III. LERA REFERED PAPERS: LABOR MARKET ECONOMICS, HUMAN RESOURCES, COLLECTIVE BARGAINING

Declining Access to Retiree Health Insurance Keeps Older Workers in the Labor Force

Christian Weller
Center for American Progress

Jeffrey Wenger
University of Georgia-Athens

Introduction

    The gift of increased life expectancy has rising costs associated with it. Chief among them is health care. Those 55 and older are disproportionate users of health care and have more frequent and more severe health problems than younger people, and the costs associated with their health problems are higher than for younger people with similar health problems.

    Health insurance coverage has become increasingly important as health care costs have exploded. A crucial source of health insurance coverage is employer-sponsored health insurance (ESI) coverage. Many employers, however, have reduced ESI access for their employees, either by no longer offering this benefit or by reducing the value of this benefit to contain costs. Declining ESI access poses a problem because there are no realistic substitutes. Early retirees are often not eligible for Medicare, those 65 and older, who are eligible for Medicare, lack important coverage (e.g., for prescription drugs), and private, nongroup insurance is too expensive.

    Thus, workers approaching retirement age have three options. They can apply for disability insurance, they can save more for higher health insurance costs during their working years, or they can decide to work longer. The results of our research show that recent increases in the labor force participation rate of older workers are due to the declining ESI access.

    This increase in labor force participation rates as a result of declining ESI access raises serious concerns. Employers will likely face rising health care costs as the share of workers in poor health with ESI coverage has grown faster than that of employees in good health. It is also possible that the overall health of the labor force is declining, which potentially reduces productivity. Moreover, employers may reduce ESI access, especially for employees in poor health, even faster if they offer health insurance for a growing share of employees in poor health. This could result in rising demands on public health insurance.

Health, Health Insurance, and Retirement

       Health improvements for the near-elderly, those from 55-64 years old, and the elderly, those 65 years of age and older, appear to be slow. The share of the near-elderly who reported themselves to be in fair or poor health declined from 26.6 percent in 1982 to 18.5 percent in 1999 (National Center for Health Statistics 2002b). The health of those 55 and older appears to have improved or at least remained stable during the recent economic downturn, but at a high level. The share of near-elderly retirees in poor health remained above 20 percent in 2002, and the share of elderly retirees in poor health stabilized at about 35 percent (Weller and Wenger 2004); however, the share of those older than 65 years with functional limitations remained stable, at around 7 percent, from 1983 to 1996 (National Center for Health Statistics 2002a). Although there may have been some improvements in people's health in the 1980s, these improvements may have leveled off in the 1990s, and the objective health status of the elderly appears unchanged.

    Older individuals are generally not as healthy as younger ones. Compared to 35-44 year olds, people between 55 and 64 years of age were twice as likely to report themselves in fair health and four times as likely to report themselves in bad health in 1996. Also, they were four times as likely to have had a stroke or cancer, seven times as likely to have had a heart attack, and five times as likely to have had heart disease as the younger comparison group (Gruber and Madrian 1996).

    Deteriorating health also means more medical spending. The nearelderly were twice as likely in 1987 to be admitted to the hospital and 40 percent more likely to need prescription medication, with twice as many medications, than 35-44 year olds. Thus, medical spending for 55-64 year olds was almost twice as much as for 35-44 year olds (Gruber and Madrian 1996, 2002). Even holding health constant, older people have greater health care needs than younger ones. Nichols (2001) estimated that men between 55 and 64 years of age and in good health spent about 2.5 times as much on health care as males between 21 and 29, if the older men were in poor health they typically spent 4.2 times as much. For females, the ratios were 1.3 and 1.9, respectively.

    Health status is an important determinant of retirement decisions. Burtless (1987) reported that about 26 percent people indicated that they left a job due to health considerations. Several other studies found similar proportions of workers in poor health, who were thus more likely to retire (Quinn 1977; Sammartino 1987; Quinn, Burkhauser, and Myers 1990; Leonesio, Vaughan, and Wixon 2000).

    Not everybody has the same chance of deteriorating health in old age. The chance of being in poor health varies with occupation, race, sex, education, and income1. These groups are also more likely to exit the labor force earlier, ceteris paribus, than their counter parts.

    Health insurance access, in addition to health, matters for the decision to retire. Karoly and Rogowski (1994) found that retiree health insurance coverage raised the chance of retirement by 47 percent; Rust and Phelan (1997) concluded that access to coverage decreased the probability of full-time work by 25 percent at ages 62-63; Headen, Clark, and Shumaker Ghent (1997) found that coverage increased the probability of retirement by 30 percent; and Karoly and Rogowski (2000) calculated that coverage increased the retirement probability by 62 percent for full-time male workers between 51 and 61 years old.

    At the same time that the health status of older workers has remained fairly stable, the costs of providing health care and health insurance have risen dramatically, raising the importance of retiree health insurance coverage. The costs of health care rose dramatically in the 1980s and 1990s. While overall prices increased by 110 percent from 1980 to 2002, medical costs rose by 278 percent, prescription drug costs by 327 percent, and the costs of hospital services by 422 percent. The costs of health insurance also rose much faster than overall prices, in particular by 200 percent from 1980 to 2001, the last year for which data are available (Bureau of Economic Analysis 2003; Bureau of Labor Statistics 2003; Center for Medicare and Medicaid Services 2003).

    Retiree health insurance access depends largely on employers offering this benefit, but ESI access is declining and costs for retirees are rising. In 1999, 35 percent of firms with more than 500 employees and 76 percent of employers with more than 1,000 employees offered ESI coverage to their retirees (Fronstin 2000). In 1988, however, 66 percent of large firms„with more than 200 employees„offered health coverage to retirees, compared to 34 percent in 2002. Among small firms, with fewer than 200 employees, only 5 percent offered employer-sponsored health insurance coverage in 2002 (Kaiser Family Foundation and Health Research and Educational Trust 2002). Furthermore, many companies that still offer ESI reduce costs through more stringent eligibility requirements, increasing employee shares of the premiums paid, reducing prescription drug coverage, lowering early retirement benefits, or capping their contributions (Freudenheim 2002; Kaiser Family Foundation and Health Research and Educational Trust 2002; Mercer Human Relations Consulting and Foster Higgens 2002; Watson Wyatt 2002).

    Declining ESI access care does not necessarily mean that fewer retirees have health insurance. In 2002, 32.0 percent of early retirees had ESI coverage from their former employers, whereas 19.0 percent received it as dependents (Weller and Wenger 2004), which suggests that almost 50 percent of early retirees were not covered by it. Moreover, more early retirees„ 21.1 percent„had no health insurance coverage than were covered as dependents on an ESI plan in 2002. Among elderly retirees, ESI as the sole source of health insurance coverage is negligible, because this group is eligible for Medicare. ESI coverage, however, is a substantial supplement to public health insurance; 28.4 percent of elderly retirees had ESI coverage in addition to public health insurance But ESI was less important than nongroup health insurance coverage, such as Medigap, in addition to public health insurance, which covered 29.8 percent of elderly retirees in 2002 (Weller and Wenger 2004).

    The possible responses include an increased propensity to apply for disability claims to increase access to publicly provided health insurance. Another response could be greater preretirement savings to cover rising medical expenditures in retirement. And the third option would be to delay retirement by working longer in an effort to keep health insurance coverage.

    Applying for disability insurance is one possible response to falling retiree health insurance coverage, but it is an imperfect substitute for longer work or early retirement with retiree health insurance. Gruber and Kubik (1994) found that a 10 percent increase in disability claim denial rates would lead to a 2.7 percent decrease in nonparticipation among men between 45 and 64 years of age, which implies that most individuals applying for disability would drop out of the labor force rather than continue employment in their conditions. Mitchell and Phillips (2000) found that, in the event of early retirement benefit reductions, individuals would be twice as likely to delay retirement until normal retirement age than to seek disability retirement benefits, which implies that disability benefits are not a substitute for retirement benefits. Furthermore, the median male who applies for benefits waits seven years after the onset of disability before submitting an application; the median female waits 8 years before applying (Burkhauser, Butler, and Weathers 2001).

    Workers could also save more for their retiree health insurance while they are still working. Some employers have begun defined contribution health plans for retirees, whereby current workers contribute to a fund that will allow them to pay for some of their retirement health care costs (Schmidt 2002). These so-called voluntary employee beneficiary associations (VEBAs) or retiree medical accounts (RMAs) require employees to assume full responsibility for their own health coverage in exchange for some tax savings, although contributions are made tax after taxes (Lee 2002; Gunsauley 2002). VEBA funds are placed into a qualified trust and can only be used for their predesignated purpose, thereby offering a certain amount of security because the money will not be used for other uses by the company (Lee 2002; Gunsauley 2002).
      There have been highly visible incidences of VEBAs. Perhaps most notable was the steelworkers' utilization of VEBAs to fund health benefits for the retirees of some bankrupt steel companies and as a means to prefund retiree health benefits outside of bankruptcy (Greenwald 2002; Fleet 2002). Despite provisions to ensure adequate capitalization of VEBA funds, these efforts have been largely unsuccessful because of ultimately insufficient funds. The effects of the inadequate VEBA funding can be devastating. For example, at LTV, one-fourth of all retirees were not eligible for Medicare, the vast majority of whom could not afford continuing insurance under COBRA, and thus had to go without medical coverage when the company went under.
      Prefunding of retiree health insurance is a poor substitute for ESI simply because an insurance program is replaced with individual accounts. The primary concerns associated with prefunding retiree health insurance are similar to those associated with defined contribution retirement savings plans. All risks are borne by the individual. In the case of defined contribution health care plans, these risks include in addition the rising costs of health care, the possibility of an employer's inability to pay promised benefits, the lack of predictability of future health expenditures, and hence the risk of households saving too little for retirement.

     Overall, estimates seem to predict, at the most, 20 percent of variation in health care costs. Pope et al. (2000) illustrate the expected health care expenditure variation by use of the principal inpatient diagnostic cost group (PIPDCG). In this model, the demographic factors are added to "add-on" factors, which take into consideration mostly serious illness situations requiring hospitalization, to calculate a risk factor for an individual. In an example used by the authors, a 69-year-old male with no disability, not on Medicaid or another health insurance plan, and with no hospitalization, is expected to incur 54.1 percent of the average individual health care expenditures for this year. Van Vliet (1992) found that at least 80 percent of an individual's health care costs cannot be predicted. For persons 50 years of age or older, the maximum observed variance that could be explained was 13.4 percent, and for persons 30 and younger 11.6 percent of the observed variance was explained. Garber, MaCurdy, and McClellan (1997) argued that individuals who incur high medical costs in one year are likely to incur high costs in subsequent years, thus allowing medical expenditures to be predicted more accurately, increasing the accuracy of the prediction by 7 percent.

     The third option for employees to address the declining access to employer-sponsored retiree health insurance is to work longer. In particular, do people in poor health work longer or do they retire earlier? People in poor health are substantially less likely to work than those in good health. Although typically more than 80 percent of those in good health work, fewer than 70 percent of those in poor health do. Further, from 1996 to 2000, there appeared to be a trend toward early retirement among those in self-reported poor health. Whereas 68.0 percent of those in poor health were in the labor force in 1996, only 63.3 percent were in the labor force in 2000; however, the share of those in poor health who were still employed rose sharply in 2001, to 68.3 percent from 63.0 percent in 2000, and declined only slightly to 67.3 percent in 2002 (Table 1).2 In comparison, among those in good health the share of employees grew from 80.5 percent to 81.8 percent from 200 to 2002.

     In light of a fairly stable labor force breakdown between those in good health and poor health among the near-elderly, and that the share of the near-elderly in poor health has remained steady, it is important to keep in mind that the labor force participation rates of the near-elderly have 

increased consistently since the mid-1980s. In late 2002, the labor force participation rates of the near-elderly reached their highest point, at 62.9 percent (Weller and Singleton 2003). Because the rate of reentrants declined in that period, the increase in labor force participation rates suggests that the near-elderly delayed retirement instead (Weller 2003).

The literature supports the view that increased labor force participation rates may be determined by health insurance access. If ESI coverage is an important reason to remain employed and if ESI access is declining, we should see stable or possibly even increasing ESI among employees. We should also see ESI coverage among retirees increase less than for employees or possibly even decline. Tables 2 and 3 show that this is the case. The share of the near-elderly employees with any ESI coverage rose by 3.7 percentage points from 65.8 percent in 1996 to 69.5 percent in 2002 (Table 2). In comparison, ESI coverage for early retirees rose only by 0.3 percentage points over the same period (Table 3). Near-elderly employees became increasingly more likely„compared to near-elderly retirees„to be covered by ESI from 1996 to 2002.
      Although the figures support the view that the rise in the labor force participation rates of the near-elderly may be due to ESI access, there is little evidence that those in poor health are more likely than those in good health to stay in the labor force to keep ESI. Among the near-elderly employees and early retirees, those in poor health were more likely than those in good health to see an increase in ESI coverage (Tables 2 and 3).

ESI access may, however, be declining for early retirees in poor health. This was the only group for whom there was a noticeable substitution of ESI coverage as dependents by ESI coverage in one's own name from 1996 to 2002 (Table 3). This substitution effect held both for men and women, despite differential overall trends in ESI coverage (Table 4). It is possible that this substitution reflects a difference in out-of-pocket expenditure growth for early retirees. Considering that employers are reducing ESI access, it seems reasonable to assume that the substitution effect from 1996 to 2002 was merely a temporary phenomenon.

Another aspect that supports the view that access to health insurance, or lack thereof, may have been a cause for the rise in the labor force participation rates of the near-elderly is the fact that the share of early retirees without health insurance has risen substantially more than the share of near-elderly employees without it. From 1996 to 2002, the share of the nearelderly employees without health insurance rose by only 0.7 percentage points to 13.5 percent in 2002 (Table 2). In comparison, the share of early retirees without health insurance rose by 2.6 percentage points, to 21.1 percent (Table 3). In both cases, the decline in health insurance coverage is partially attributable to a decline in the coverage by nongroup health insurance.

      In fact, to some degree, ESI coverage became a substitute for nongroup health insurance for early retirees. As employers covered relatively more early retirees, as a share of early retirees, and as their costs are rising, it is possible that employers will further reduce access to retiree health insurance in the future.

     It is not only the near-elderly who seem to be staying in the labor force longer, but also the elderly. Since reaching its low point in July 1985, the labor force participation rate of the elderly rose from 10.5 percent to more than 13 percent by late 2002 (Weller and Singleton 2003). A labor force participation rate of 13.6 percent in June and August 2002 were the highest such rates for the elderly since June 1978. But the increase in the labor force participation rate of those older than 65 years has proceeded at a gradual pace, without the acceleration after 2001 that was observed for the near-elderly.
      For the elderly, the changes in labor force participation rates may be a result of the loss of supplemental health insurance. For those over 65 years in general, nongroup health insurance coverage has declined dramatically. The trends from 1996 to 2002 show a rising share of retirees relied solely on public health insurance as their share rose by 6.7 percentage points, from 33.7 percent in 1996 to 40.5 percent in 2002. At the same time, the share of elderly retirees with supplemental nongroup private insurance declined by 6.4 percentage points, from 36.2 percent to 29.8 percent (Table 5). This implies that the big changes in elderly retirees' health insurance occurred in the private market, where many older workers have dropped their private plans.

     Disaggregating health insurance coverage trends for those older than 64 years also suggests that less ESI access for elderly retirees contributed to the rise in their labor force participation rates. Separating the sources of health insurance by age for the elderly indicates that the elderly between the ages of 65 and 74 have seen their ESI coverage decline, whereas the elderly„those 75 years of age and older„have seen it increase3. The share of the younger elderly retirees with ESI in their own name declined by 2.8 percentage points, from 34.5 percent in 1996 to 31.7 percent in 2002. In comparison, the share of older elderly retirees with ESI coverage rose by 1.9 percentage point, from 22.6 percent in 1996 to 25.5 percent in 2002 (Table 6).

Conclusion

     Retiree health insurance coverage has become an increasingly important aspect of retirement income security as health care costs are rising, while people are living longer. Employers, however, are reducing access to employer-sponsored retiree health insurance by either abandoning this benefit or shifting costs to employees and retirees. One result of the decline in access has been a rise in the labor force participation rate of the near-elderly

and less generous early retiree health insurance benefits, as reflected in a declining dependent coverage by employer-sponsored retiree health insurance plans. Another reflection that this is a fairly recent phenomenon is the fact that the employer-sponsored health insurance coverage for the elderly has remained largely unchanged, in contrast to the declining coverage for the near-elderly. Our results suggest, however, that ESI coverage for elderly retirees may decline in the future, as those between 65 and 74 have already seen declining coverage.

Notes 

1. See Weller and Wenger (2004) for a detailed literature review on this issue.

2. All figures are calculated from the March supplement to the Current Population Survey, various years. For a detailed description of the data and the underlying analysis, the reader is referred to Weller, Wenger, and Gould (2004).

3. This could be a survivor isse (i.e., those with health insurance are more likely to live beyond 75, raising the percentage with health insurance as they age). 

References

Bureau of Economic Analysis. 2003. National Income and Product Accounts. Washington, D.C.: Bureau of Economic Analysis.

Bureau of Labor Statistics. 2003. Consumer Price Index, Washington, D.C.: Bureau of Labor Statistics.

Bureau of Labor Statistics. 1994. Report on the American Workforce. Washington D.C: Bureau of Labor Statistics.

Burkhauser, R. V., J. S. Butler, and Robert R. Weathers II. 2001, "How Policy Variables Influence the Timing of Social Security Disability Insurance Applications." Social Security Bulletin, Vol. 64, no. 1, pp. 52-83.

Burtless, G., 1987, "Occupational Effects on the Health and Work Capacity of Older Men." In G. Burtless, ed., Work, Health, and Income among the Elderly. Washington, D.C.: Brookings Institution.

Center for Medicare and Medicaid Services. 2003. National Health Accounts, Washington, D.C.: Center for Medicare and Medicaid Services.

Fleet, S. 2002. "PNC Bank Signs on as Health Plan Principal Trustee for LTV Steel Retirees and Former Employees through Steel Worker Benefits Trust." Business Wire, May 23.

Freudenheim, M., 2002, Companies Trim Health Benefits for Many Retirees as Costs Surge, The New York Times, page A01, May 10.

Fronstin, P. 2000. "The Erosion of Retiree Health Benefits and Retirement Behavior: Implications for the Disability Insurance Program." Social Security Bulletin, Vol. 63, no.4, pp. 38-46.

Garber, A.M., T. E. MaCurdy, and M. C. McClellan. 1997. Persistence of Medicare Expenditures among Elderly Beneficiaries. NBER Working Paper No. 6249. Cambridge, Mass.: National Bureau of Economic Research [http://www.nber.org/papers/w6249].

Greenwald, J. 2002. "Ex-LTV workers form VEBA; Unusual Move Raises Questions, Benefit Experts Say." Business Insurance, April 22.6.

Gruber, J., and J. Kubik. 1994. "Disability Insurance Rejection Rate and the Labor Supply of Older Workers." National Bureau of Economic Research Working Paper No. 4941, Cambridge, Mass: National Bureau of Economic Research.

Gruber, J., and B. Madrian. 2002. "Health Insurance, Labor Supply, and Job Mobility: A Critical Review of the Literature." National Bureau of Economic Research Working Paper No. 8817. Cambridge, Mass.: National Bureau of Economic Research.

Gruber, J., and B. Madrian. 1996, "Health Insurance and Early Retirement: Evidence from the Availability of Continuation Coverage." In D.A. Wise, ed., Advances in the Economics of Aging. Chicago: University of Chicago Press, pp. 115-43.

Gunsauley, C. 2002. "Benefit Liabilities Threaten Retiree Medical Care." Employee Benefit News, November 1.

Headen, A. E., Jr., R. Clark, and L. Shumaker Ghent. 1997. "Effects of Retiree Health Insurance and Pension Coverage on the Retirement Timing of Older Workers: Sensitivity of Estimates." Unpublished paper, North Carolina State University.

Kaiser Family Foundation and Health Research and Educational Trust. 2002. "Retiree Health Benefits." Employer Health Benefits, 2002 Annual Survey. Menlo Park, Calif.: Kaiser Family Foundation, sect. 11, pp. 141-153.

Karoly, L. A., and J. Rogowski. 1994. "The Effect of Access to Post-Retirement Health Insurance on the Decision to Retire Early." Industrial Labor Relations Review, Vol. 48, no. 1, pp. 103-23.

Lee, K. 2002. "Education the Best Option: Funding Mechanisms for Retiree Health Care Deemed Insufficient." Employee Benefit News, December 1.

Leonesio, M. V., D. R. Vaughan, and B. Wixon. 2000. "Early Retirees under Social Security: Health Status and Economic Resources." ORES Working Paper Series No. 86. Washington D.C.: Social Security Administration, Office of Policy, Office of Research, Evaluation and Statistics.

Mercer Human Relations Consulting and Foster Higgens. 2002. Employer Sponsored Health Plans: Key Findings.Washington, D.C.: Mercer Human Relations Consulting.

Mitchell, O., and J. Phillips. 2000. "Retirement Responses to Early Social Security Bene- fit Reductions." National Bureau of Economic Research Working Paper No. 7963. Cambridge, Mass.: National Bureau of Economic Research.

National Center for Health Statistics. 2002a. Functional Limitation by Age, Sex, and Race (National), 1983-1996. Hyattsville, Md.: National Center for Health Statistics [http://www.cdc.gov/nchs/about/otheract/aging/trenddata.htm#Functional Status].

National Center for Health Statistics. 2002b. Self-Reported Health by Age, Sex, and Race: United States, 1982-1999. Hyattsville, Md.: National Center for Health Statistics [http://www.cdc.gov/nchs/about/otheract/aging/trenddata.htm#Health Status].

Nichols, L. 2001. "Policy Options for Filling Gaps in the Health Insurance Coverage of Older Workers and Early Retirees." In P. Budetti, R. Burkhauser, J. Gregory, and H. A. Hunt, eds., Ensuring Health and Income Security for an Aging Workforce. Kalamazoo, Mich.: W. E. Upjohn Institute for Employment Research.

Pope, Gregory C., Randall P. Ellis, Arlene S. Ash, Chuan-Fen Liu, John Z. Ayanian, David W. Bates, Helen Burstin, Lisa I. Iezzoni, and Melvin J. Ingber. 2000. "Principal Inpatient Diagnostic Cost Group Model for Medicare Risk Adjustment." Health Care Financing Review, Vol. 21, no. 3, pp. 93-118.

Quinn, J. F. 1977. "Microeconomic Determinants of Early Retirement: A Cross-Sectional View of White Married Men." Journal of Human Resources, Vol. 12, no. 3, pp. 329-46.

Quinn, J. F., R. V. Burkhauser, and D. A. Myers. 1990. Passing the Torch: The Influence of Financial Incentives on Work and Retirement. Kalamazoo, Mich.: W. E. Upjohn Institute on Employment Research.

Rogowski, J., and L. A. Karoly. 2000. "Health Insurance and Retirement Behavior: Evidence from the Health and Retirement Survey." Journal of Health Economics, Vol. 19, no. 4, pp. 529-39.

Rust, J., and C. Phelan. 1997. "How Social Security and Medicare Affect Retirement Behavior in a World of Incomplete Markets." Econometrica, Vol. 65, no. 4, pp. 781-831.

Sammartino, F. 1987. "The Effect of Health on Retirement." Social Security Bulletin, Vol. 50, no. 2, pp. 31-47.

Schmidt, R. L. 2002. "Strategies for Dealing with Retiree Medical" Employee Benefit News, August 1.

Van Vliet, Ren³ C. J. A. 1992. "Predictability of Individual Health Care Expenditures." Journal of Risk and Insurance, Vol. 59, no. 3, pp. 443-60.

Watson Wyatt. 2002. Retiree Health Benefits: Time to Resuscitate?Washington, D.C.: Watson Wyatt.

Weller, C. 2003. "Older Working Staying in the Labor Force." Economic Snapshot, July 16. Washington, D.C.: Economic Policy Institute.

Weller, C., and L. Singleton. 2003. "Between a Rock and a Hard Place: How the Loss of Health Care Coverage Puts Older Workers in a Bind." Indicators (summer).

Weller, C., and J. Wenger. 2004. "The Interaction between Health, Health Insurance, and Retirement." In T. Ghilarducci and J. Turner, eds., Work Options for Older Americans. Notre Dame, Ind.: University of Notre Dame Press, forthcoming.

Weller, C., J. Wenger, and E. Gould. 2004. The Other Side of Retirement Income Security. Washington, D.C.: Economic Policy Institute, forthcoming.