eliminating the earnings test would bolster labor supply ... at a small fiscal cost.

The Social Security earnings test generates some of the highest marginal tax rates in the economy today, according to an NBER Working Paper by Leora Friedberg. Previous research suggested that the earnings test has only a minor impact on behavior; Friedberg's work casts doubt on these earlier studies. In The Labor Supply Effects of the Social Security Earnings Test (NBER Working Paper No. 7200), she finds that the "earnings test is an important consideration in understanding the decisions of the elderly who continue to work."

The Social Security earnings test, introduced in 1939, reduces Social Security benefits after earnings pass a threshold amount. Once the beneficiary earns more than that, his or her benefits are reduced at a rate proportional to additional earnings, until benefits are exhausted. The original intent of the earnings test was to move older workers out of the labor force--beneficiaries lost an entire month's benefits when monthly earnings exceeded $15.

Since the 1950s the earnings test gradually has been relaxed as the emphasis has shifted toward encouraging work and saving. Still, in 1998 the earnings test slashed $1 in benefits for every $2 in earnings above $9,120 for those in the 62-64 age bracket--a 50 percent tax on wages. A beneficiary aged 65-69 with earnings above $14,500 faced a less restrictive 33 percent earnings test tax rate. (Once a person turns 70 the earnings test no longer applies.) Many beneficiaries appear unaware that their future benefits will be raised if they lose benefits today. However, the delay in benefit payouts results in little or no long run cost saving to the government.

Friedberg reaches her conclusions by studying several changes to the earnings test rules initiated between1978 and 1990. These changes applied to beneficiaries of some age groups and not to others. Comparing the reactions of beneficiaries before and after the changes, using the unaffected groups to control for other changes in the labor supply, Friedberg isolates the response to the earnings test using a combination of methods.

Beneficiaries in the Current Population Survey satisfy the strongest prediction: many keep their earnings just below the exempt amount, and this bunching shifts with the changes in the earnings test rules. Furthermore, the clustering disappears when the earnings test is eliminated. The clustering is evidence that the earnings test leads some individuals to limit their labor supply.

Using more formal methods, Friedberg predicts that eliminating the earnings test would bolster labor supply, raising average hours worked by 5.3 percent for those currently at or above the exempt amount, and at a small fiscal cost. In contrast, a slight decrease in labor supply among 65-69 year old men is predicted from the recently legislated increase in the exempt amount to $30,000, which will occur by 2002. The positive effect on hours worked by low earners is offset by a negative effect for high earners, since the exempt amount will bind for a new group of people with higher earnings, who would reduce their labor supply considerably.

These findings reaffirm the deleterious effect of high taxes on older workers, who are more sensitive to tax and transfer rules. According to the author, the results suggest a potentially severe negative effect on labor supply if Social Security benefits are means tested, a proposal that has gained recent attention. Another area of concern is 62-64 year olds, who continue to face a more restrictive earnings test, with a 50 percent tax rate. These tighter rules will be extended to 65 and 66 year olds as the normal retirement age begins to rise from 65 to 67 in 2000. With the median retirement age falling from 65 in the 1970s to 62 today, the earnings test is growing more binding for the younger group.

-- Marie A. Bussing-Burks