Unemployment and Older Workers

August 29, 2007 (RL32757)

Contents

Figures

Tables

Summary

This is one in a series of papers that explore issues of our aging society. This report examines how unemployment has a different impact on the older worker. As workers age, negative—but previously temporary—events such as unemployment may push otherwise firmly entrenched workers out of the labor force. Older workers are less likely than others to experience a spell of unemployment, but those older workers who do experience unemployment have a higher incidence of withdrawing from the labor market. Some studies have found that unemployment in older workers contributes up to a one-third increase in the probability of retirement.

The pattern of unemployment leading to unexpectedly early retirement is not a new development. Rather, it is the relative scale of the phenomenon to the overall workforce that is new. The shifting demographics of the workforce have made what was once a fairly small policy issue grow in importance.

Depending on the age of the older unemployed workers, new alternative income sources such as retirement benefits and early Social Security benefits may be used while previous pillars of support such as unemployment compensation become less helpful in replacing income. Facing lowered expected wages and lower chances of rehire, older workers find themselves faced with new decisions: should they search for a new job; create a new job through self-employment; spend down any non-retirement personal savings they may have accrued during their working years; opt to withdraw funds from retirement savings plans, and/or opt to receive Social Security benefits?

One policy issue is how to address the inherent tensions among the Unemployment Compensation (UC) system, alternative working arrangements, and eligibility for and receipt of various types of retirement income including Social Security benefits. Another is how federal programs might better balance providing income support for older Americans with providing appropriate work incentives for those who would prefer to remain engaged in some type of work. This report will be updated as needed.


Unemployment and Older Workers

Overview

This report examines how unemployment has a different impact on the older worker. As workers age, negative—but previously temporary—events such as unemployment may push otherwise firmly entrenched workers out of the labor force. Older workers are less likely than others to experience a spell of unemployment, but those older workers who do experience unemployment have a higher incidence of withdrawing from the labor market. The workers find themselves faced with new decisions: should they search for a new job; create a new job through self-employment; spend down any non-retirement personal savings they may have accrued during their working years; opt to withdraw funds from retirement savings plans and/or receive Social Security benefits? Some studies have found that unemployment in older workers contributes up to a one-third increase in the probability of retirement.1 Once these workers have entered retirement, they then face a different set of incentives that may make it less rewarding for these retired workers to return to the labor market.

The pattern of unemployment leading to unexpectedly early retirement is not a new development. Rather, it is the relative scale of the phenomenon to the overall workforce that is new. The shifting demographics of the workforce have made what was once a fairly small policy issue grow in importance.

One policy issue is how to address the inherent tensions among the Unemployment Compensation (UC) system, alternative working arrangements, and eligibility for and receipt of various types of retirement income. Another is how federal programs might better balance providing income support for older Americans with providing appropriate work incentives for those who would prefer to remain engaged in some type of work.

Options for Older Unemployed Workers

When older workers find themselves unemployed, they face a substantially different set of decisions to make than if they were a few years younger. While all qualified unemployed workers may receive unemployment compensation, older workers may replace lost earnings (conditional on age) with retirement savings and pension programs. For example, workers who are at least 59½ years old may begin to take funds out of their IRA based savings without incurring the 10% early withdrawal penalty. At age 60, qualified widows and widowers may begin to receive their widow(er)s' Social Security benefits. At age 62, all workers may receive a reduced, or "early," Social Security benefit. Finally, at some age between 65 and 67 (depending on their birth year) workers can receive their "full," or non-reduced, Social Security benefit.

Thus, depending on the age of the older unemployed workers, new options are incorporated into their labor force participation decisions and UC benefits for unemployed workers become relatively less important in income replacement. Previously, these workers were motivated to stay in the labor market and use UC benefits to give themselves adequate time to engage in an effective job search. As they age, the older workers' job searches change. In order to attain new employment, older workers may (1) need to form new (potentially lower) expectations of wages that reflect the loss of tenure and firm specific knowledge, (2) need to obtain new skills through training programs in order to remain competitive in the labor market, or (3) wish to use a spell of unemployment as a bridge to easing into retirement.

The Aging Demographics of the Population

The pattern of unemployment leading to unexpectedly early retirement is not a new development. Rather, it is the scale of the phenomenon relative to the overall size of the workforce that is new. The shifting demographics of the workforce have made what was once a fairly small policy issue grow in importance. Over the next 25 years, the age profile of the U.S. population and the U.S. workforce will change dramatically. Members of the "baby boom" generation, those who were born between 1946 and 1964, are currently between 40 and 59 years old. This boom creates a peak in the U.S. population chart presented in Figure 1. As the baby boom ages, the demographic profile of the United States substantially alters, as the share of the population aged 35-55 decreases and the share of older American increases.

The shift in the age profile has particular implications for older Americans and in the 55 to 69 year old demographic group. In 2000, this group comprised 12.0% of the total population. Predictions for 2030 increase its proportion to 18.0%. This 6 percentage point increase is equivalent to just under a 50% growth in the demographic group's share of the population. In 2050, the proportion of the 55 to 69 year olds relative to the total population is expected to fall to 15.8%; however, this is still more than a 30% increase from 2000.

Figure 1. Age Profile of the U.S. Population 2000 to 2050

Source: CRS graph using estimates from U.S. Census Bureau National Population Projections, Middle Series. http://www.census.gov/population/www/projections/natsum-T3.html/.

The demographic shift in the potential workforce will not affect all occupations equally. Research suggests that the educational and health services industries will be disproportionately affected when workers in those industries begin to retire since they currently have an age distribution which is older than that of other industries.2

Federal policy has recognized the problems that this demographic trend may create by encouraging workers to extend their labor force participation through increasing the full-retirement age for Social Security benefits while decreasing the rate at which earnings offset benefits3 and through abolishing mandatory retirement ages for most jobs.4

These federal efforts to encourage workers to work longer have not been mirrored by supports to keep older unemployed workers in the workforce.5 With a few exceptions, the UC system generally does not treat common older worker employment patterns favorably. Depending on the particular state, a desire for part-time work or self-employment may disqualify a worker from receiving UC benefits. At the same time, retirement income may lower UC benefits, which are especially important for those workers who are bridging into retirement. This retirement income offset in the UC program will decrease the amount of income connected with an older worker's job search and may cause unemployed older workers to completely exit the labor market. As the ratio of older workers to younger workers increases, the impact on the economy of older worker responses to spells of unemployment increases.

Labor Force Participation

Labor force participation rates measure the ratio of the civilian non-institutionalized civilian population that (1) has a job or (2) has searched for a job in the previous four weeks. Labor force participation rates climb rapidly in the population aged 20-24 as formal schooling is finished and students enter the labor market. Between the ages of 25 and 54, labor force participation is fairly consistent and high with approximately 83% being actively engaged in the labor market. Rates begin a substantial decline as workers reach age 55. Just under 62% of those between the ages of 55 and 64, and just over 13% of those 65 and older are actively working or searching for work.

The unemployment rate is the ratio of those workers who did not have a job but had searched for a job in the previous four weeks to all workers active in the labor market (either searching or employed). Typically, unemployment is a large social concern for the youngest and least skilled of workers. As workers age and gain skills and tenure, the likelihood that they are unemployed decreases. In fact, one study found that the aging labor force accounted for half of the decrease in unemployment in the 1990s.6 In 2003, 4.8% of prime-age (35-54 years of age) workers were unemployed on average. Older workers continued to have lower rates of unemployment: 3.9% for those aged 55-64 and 3.6% for those 65 and older. However, between 2000 and 2002, older workers experienced rising unemployment rates that were greater in relative magnitude than those for younger workers during the same period.7

Table 1 lists the participation and unemployment rates for workers aged 35 and older in 2003. Notice that labor force participation rates are heavily influenced by the age at which older Americans become eligible for programs such as early retirement for Social Security (age 62). The unemployment rates do not vary substantially for older Americans; this may be in part due to older unemployed workers choosing to declare themselves out of the labor force. For example, it is unlikely that all of the 56.2% of the Americans aged 62-64 who are out of the labor force would prefer not to have a job. Some may choose to declare themselves retired rather than admit to being unemployed. There is weak evidence that this may be true from the otherwise unexplained 10% decrease in the reported unemployment rate from 4.4% for Americans aged 60-61 to 4.0% for those aged 62-64.

Table 1. Annual Average Labor Force Participation Rates and Unemployment Rates in 2005, by Age

Age

Labor Force
Participation Rate

Unemployment
Rate

35-44

83.8%

3.9%

45-54

81.7%

3.5%

50-54

79.8%

3.3%

55-59

71.4%

3.4%

60-64

51.6%

3.2%

65-69

28.3%

3.4%

70-74

16.3%

3.7%

75+

6.4%

3.3%

Source: Current Population Survey, 2005.

Table 1 demonstrates that workers aged 35-54 (prime-age workers) have the highest labor force participation rates with more than 80% of the non-institutionalized civilian population active in the labor market. There is a rapid decline of labor force activity as workers reach their late 50s (55-59)—the approximate age when many buy-outs and some private pension plans allow for early retirement. This is reflected in a pronounced drop in labor force participation of 8.4 percentage points (from 79.4% to 71.4%) from what was reported for workers in their early 50s. This drop is followed by a subsequent 19.8 percentage point drop as workers enter their early 60s, when widow(er)s' Social Security benefits (age 60) and early retirement (age 62) become available.

In the span from the 50-54 age group to the 60-64 age group, the labor force participation rate falls to just under two-thirds of what it had been. This is not, perhaps, surprising given that at age 62 a reduced Social Security benefit becomes available. This reduction in Social Security benefits has been constructed to be actuarially fair on average. (That is, it does not set a penalty for early participation but rather reduces the Social Security benefit so that the discounted value of the benefit is equal to that of the full retirement benefit for an average worker.)

Definitions of Retirement and Labor Force Participation

For the purposes of this report, the term worker refers to all participants in the labor market. Individuals who are not employed or not actively searching for a job and are at least 55 years old are labeled as being retired, regardless of pension or Social Security benefit income. This is substantially different from the definition often used in retirement analyses. Generally, retirement is defined with reference to two characteristics: nonparticipation in the labor force, and receipt of income from pensions, Social Security, or other retirement savings vehicles. An individual who does not work for compensation and who receives income only from these retirement-based income replacements would meet this definition of retirement.

Between these two extremes, however, are those who might be considered to have retired based on one part of the definition but not the other. For example, of recipients of employer pensions and retirement savings plans, 40% of the male recipients and 34.5% of the female recipients were employed.8 Sometimes, these workers are employed in " bridge" jobs that help to transition them from full labor force participation into retirement. Other workers may be fully employed in another career while enjoying the retirement benefits from a previous job.9 In such cases, retirement from a particular job does not necessarily signal that the worker has withdrawn from the workforce and moved into the strict definition of retirement.

On the other hand, some of those who retire from full-time employment continue to work part-time to supplement their retirement-based income but receive the majority of their income through Social Security, pensions, and savings. These workers would typically be classified as retired, even though they continue to engage in paid employment. As these examples suggest, not everyone who receives pension income is retired, and some people who work for pay may be considered retired.

The Process of Retirement

Generally, the transition from the labor market into full retirement is an orderly and fully planned process. For many workers, their retirement plans may involve an abrupt move from being full-time workers to enjoying a full retirement without any further labor force participation. Other workers may have retirement plans for a gradual transition into retirement.10 These transitional steps may include creating alternative work arrangements (moving from full-time to part-time status) or switching to less demanding positions—which may not be in the same occupation or industry (a bridge job). Furthermore, some individuals may churn or cycle from bridge jobs back to "career" jobs and then gradually transition into full retirement with no labor force participation.11

Unemployment and the Disruption of the Retirement Process

For some workers, however, an unexpected loss of employment may disrupt their retirement plans and push the workers into an earlier—and unexpectedly abrupt—retirement as their UC benefits are exhausted (or if they choose not to receive UC benefits, as they decide to stop working earlier than they had originally intended). These unplanned retirements result in permanent decreases in retirement benefits as the workers begin drawing benefits earlier (and in the case of Social Security, early retirement penalties occur) while discontinuing the accumulation of retirement assets (possibly leading to lower total retirement assets).

Unemployment and Older Workers

Many factors that help explain why unemployment has such a different impact on older workers than on younger workers. Empirically, research suggests that older workers have an increased difficulty in obtaining a similar position at the same level of compensation, because their higher wages often are difficult to fully replace through reemployment. Workers lose their wage premium attributed to longevity, tenure, or employment contract.12 Older workers also may lose the wage premium for skills specific to their previous firm. If workers find new jobs in different industries, they lose the premium for that industry-level knowledge as well.13

Unemployment Compensation

UC in the United States is provided through a federal-state system that provides temporary partial wage replacement to active job seekers who are involuntarily out of work. Older workers generally have long-term, stable work histories and are less likely to be unemployed than other workers. As a result, older workers who lose their jobs are more likely to receive UC benefits and have higher expected benefits because of their stronger work histories and greater understanding of the UC system.14 Older workers are also more likely to exhaust their benefits.15

Dislocated Older Workers

Studies suggest that older unemployed workers take longer to find new jobs; are less likely to find a job; and, their new jobs replace a smaller fraction of previous earnings. Older workers are much more likely to be dislocated from their jobs. That is, they are more likely to have lost a job where they had long tenure and the separation from the employer is permanent.16 Dislocated workers have a lower chance of finding new employment. In addition, those who do find employment typically earn substantially less than they did in their previous job.17 For older workers a job dislocation has more of an impact on earnings than for younger workers. A significant proportion of their previous high salaries may be attributed to job tenure; thus, wages from new jobs may be substantially lower. Likewise, facing lower levels of replaced earnings, older workers are also less likely to continue to work after job dislocation. Subsequently this increases their chances of early withdrawal from the labor market.18

Encouraging and Discouraging Older Workers to Remain in the Labor Force

If older workers do not qualify for, cannot afford, do not opt for, or have exhausted support from the UC program, they may turn to other forms of income support intended for retirement purposes. These earnings substitutes listed in Table 2 include participation in early Social Security benefits and early withdrawal of personal retirement funds. These substitutes often change the structure of work incentives for older Americans and may contribute to a decision to permanently withdraw from the labor market.19

Table 2. Age at Which Potential Earnings Substitutes Become Available and Changes in Social Security, Pension, and Retirement Savings Incentives

Age

Earnings Substitute

Changes in Incentives

55

IRA savings purchased as life-time annuity

Removal of 10% early withdrawal penalty.

59½

IRA savings

No early withdrawal penalty.

60

Widow(er)s Social Security benefits

Earnings up to maximum ($97,500 in 2007)
subject to employment taxes. (Shared
between employer and employee.) Each $1
of earnings over $12,960 (2007) causes
benefit reduction of $.50. Early retirement
benefit reduction applies.

62

Early Social Security retirement benefits

Earnings continue to be subject to
employment taxes. Each $1 of earnings over
$12,960 (2007) causes benefit reduction $.50.
Early retirement benefit reduction applies.

Year of full-
retirement age

Social Security retirement benefits

Earnings continue to be subject to
employment taxes. Early retirement benefit
reduction applies. Each $3 of earnings over
$34,440 (2007) causes benefit reduction of $1
benefits. Only earnings before the month of
full retirement age are counted against the
$34,440.

Full-retirement age
(65-67) based on
birth year

Social Security retirement benefits

Earnings continue to be subject to
employment taxes. Earnings do not reduce
benefit. Delayed retirement credit applies if
Social Security is deferred.

Full-retirement
age (65-67) based on birth year

Aged SSI

Needs-based benefit. After $65 per month,
SSI benefits are reduced $1 for every $2 of
earned income. For unearned income such as
pensions, Social Security benefits, and
workers' compensation, after the first $20 per month each $1 reduces the SSI benefit by $1.

70

Must begin collection of Social Security benefits

Earnings continue to be subject to
Employment tax. Delayed retirement
credit is terminated.

70½

Must begin to take IRA distributions

 

Source: CRS compilation of rules reported in the Social Security Administration's 2005 fact sheet http://www.ssa.gov/legislation/2005_factsheet.doc, and CRS Report RL31770, Individual Retirement Accounts and 401(k) Plans: Early Withdrawals and Required Distributions, by [author name scrubbed].

Although the earnings substitutes have disincentives for labor force participation, there are several other public policies and programs that encourage and aid older workers in their continued participation in the labor force. These policies and programs reflect several underlying—but at times competing—objectives: alleviating Social Security financing pressures, rewarding and encouraging longer labor force attachment, and eliminating employment discrimination based upon age. The following examples of incentives and disincentives do not comprise an exhaustive list of all legislation and programs that encourage or discourage work by older Americans, but rather represent the keystones to this approach.

Incentives to Continue to Work

Age Discrimination in Employment Act of 1967 (ADEA, P.L. 90-202)

This act protects workers (40 and older) from age discrimination in their jobs. It eliminated mandatory retirement age for most workers except where age is a bona fide condition of employment. An example of a position where age is considered to be a bona fide condition includes many (but not all) protective services workers. Until a 1986 amendment (P.L. 99-592), workers over the age of 70 were not protected by ADEA.

Unemployment Compensation

For older workers, the UC system is a source of income security, just as it is for younger workers. Unlike the case of younger workers, however, older workers' UC benefits become a factor in the decision to retire. Unlike receiving Social Security benefits or certain other retirement income where benefits may be permanently reduced, receiving UC benefits does not have permanent adverse effects on future UC benefits. For example, after approximately 15 to 18 months of reemployment, a worker's UC benefit is strictly determined by the new employment activity. However, workers who are bridging to retirement through part-time work or self-employment are sometimes excluded from receiving UC benefits and workers receiving some pension or other retirement benefits may have their UC benefits substantially offset. Table 3 summarizes how states' UC programs treat factors that might impact older workers: part-time work, self-employment, pension income offsets, and Social Security income offsets.

In most states' UC systems workers who have had their hours reduced or who are working short-term in part-time jobs while looking for a permanent full-time job are able to receive some UC benefits. To encourage workers to remain in the labor force, all states disregard some earnings as an inducement to take part-time work. The worker's UC benefit will generally equal the difference between the weekly benefit amount and earnings plus a small allowance.

Table 3. Treatment of Typical "Bridging" Older Worker Labor Force Behaviors in State Unemployment Compensation Systems, January 2007

State

Explicitly May
Search for Part-time Only

Self-employment Assistance
Program

Compulsory Retirement Is
Treated as "Good Cause"

UC Offset of
Pension Only If
Period in Base-
year Increased Pension Incomea

Disregard Portion
of Employee Contribution to
the Pensionb

No UC Offset
of Social Security Benefits

Alabama

 

 

 

Y

 

Y

Alaska

 

 

Y

Y

Y

Y

Arizona

 

 

Y

Y

Y

Y

Arkansas

Y

 

Y

 

Y

Y

California

Y

 

Y

Y

Y

Y

Colorado

Y

 

Y

 

 

 

Connecticut

 

 

Y

Y

Y

Y

Delaware

Y

Y

Y

 

Y

Y

District of Columbia

Y

 

 

 

Y

 

Florida

Y

 

Y

 

Y

Y

Georgia

 

 

 

Y

Y

Y

Hawaii

Y

 

Y

Y

Y

Y

Idaho

 

 

Y

 

Y

Y

Illinois

Y

 

Y

 

Y

 

Indiana

 

 

Y

 

 

Y

Iowa

Y

 

Y

Y

Y

Y

Kansas

Y

 

Y

Y

Y

Y

Kentucky

 

 

Y

Y

Y

Y

Louisiana

Y

Y

Y

 

 

 

Maine

Y

Y

 

Y

Y

 

Maryland

 

Y

 

 

Y

Y

Massachusetts

Y

 

Y

Y

Y

 

Michigan

 

 

Y

 

Y

Y

Minnesota

Y

 

Y

 

 

 

Mississippi

 

 

Y

 

 

Y

Missouri

 

 

Y

Y

 

Y

Montana

Y

 

Y

Y

Y

Y

Nebraska

Y

 

Y

 

Y

Y

Nevada

 

 

Y

Y

Y

Y

New Hampshire

Y

 

Y

Y

Y

Y

New Jersey

Y

Y

Y

Y

Y

Y

New Mexico

Y

 

Y

 

Y

Y

New York

Y

Y

Y

Y

Y

Y

North Carolina

Y

 

 

 

 

Y

North Dakota

Y

 

Y

Y

Y

Y

Ohio

 

 

Y

 

Y

 

Oklahoma

Y

 

Y

Y

Y

Y

Oregon

 

Y

Y

 

Y

Y

Pennsylvania

Y

Y

Y

Y

Y

Y

Rhode Island

 

 

Y

Y

Y

 

South Carolina

 

 

 

 

Y

Y

South Dakota

Y

 

Y

 

Y

 

Tennessee

 

 

Y

Y

Y

Y

Texas

 

 

Y

 

Y

Y

Utah

 

 

Y

Y

 

 

Vermont

Y

 

 

 

Y

Y

Virginia

 

 

Y

 

 

Y

Washington

Y

 

Y

Y

Y

Y

West Virginia

 

 

 

Y

 

Y

Wisconsin

Y

 

 

Y

Y

Y

Wyoming

Y

 

 

 

Y

Y

Total

30

8

40

27

39

40

Source: CRS compilation of rules reported in Comparison of State Unemployment Insurance Laws 2007, U.S. Department of Labor, Employment and Training Administration, Office of Workforce Security, 2007.

Notes: "Y" indicates that the state has this policy/program.

a. UC benefits are offset only by the pension income that can be directly attributed to the base-year months used in the calculations for UC benefits.

b. UC benefits are offset only by the pension income that can be attributed to employer contributions.

Several states have a Self-Employment Assistance (SEA) program to aide workers in creating their own jobs by creating small businesses. To be eligible for SEA payments, workers must be eligible for UC, permanently laid-off from their previous jobs, identified as likely to exhaust their benefits, and participating in self-employment activities such as entrepreneurial training, business counseling, and technical assistance.20 Workers enrolled in an SEA program receive weekly self-employment payments (the same amount as the workers' regular UC benefits) while working full-time on starting a business.

Employment Service (ES)

The ES21 is linked to the UC system; generally, state laws mandate registration and participation in ES programs in order to remain eligible for UC benefits. The ES system provides employment and training programs and services to workers through a joint federal-state partnership.

Two federal ES programs attempt to aid older workers remaining in the workforce. The Senior Community Service Employment Program22 (SCSEP), places eligible applicants into a non-profit, public service, or community service agencies where they receive on-the-job training. The second program, the Alternative Trade Adjustment Assistance (ATAA) Program for Older Workers, was established as an alternative to the benefits offered under the regular Trade Adjustment Assistance (TAA) program. Participation in ATAA allows older workers who are eligible for Trade Adjustment Assistance, but for whom retraining may not be appropriate, to accept reemployment at a lower wage and receive a wage subsidy. Eligible workers age 50 or older who obtain new employment at wages of less than $50,000 within 26 weeks of losing their job may receive a wage subsidy of 50% of the difference between their old and new wages, equal up to $10,000, for up to a two-year period.

Social Security

Before December 31, 1999, workers who were under age 70 faced steep earnings tests and substantial deductions in their benefits. The Senior Citizens Freedom to Work Act, 2000 (P.L. 106-182) eliminated the retirement earnings test23 for those at or above full-retirement age, and relaxed the earnings test for those between 62 and full-retirement age. Thus, workers at or above full-retirement age are now able to earn unlimited wages without relinquishing Social Security benefits. Those workers who are receiving early retirement benefits now have a greater amount of money they may earn before their Social Security benefit is subject to reduction and the rate at which their Social Security benefit is offset is smaller. Workers who are at or above full retirement age (age 65 and six months if born in 1957) may keep all of their Social Security benefits, with no earnings tests. Workers younger than the full-retirement age for the entire year, have an earnings limit. For every $2 earned above $12,960 (in 2007) there is a $1 deduction in the benefit. The earnings test is altered if a worker will achieve full-retirement age during the year. In this case, for every $3 earned above $34,440 there is a $1 reduction in benefits until the month the worker achieves full retirement age at which point there is no reduction in benefits.24

The Social Security Amendments of 1983 (P.L. 98-21) gradually increased full-retirement age for those born after 1937 from 65 to 67 (for those born after 1959). As a result, workers born after 1937 effectively face lower Social Security benefit levels than workers born before 1938. Thus, a worker would also find it more expensive to choose early retirement at age 62, forgoing a greater percentage of his or her full-retirement benefit. An expected result of this degradation of benefit level would be that workers would work longer and save more to replace their reduced benefits.

The changes in the earnings test and the increase in the full-retirement age are fairly new and have only recently been analyzed. Before the elimination of the earnings test for individuals at or above the full retirement age, most studies, such as that by produced by Gruber and Orzag (2000) suggested that eliminating the test would have a modest effect on labor supply. One study, Friedberg (2000), predicted that eliminating the earnings test would increase the hours worked by 65-69 year old men by about 5%.25 The first study to investigate the preliminary effects of removing the test, Song (2003), concludes that while the removal increased hours worked for those with higher earnings, the effect on lower earners is not statistically significant and there is no clear evidence of the effect of the removal on the labor force participation rate. But others have found no evidence of change in labor supply.26

Pension and Retirement Savings

Recognizing that individuals might need to access their retirement funds before retirement, Congress permits early withdrawals of some pension savings without penalty in certain situations.27

401(k)s

Individuals may take "hardship distributions" from their 401(k) accounts without penalty in the case of "an immediate and heavy financial need of the employee." These needs include expenses for medical care of the plan participant or family members, purchase of a principal residence, college tuition and related expenses, expenses to prevent eviction or foreclosure on a principal residence, and funeral expenses. A spell of unemployment is not considered to be a reason for a 401(k) hardship distribution.

Roth IRAs

Withdrawals of tax-deferred IRA funds are taxed when received. However, Roth IRA funds are not taxed when withdrawn since contributions to a Roth IRA are from taxable, rather than pre-tax income. Withdrawals from IRA funds before age 59½ generally are assessed a 10% early withdrawal penalty. There are exceptions, including the payment of health insurance premiums if the worker is unemployed at least 12 weeks. Consequently, an unemployed older worker could choose to access retirement savings accounts to ease the worker through periods of unemployment.

Employer Provided Pensions

Under current law, a pension plan cannot pay benefits unless the recipient has either separated from the employer or reached the pension plan's normal retirement age. Some employers would like to pay partial pension distributions to workers at the plan's early retirement age and limit participation to workers in specific occupational categories. However, targeted participation may cause a pension plan to violate the provisions of the tax code that prohibit retirement plans from discriminating in favor of highly compensated employees.28

Disincentives to Continue to Work

Unemployment Compensation

In some states, the UC system excludes workers who were employed in part-time positions and who are searching for part-time work. If an older worker was already in a bridge job, working fewer hours, the worker may no longer have enough recorded earnings to be sufficient for the purpose of receiving UC benefits. This inflexibility also is seen in the treatment of searching exclusively for part-time work. It is only in 29 states and the District of Columbia that a worker need not be explicitly available for full-time work. Two states (Michigan and West Virginia) summarily exclude workers from benefits if they search for only part-time positions.

As workers age, self-employment increases. In 2003, 6.9% of all non-farm workers were self-employed. In comparison, 10.5% of those aged 55-64 and 15.3% of those 65 and older were self-employed.29 After a spell of unemployment, the probability that workers begin their own business or are otherwise self-employed increases.30 However, while these workers may be actively creating a job they generally do not qualify for UC benefits since they are not seeking non-self-employed work.

Although the ADEA prohibits most employers from having a policy of compulsory retirement on account of age, there are a few occupations that have bona fide reasons for age-based compulsory retirement.31 Most states allow workers to collect UC benefits if they no longer have a position due to compulsory retirement, treating the situation as a spell of unemployment. However, eight states consider compulsory retirement to be a voluntary quit and either exclude the workers from collecting UC, or limit or reduce benefits.

Understanding the impact of unemployment on older workers is complicated by the way that non-work income, in particular Social Security and pension income, may reduce unemployment compensation. The Multi-employer Pension Plan Amendments Act of 1980 (P.L. 96-364) requires that all retirement income from an employer when the earnings from that employer are used to calculate the UC benefit (a base period employer32) be offset against the UC benefit. This is to ensure that a worker who retires does not also collect UC benefits from the job from which they retired. States are permitted to reduce UC benefits on less than a dollar-for-dollar basis by taking into account the contributions made by the worker to finance the plan. This requirement applies only to those payments on a periodic (not lump-sum) basis. As listed in Table 3, columns 4 and 5, states can impose broader offsets than federal law requires for pension income. The District of Columbia and Virginia offset all pension income (as well as excluding lump-sum payments). The remaining states restrict the offset to pension income paid by base-year employers only. Some states offset this income only if the amount were increased by the base-period employment. Thirteen states do not disregard the employee contribution portion of pensions.

UC benefits may be offset by Social Security payments. Currently nine states and the District of Columbia at least partially reduce UC benefits if the worker is also receiving Social Security payments.

Employment Service

Few ES programs directly address older worker concerns.33 In fact, the Workforce Investment Act of 1998 (WIA, P.L. 105-220) eliminated set-asides that had been targeted for older workers in the previous program (Job Training and Partnership Act of 1982, P.L. 97-300) for adult training needs. Furthermore, WIA performance measures may give individual programs incentives to avoid enrolling older workers.34 Research indicates that most employment and training programs don't seem to help older workers find new, well-paying jobs. The loss of firm-specific knowledge, job tenure, and the likelihood that the worker is in a low-demand occupation or in a declining industry are likely to contribute to decreased reemployment wages. Generally, as workers age, cognitive abilities begin to slow and physical condition begins to decline. These factors combine to depress reemployment wages.35

Social Security

Although the earnings tests and deductions have been substantially reduced since 1999, they still constitute a substantial effective tax (up to 50%) on any meaningful work before the full retirement age. These age-based changes are listed in Table 2. As mentioned previously, workers who are at or above full retirement age (age 65 and 10 months if born in 1942) may keep all of their Social Security benefits, with no earnings tests. Workers younger than the full-retirement age for the entire year, have an earnings limit. For every $2 earned above $12,960 (in 2007) there is a $1 deduction in the benefit. The earnings test is altered if a worker will achieve full-retirement age during the year. In this case, for every $3 earned above $34,440 there is a $1 reduction in benefits until the month the worker achieves full retirement age.36

Income and Older Americans

As an older worker transitions from work-based income security to retirement-based income security, the worker's incentives to remain in the labor force change. This changing incentive structure is a result of the worker moving from participating in a program designed to aid an effective job search and quick return to the labor force towards a series of programs that are primarily intended to ensure retirement income adequacy. While it is difficult to measure the exact effect that the incentives and disincentives have on older Americans' participation in the labor force, an examination of the changing patterns of income sources as Americans age may capture the end result.

The effect of these potential earnings substitutes (Table 2) and the changes in tax rates shown (Table 3) may be seen in Table 4 where the changing distribution of income sources for older Americans is shown. Table 4 describes the sources of income for Americans 35 years and older in 2005 by age category. From ages 35 to 44 approximately 84% of Americans have earnings income and less than 3% report Social Security income. However, from ages 62 to 64 only 53.8% have earnings income while 43.0% report Social Security income. From ages 65 to 69, one-third of the group has earnings income while 81.8% have income from Social Security.37

The percentage of Americans who report UC income declines from 3.6% for ages 35-44 to 1.0% by ages 65-69. This may be attributed to both a decrease in eligibility based on the types of jobs that older workers hold, and the program rules that offset UC benefits by pension and Social Security income.

Table 4. Sources of Income and Percentage of Individuals Receiving Income in 2005, by Age

Source of Income

35-44

45-54

55-61

62-64

65-69

70+

Any income

93.0%

93.0%

93.2%

94.8%

96.8%

97.0%

No income

7.0%

7.0%

6.5%

5.2%

3.2%

3.0%

Earnings

84.1%

82.7%

72.3%

53.8%

32.0%

112.1%

Retirement benefits

 

Social Security

3.0%

4.6%

8.5%

43.0%

81.8%

90.3%

 

Public pensions

0.5%

1.6%

5.6%

9.3%

10.2%

12.4%

 

Private pensions or annuities

0.7%

1.6%

7.2%

15.9%

20.6%

24.9%

Income from assets

48.6%

53.4%

59.7%

57.1%

55.7%

55.3%

Veterans' benefits

0.4%

0.8%

2.1%

1.8%

1.9%

3.4%

Unemployment compensation

3.6%

3.4%

4.2%

2.2%

1.0%

0.2%

Workers' compensation

0.9%

1.2%

1.2%

1.0%

0.6%

0.3%

Public assistance

2.9%

3.1%

3.3%

3.8%

3.6%

3.4%

Personal contributions

5.7%

2.8%

1.0%

0.7%

0.8%

0.6%

Total (thousands)

43,121

42,797

21,612

7,683

10,231

25,274

Source: CRS calculations, CPS Mar. Demographic File, 2006.

Note: Individuals will have multiple sources of income. Therefore, percentages will not add up to 100%.

Policy Issues

One policy issue is how to address the inherent tensions among the UC system, alternative working arrangements, and eligibility for and receipt of various types of retirement income including Social Security benefits. Currently, each state's UC system treats these typical older-worker employment and income patterns in a different manner. Some states allow flexibility in the job search, allowing those workers who are only willing to work part-time to collect UC benefits. Some states treat retirement and pension income and Social Security benefits as earnings that offset the UC benefit. Other states offset only the portion of the increase in pension income that is attributable to the time period that also qualifies the unemployed worker for UC benefits. Ideally, there should exist a balance where the UC system will be able to recognize the need for older workers to bridge into retirement and supplement their earnings with a variety of non-work income sources while still maintaining the program's work emphasis.

Another policy issue is how federal programs might better balance providing income support for older Americans with providing appropriate work incentives for those who would prefer to remain engaged in some type of work. The abrupt changes in income support eligibility or program offsets on account of employment earnings can create extreme disincentives to remain engaged (or re-engage) in the labor market for certain groups of older unemployed workers.

For example, from age 62 until the year before full retirement age, early Social Security benefits are offset by rates of up to 50% as well as being subject to the employment taxes. This provides a disincentive for full-time workers to take early Social Security benefits as they would see a fairly small change in their total income while subjecting themselves to a permanent reduction in Social Security benefits. However, viewing these same offset rules from the vantage point of an early retiree, these offsets also erode the incentives for those early retirees to re-engage themselves in the labor market.

Footnotes

1.

See, for example, Sewin Chan and Ann Huff Stevens, "Job Loss and Employment Patterns of Older Workers," Journal of Labor Economics, vol. 19, no. 2 (2001), pp. 484-521. Also see Sara Six, "Update on the Older Worker: 2002," Data Digest, no. 88, AARP Public Policy Institute.

2.

Arlene Dohm, "Gauging the Labor Force Effects of Retiring Baby-boomers," Monthly Labor Review, July 2000.

3.

For example, P.L. 106-182, the Senior Citizens Freedom to Work Act, effective January 1, 2000, eliminated the retirement earnings test for those at or above full-retirement age (and under age 70; those over age 70 were already exempt). The stated intent of the act was to reduce the labor force participation and earnings disincentives that recent retirees faced. (The previous rules for workers ages from ages 65 up to 75 had an exemption of the first $15,500 earned and a one-third reduction in benefits thereafter.)

4.

P.L. 90-202, the Age Discrimination in Employment Act of 1967, amended in 1986 by P.L. 99-592 removed the upper age limitation of age 70 of the act; this amendment extended coverage to all individuals who are at least 40 years of age.

5.

For example, the UC program does not allow extra time for older workers' job searches nor does the UC program allow for easy fulfillment of UC program requirements by creating self-employment.

6.

Abbigail Chioto and Michael Owyang, "Low Unemployment: Old Dogs or New Tricks?" Regional Economist, October 2001, Federal Reserve Bank of St. Louis.

7.

Sara Six, "Update on the Older Worker: 2002," Data Digest, No. 88, AARP, June 2003.

8.

See CRS Report RL30629, Older Workers: Employment and Retirement Trends, by [author name scrubbed] (pdf).

9.

For example, individuals who have retired from careers in law-enforcement or the military—both of which typically provide pensions after 20 years of service—often work for many years at other jobs while concurrently receiving a pension from prior employment.

10.

For a detailed statistical description of these patterns see [author name scrubbed], "Older Workers: Employment and Retirement Trends," Monthly Labor Review, October 2000, pp. 19-30.

11.

For more discussion on this topic and how there are impediments to partial retirement and part-time work, see Legal and Institutional Impediments to Partial Retirement and Part-time Work by Older Workers, by Rudolph Penner, Pamela Perun, and Eugene Steuerle, Urban Institute, http://www.urban.org/UploadedPDF/410587_SloanFinal.pdf.

12.

Roger Gordon and Alan Blinder, "Market Wages, Reservation Wages, and Retirement Decisions," NBER Working Paper W0513, 1981. Advancing age may pull wages downward as productivity for many workers decreases. See also Henry Farber, "Job Loss in the United States: 1981-2001," NBER Working Paper 9707, May 2003.

13.

Diane Herz and Philip Rones, "Institutional Barriers to Employment of Older Workers" Monthly Labor Review, April 1989, pp. 14-21.

14.

Christopher O'Leary and Stephen Wandner.,"Unemployment Compensation and Older Workers,"in Peter Budetti, Richard Burkhauser, Janice Gregory, and H. Allan Hunt, eds., Insuring Health and Income Security for an Aging Workforce (Kalamazoo, Michigan: W.E. Upjohn Institute for Employment Research, 2001). (Hereafter cited as O'Leary and Wandner, Unemployment Compensation and Older Workers.)

15.

For more discussion on this topic, see CRS Report RL32111, Unemployment Compensation (UC)/ Unemployment Insurance (UI): Trends and Contributing Factors in UC Benefit Exhaustion, by [author name scrubbed].

16.

Sewin Chan and Ann Huff Stevens, "Job Loss and Employment Patterns of Older Workers," Journal of Labor Economics, April 2001. (Hereafter cited as Chan and Stevens, Job loss and Employment Patterns of Older Workers.)

17.

See Henry Farber, Job Loss in the United States, 1981-2001, NBER Working Paper 9707, May 2003, http://papers.nber.org/papers/w9707.pdf.

18.

See Evaluation of the Workforce Development Partnership Program in 1994, 1995 and 1996, New Jersey: John J. Heldrich Center for Workforce Development, Rutgers University, 2000.

19.

For a detail description of all taxes that older workers face, see "The Implicit Tax on Work at Older Ages," National Tax Journal, Bol LVIV, No. 2, June 2006, by Barbara Butrica, Richard Johnson, Karen Smith, and Eugene Streuele.

20.

 Federal law requires that no more than 5% of workers receiving regular UC may be part of the SEA program.

21.

The Wagner-Peyser Act of 1933 established a nationwide system of public employment offices, and it was amended in 1998 to be part of the One-Stop delivery system. State Employment Security Agencies are affiliated with the U.S. Employment Services (ES) and assist job seekers in finding jobs and employers in finding qualified workers.

22.

See CRS Report RL30055, Older Americans Act: 2000 Reauthorization Legislation, by [author name scrubbed] (pdf) for more information on the program. In FY2004 the program was allocated $439 million in funds.

23.

The retirement earnings test refers to the amount of earnings a worker may receive without offsetting Social Security benefits. After this amount is reached, earnings above this level offset Social Security benefits until all benefits have been eliminated through the offset.

24.

It is possible that mid-year full-retirement aged retirees have earned more than the yearly earnings limit. Under OASI rules these retirees will receive a full OASI benefit for any whole month that they are retired, regardless of their yearly earnings.

25.

See Leora Friedberg, "The Labor Supply Effects of the Social Security Earnings Test," NBER Working Paper 7200, June 1999. Jonathan Gruber and Peter Orszag in "Does the Social Security Earnings Test Affect Labor Supply and Benefit Receipt," NBER Working Paper 7923, September 2000, suggest that the effect is negligible for men but does matter for women.

26.

See Jae Song, "The Effects of the Removal of the Retirement Earnings Test in 2000," paper presented at the 4th International Research Conference on Social Security, May 2003.

27.

For tax-deferred 401(k) plans, the 10% early withdrawal penalty is waived if the participant: died; became permanently disabled; used the funds to purchase a lifetime annuity; retired on or after age 55; incurred medical bills large enough to qualify for itemized deduction treatment under the federal income tax; transferred the funds within 60 days to another tax-deferred retirement account; or, received the funds in the form of a repayable loan.

28.

See CRS Report RL30629, Older Workers: Employment and Retirement Trends, by [author name scrubbed] (pdf) for more details.

29.

See CRS Report RL32387, Self-Employment as a Contributor to Job Growth and as an Alternative Work Arrangement, by [author name scrubbed] for a detailed description of how self-employment is important in the labor force participation of older workers.

30.

Chan and Stevens, Job lobs and Employment Patterns of Older Workers.

31.

These occupations include some protective services workers.

32.

A base-period employer refers to any firm for which the UC recipient received a wage during the period that will determine UC benefits. Typically this period begins five quarters before the quarter of unemployment and ends one quarter before the quarter of unemployment.

33.

The Worker Profiling and Reemployment Services (WPRS) system, a statistical model developed by the USDOL, is used by many states to aid in early identification of UC benefit recipients who are likely to exhaust their benefits before finding a new job. While it was determined during the development of the WPRS system that age was a significant variable, USDOL determined that the use of that variable was prohibited by federal civil right legislation. See Christopher O'Leary and Stephen Wandner, UC and Older Workers, NASI Conference on Health and Income Services for an Aging Workforce, January 2000.

34.

U.S. General Accountability Office, GAO Report 03-350, a report to the Ranking Minority Member, Subcommittee on Employer-Employee Relations, Committee on Education and the Workforce, House of Representatives, Older Workers: Employment Assistance Focuses on Subsidized Jobs and Job Search, but Revised Performance Measures Could Improve Access to Other Services, January 2003.

35.

One New Jersey program for dislocated workers found that of the 57% of older workers who had found a new position three years after first becoming unemployed, they were earning 77% of the previous job. See Evaluation of the Workforce Development Partnership Program in 1994, 1995 and 1996, (New Jersey: John J. Heldrich Center for Workforce Development, Rutgers University), 2000. Similarly, a study of Washington state community college retraining found older displaced workers were less likely to participate in schooling although those who did had small positive rates of internal returns of between 2%(men) and 4% (women). See Louis Jacobson, Robert LaLonde, and Daniel Sullivan, Should We Teach Old Dogs New Tricks? The Impact of Community College Retraining on Older Displaced Workers, Federal Reserve Bank of Chicago, WP 2003-25, November 2003.

36.

It is possible that mid-year full-retirement aged retirees have earned more than the yearly earnings limit. Under OASI rules these retirees will receive a full OASI benefit for any whole month that they are retired, regardless of their yearly earnings.

37.

For more detailed analysis of sources of income for those 65 and older, see CRS Report RL33387, Income of Americans Age 65 and Older, 1969 to 2004, by [author name scrubbed] and Debra B. Whitman.