Health Affairs

This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

H E A L T H  T R A C K I N G:  
T R E N D S
W E B  E X C L U S I V E   

23 July 2003  
  

Employer-Sponsored Health Insurance
And Prescription Drug Coverage For New Retirees:
Dramatic Declines In Five Years

A scenario that adds real urgency to the debate over a Medicare drug benefit.


by Bruce Stuart, Puneet K. Singhal, Cheryl Fahlman, Jalpa Doshi, and Becky Briesacher

ABSTRACT:

Employer-sponsored health insurance is often described as the most reliable private source of Medicare supplementation, particularly for prescription drug benefits. This study’s findings show that employer coverage is becoming an increasingly less dependable source of coverage for new retirees, and the problem is likely to get worse. We found that the proportion of Medicare beneficiaries ages 65–69 with employer coverage declined from 46 percent in 1996 to 39 percent in 2000. The proportion with drug coverage from an employer declined from 40 percent in 1996 to 35 percent in 2000. Losses among males, the group most affected, would have been even greater had it not been for a slight increase in benefits from spouses’ policies.

Employer-sponsored health insurance is often described as the most reliable private source of Medicare supplementation, particularly for prescription drug benefits.¹ However, employers are cutting back on their offer of health benefits to both current employees and retirees. According to annual Mercer/Foster Higgins surveys, the number of large employers (500 employees or more) offering coverage to Medicare-eligible retirees declined from 57 percent in 1987 to 23 percent in 2001.² Surveys conducted by the Henry J. Kaiser Family Foundation/Health Research and Educational Trust (KFF/HRET) report similar findings for smaller firms, although fewer small firms have ever offered coverage to either employees or retirees.³ Also, most firms that maintained retiree coverage during the 1990s have since reduced benefits or increased employees’ share of premiums, or both.⁴ Health insurance take-up rates among eligible employees and retirees have fallen as a result.⁵

These statistics seem to imply a marked decline in employer coverage among retired Medicare beneficiaries. However, published data from the Medicare Current Beneficiary Survey (MCBS) indicate that the percentage of beneficiaries with supplemental medical and drug coverage from employer plans has remained virtually unchanged since the mid-1990s.⁶ One reason why actual employer coverage rates have not fallen as sharply as the surveys otherwise suggest is that employers typically eliminate coverage for future rather than current retirees.⁷ But this cannot be the only explanation. Our analysis focuses on a factor that has not been addressed in previous research: namely, the impact of women’s increasing workforce participation rates on employer coverage for new retirees. In 1980 only 41 percent of women in the preretirement age band, 55–64, were in the workforce; by 2000 this number had risen to 52 percent.⁸ Increased workforce participation for women should produce higher rates of employer coverage upon retirement for both sexes. Women gain because they are more likely to meet longevity-of-work requirements for retiree benefits. Men with working or retired spouses gain because they have greater opportunities for dependent coverage.

Have increases in employer coverage eligibility among women offset reductions in offer rates to new retirees? To find out, we tracked employer-sponsored medical and drug coverage for new retirees each year from 1996 through 2000. To assess the impact of increased workforce participation among women, we examined trends in coverage by sex and source of benefit (own versus spouse policy). We also tracked alternative sources of Medicare supplementation among people in this age bracket to determine whether beneficiaries substituted other types of coverage for employer coverage during the study period.

Data And Methods

We used the MCBS Cost and Use surveys from 1996 to 2000. These surveys report data from in-home interviews of a nationally representative sample of approximately 12,000 Medicare beneficiaries each year. They contain detailed information on beneficiaries’ demographics, socioeconomic characteristics, medical care and prescription drug use, estimated costs for all health services whether covered by Medicare or not, and coverage under Medicare supplemental and public plans. The survey records up to five private health plans, including starting and ending dates of coverage where applicable.⁹ Indicator variables show which plans offer prescription drug coverage. Some respondents either fail to report coverage or indicate that they do not know its source but during the surveys produce information—a receipt or insurance statement for example—indicating third-party payment. We classified beneficiaries as having employer coverage from a former or current employer based on self-reported coverage or evidence in the spending reports. Our study sample was restricted to community-dwelling beneficiaries ages 65–69.¹⁰ Although the MCBS does not provide information on beneficiaries’ retirement status, we reasoned that changes in coverage offerings to new retirees should be most evident in this age band.

We estimated annual prevalence rates of employer coverage and employer drug coverage for five annual samples of beneficiaries ages 65–69 from 1996 to 2000. We stratified these samples by sex and source of policy (own and spouse), to identify sex-specific trends.¹¹ To determine possible alternative sources of Medicare supplementation among those with no employer coverage plans, we assigned respondents to Medicare+Choice plans, public plans (Medicaid, Veterans Affairs, and various other programs), and individually purchased Medigap plans based on self-reports or evidence in the payment records.¹² We created a “multiple plan” category for beneficiaries with evidence of more than one Medicare supplement other than employer coverage. A very small proportion of respondents (under 2 percent a year) report coverage from an unknown source and could not be classified with payment-level data. These respondents are counted as not having supplemental coverage.

We conducted bivariate statistical tests of differences in mean employer coverage rates between 1996 and 2000 for each of our measures. Because the MCBS employs a complex sampling design, these tests were performed using the weighted survey estimator in STATA, version 7. This technique adjusts standard errors for clustering and stratification involved in the survey design. Statistical significance was defined at p < .05.

Study Findings

The overall proportion of all aged community-dwelling Medicare beneficiaries with health care coverage from an employer hovered at 39–40 percent over the period 1996–2000 (Exhibit 1). Prescription drug coverage from employer plans held constant at a fraction above 34 percent. None of the differences are statistically significant.

Exhibit 2 shows the trend in employer coverage between 1996 and 2000 for younger (ages 65–69) and older (age 70 and older) Medicare beneficiaries. A very different pattern is evident in these findings. The proportion of the younger group with medical benefits from an employer fell by a statistically significant six percentage points over the five-year period. The proportion of this group with employer-sponsored drug coverage declined 4.7 percentage points (also statistically significant). On the other hand, the trend in employer coverage for older beneficiaries is essentially flat, with a small, nonsignificant, but steady increase in prescription drug benefits.

The dissimilar trends for older and younger beneficiaries are the result of different demographic forces. The apparent stability in employer coverage among people older than age seventy is actually the product of high mortality among the oldest old (who have the lowest employer coverage rates of any age group) and their replacement by beneficiaries who are more likely to have employer coverage. The opposite dynamic is at work among those ages 65–69: Beneficiaries exiting the age band have higher rates of employer coverage than those entering it. Obviously, if current trends persist, employer coverage rates among Medicare beneficiaries will eventually decline across the entire age span.

Exhibit 3 shows the change in employer coverage for younger beneficiaries by sex (data for 1996 and 2000 only are shown, but the trend in the intervening years was consistently downward). While medical coverage fell for both sexes, the rate of decline for men (nine percentage points) far surpassed that experienced by women (three percentage points). The decline was statistically significant only for men. As a result, the gender gap evident in 1996 had largely disappeared by the turn of the century. Trends in drug benefits followed the same basic pattern. However, the loss of drug benefits was proportionally greater for men than for women (again, the differences were significant only for men).

Exhibit 4 shows changes in the source of employer coverage among these younger Medicare beneficiaries. In both 1996 and 2000 retiree policies earned by beneficiaries themselves (own policies) predominated; however, spousal coverage has become a relatively more important source of coverage for both sexes over time. Indeed, had it not been for spousal benefits, the decline in employer coverage observed in Exhibits 1–3 would have been much more severe. In just five years, the percentage of younger elderly men with employer coverage under their own policies fell 26 percent (statistically significant). In 1996 fewer than one in six males with employer benefits obtained them from a spouse; by 2000 one in four did so. A similar pattern is evident among female retirees, but the change is not statistically significant.

These findings underscore a dramatic transformation in the structure of retiree coverage during the late 1990s. The question remains how new retirees coped in the face of declining opportunities from their former employers. Exhibit 5 shows a small (albeit nonsignificant) increase in the proportion of younger beneficiaries with no Medicare supplementation (rising from 8.2 percent in 1996 to 9.5 percent in 2000). In other words, most new retirees found alternative sources of Medicare supplementation. These alternatives included Medicare+Choice (M+C) plans, public plans (Medicaid, Veterans Affairs, and various other programs), individually purchased private insurance (Medigap plans), or coverage under multiple plans. The most significant shift between 1996 and 2000 was from employer coverage into M+C plans and multiple coverage. The slight increase in proportion of beneficiaries in public plans is not statistically significant. On the other hand, the decline in employer coverage was matched by an equally large and statistically significant decline in self-purchased policies.

Gender matters here as it does with employer coverage. Men and women had very different patterns in supplementary coverage in both 1996 and 2000 (Exhibits 6 and 7). Switching to M+C is evident for both sexes, but the swing was almost twice as great for women: a 6.4-percentage-point increase for women versus 3.5 percentage points for men (only the former change is statistically significant). Men were somewhat less likely than women were to depend on multiple supplements in 1996, but that pattern had reversed by 2000 (the increase in multiple-plan coverage is statistically significant for men but not women). Both sexes moved away from reliance on self-purchased Medigap plans over the study period. The decline was significantly greater among women than among men.

Discussion And Policy Implications

The dramatic decline in employer coverage among Medicare beneficiaries ages 65–69 highlighted in this study is consistent with declining offer rates reported in employer surveys. Women fared much better than men. This, too, was expected, given rising workforce participation rates among preretirement-age women in the 1980s and early 1990s. We did not anticipate the shift from own to spousal policies—a trend most notable among males. Yet it represents a rational response to declining opportunities for retiree benefits. Had this shift not occurred, the overall drop in employer coverage for this age group would have been even more pronounced.

These trends have important implications for new retirees, particularly with respect to prescription drug coverage. The fact that almost all employer policies have drug coverage means that loss of employer coverage places retirees at risk of losing what is for many the single most valuable benefit from Medicare supplementation. Only a small proportion of new retirees are eligible for publicly funded drug coverage. None of the standardized Medigap policies that cover prescription drug expenses (the H, I, and J plans) and very few M+C plans available today provide the same generosity of drug coverage that is available from the typical employer coverage policy. In 2003, for example, 99 percent of all M+C enrollees faced annual benefit caps on prescription drug spending.¹³ Such caps are still uncommon among employer-sponsored plans. This might explain the growing proportion of beneficiaries with multiple Medicare supplements. As the available options become less desirable, some people are forced to move from plan to plan to maintain benefit levels, while others must depend on overlapping coverage.

Taken together, these results spell trouble for future retirees. First, there is no evidence that the pullback in offer rates for retiree medical and drug benefits has bottomed out. Indeed, questions posed to employers about possible future changes in retiree health benefits indicate that further erosion can be expected.¹⁴ Second, the historic increase in workforce participation by women has nearly reached its peak. The U.S. Bureau of Labor Statistics estimates that just 3.4 percent more women in the preretirement age band (ages 55–64) will be in the workforce in 2010 than in 2000.¹⁵ This rate of increase is too small to counteract declining employer coverage offer rates. Third, the rise in proportion of employer coverage recipients with drug benefits observed during the late 1990s is unlikely to continue into the twenty-first century. Employers have singled out drug benefits for cuts in the face of rapidly rising drug costs, and it is unreasonable to assume that retiree drug coverage will survive unscathed.¹⁶ Finally, and perhaps most important of all, the trends we have reported in this paper occurred over a very short period of time—just five years. Even if these trends continue at their current pace, the future availability of employer coverage for retirees will be seriously eroded. If the pace accelerates in the face of changing demographic and market forces, then the future of employer-sponsored Medicare supplemental coverage looks bleak indeed.

Further complicating the future for new retirees is the fact that the rest of the market for Medicare supplemental policies is also in flux. M+C enrollment peaked in 1999, and plan pullouts since then have stranded increasing numbers of beneficiaries.¹⁷ The decline in Medigap coverage among younger beneficiaries observed in the late 1990s is unlikely to turn around in the face of rapidly rising premiums. There has been an expansion in the number of state-sponsored pharmaceutical assistance programs in recent years, but these are at risk given states’ severe budget problems. In short, we can expect that greater numbers of new retirees will face the prospect of having no viable source of outpatient prescription drug coverage.

This scenario adds real urgency to the debate over a Medicare drug benefit. While it is obvious that beneficiaries who lose employer coverage lose any prescription drug benefits associated with their retiree policies, a potentially greater threat is that retirees will lose all of their supplemental health benefits because of the rising cost of prescription coverage. A properly structured Medicare drug benefit would at least provide employers with an incentive to maintain critical medical benefits that are unavailable in traditional Medicare.

The authors are grateful to Barbara Cooper, John Poisal, and Mary Laschober for helpful comments on previous drafts. This work was supported by the Commonwealth Fund. The views expressed here are those of the authors and do not necessarily represent those of the Commonwealth Fund or its directors, officers, or staff.

NOTES

1. M. Moon, “Medicare,” New England Journal of Medicine 344, no. 12 (2001): 928–931; and M.A. Laschober et al., “Trends in Medicare Supplemental Insurance and Prescription Drug Coverage, 1996–1999,”
www.healthaffairs.org/WebExclusives/Laschober_Web_Excl_022702.htm (7 July 2003).
2. William M. Mercer Inc., “Fifteenth Annual Mercer/Foster Higgins National Survey of Employer-Sponsored Health Plans,” 2001, www.imercer.com/us/imercercommentary/Healthsurvey/BB-final.pdf (2 July 2003); and D.L. DeWitt, “Emerging Facets of Retiree Benefits,” Business and Health 5, no. 10 (1988): 8–12.
3. Henry J. Kaiser Family Foundation and Health Research and Educational Trust, Employer Health Benefits: 1999 Annual Survey, October 1999, www.kff.org/content/1999/1538/KFF.pdf (2 July 2003); KFF/HRET, Employer Health Benefits: 2000 Annual Survey, September 2000, www.kff.org/content/2000/20000907a/EHBreport.pdf (2 July 2003); KFF/HRET, Employer Health Benefits: 2001 Annual Survey, October 2001, www.kff.org/content/2001/3138/EHB2001_fullrpt.pdf (2 July 2003); KFF/ HRET, Employer Health Benefits: 2002 Annual Survey, September 2002, www.kff.org/content/2002/3251/3251.pdf (2 July 2003); and KFF/HRET and Commonwealth Fund, “Erosion of Private Health Insurance Coverage for Retirees: Findings from the 2000 and 2001 Retiree Health and Prescription Drug Coverage Survey,” April 2002, www.cmwf.org/programs/medfutur/gabel_retiree_cb_506.pdf (2 July 2003).
4. KFF/HRET, Employer Health Benefits: 2002 Annual Survey; and KFF/HRET and Commonwealth Fund, “Erosion of Private Health Insurance.”
5. L. Duchon et al., Listening to Workers: Challenges for Employer-Sponsored Coverage in the Twenty-first Century, Pub. no. 361 (New York: Commonwealth Fund, January 2000).
6. Laschober et al., “Trends in Medicare Supplemental Insurance”; J.A. Poisal and L. Murray, “Growing Differences between Medicare Beneficiaries with and without Drug Coverage,” Health Affairs (Mar/Apr 2001): 74–85; and M. Davis et al., “Prescription Drug Coverage, Utilization, and Spending among Medicare Beneficiaries,” Health Affairs (Jan/Feb 1999): 231–243.
7. U.S. General Accounting Office, Retiree Health Benefits: Employer-Sponsored Benefits May Be Vulnerable to Further Erosion, Pub. no. GAO-01-374 (Washington: GAO, May 2001).
8. M. Toossi, “A Century of Change: The U.S Labor Force, 1950–2050,” Monthly Labor Review (May 2002): 15–28.
9. We use the MCBS Cost and Use files to generate statistics on employer coverage at any point during each calendar year. While most beneficiaries maintain these benefits for the entire year, some have gaps in coverage. For this reason, our estimates of coverage will differ from those generated from the MCBS Access to Care files, which provide information on supplemental coverage in hand at the time of the fall survey.
10. The MCBS survey of institutionalized beneficiaries does not collect information on type of supplemental insurance or whether prescription drug benefits are provided. Thus, it was necessary to exclude them from the analysis.
11. Approximately 0.5 percent of respondents each year reported having both their own and a spousal policy. To avoid double counting, these people were included in the own-policy group. An additional 0.5 percent reported having employer coverage but the source of the policy was missing. We assigned these people to own or spousal policies in the same proportions as beneficiaries with known sources of coverage.
12. Approximately 15 percent of beneficiaries classified as having employer coverage each year also had evidence of Medicare supplementation from another source.
13. Henry J. Kaiser Family Foundation, “Fact Sheet: Medicare+Choice,” April 2003, www.kff.org/content/2003/2052-06/2052-06.pdf (2 July 2003).
14. KFF/HRET, Employer Health Benefits: 2001 Annual Survey; and KFF/HRET, Employer Health Benefits: 2002 Annual Survey.
15. Toossi, “A Century of Change.”
16. KFF/HRET, Employer Health Benefits: 2001 Annual Survey; and KFF/HRET, Employer Health Benefits: 2002 Annual Survey.
17. C. Zarabozo, “Milestone in Medicare Managed Care,” Health Care Financing Review 22, no. 1 (2000): 61–67.


Bruce Stuart (bstuart{at}rx.umaryland.edu) is a professor and executive director of the Peter Lamy Center on Drug Therapy and Aging at the University of Maryland School of Pharmacy in Baltimore. Puneet Singhal, Cheryl Fahlman, and Jalpa Doshi are doctoral candidates in the Pharmaceutical Health Services Research program at the University of Maryland. Becky Briesacher is an assistant professor at the University of Maryland and director of research at the Peter Lamy Center.

©2003 Project HOPE–The People-to-People Health Foundation, Inc.