Employer-sponsored health insurance plans are the single largest source

Transcription

1 DataWatch Employer-Based Health Insurance In A Changing Work Force by Deborah Chollet Abstract: The loss of manufacturing jobs and the expansion of service jobs and part-time employment have contributed to a decline in the rate of employer-sponsored health insurance among workers. Not only does manufacturing provide more of its own workers with coverage compared with other industry groups, but it also is a significant net "exporter" of coverage to dependent workers in other industries. In 1991 the net export of coverage represented a 20 percent tax on manufacturing employers per covered worker, while professional services the fastest-growing industry group collected a subsidy from other industry groups equal to more than 12 percent per covered worker. Similarly, larger firms those that employed 100 workers or more paid a self-imposed tax of as much as 13 percent per covered worker to support dependent workers employed in smaller firms. Employer-sponsored health insurance plans are the single largest source of private insurance coverage among nonelderly Americans. In 1991, 140 million Americans under age sixty-five including nearly eightynine million workers had coverage from an employer-sponsored health plan. Approximately three-quarters of employer-insured workers are covered as a benefit of their own employment; all others some twenty-one million workers are covered as the dependent of an employer-insured worker, according to the March 1992 Current Population Survey. The number of Americans covered by an employer health plan has declined over the past decade. Historically, during periods of economic recession the number and proportion of workers without employer coverage has risen. However, at least since 1985 economic expansion has failed to produce growth of employer-sponsored health insurance among workers and their dependents. Between 1985 and 1991 the number of jobs that provided health insurance declined by nearly 2 percent; 1.2 million fewer workers were covered by their own employer in 1991 than had been covered six years earlier. Since 1988 the loss of employer-insured jobs has accelerated. The erosion of coverage by employer-sponsored plans coincides with major changes in the structure of the U.S. work force. The loss of manufacturing jobs characterized by relatively high rates of employer-sponsored Deborah Chollet is director of the Center for Risk Management and Insurance Research at Georgia State University in Atlanta.

2 316 HEALTH AFFAIRS Spring (I) 1994 coverage and the expansion of service jobs and part-time employment have contributed to the decline in coverage. This DataWatch describes the economic burden of employer-sponsored health insurance coverage assumed by employers and workers in different industries and different-size firms. Estimates of uninsured workers and employer-insured workers are presented, as are estimates of the net change in employer-insured jobs within major industry groups relative to the change in industry employment. These estimates indicate that the growth of employer-insured jobs has been consistently slower than job growth in industries that have experienced net job growth. In industries with declining employment, employer-insured jobs have declined even faster. Uninsured And Employer-Insured Workers It is well known that the percentage of workers who are uninsured varies among industry groups and firm sizes. Uninsured workers are counted as the difference between all workers in an industry or firm-size group and workers who report having coverage from any of a number of sources: as a benefit of the worker's own job; as the dependent of another worker; from an individual policy; or as a beneficiary of a public program, usually Medicaid. In 1991 just over half of all workers (55 percent) were covered directly by their own employer; all other employer-insured workers were covered only by another worker's plan. Nationally, for every ten workers insured by their own employer, another three were covered only as dependents. Variation by industry group. In 1991 the proportion of workers who were uninsured from any source throughout the year (or for a significant portion of the year) varied markedly among industry groups. Nearly 42 percent of agricultural workers reported having no insurance coverage (Exhibit 1). Among workers principally employed in construction and personal/entertainment services, nearly 30 percent were uninsured all year. These very high rates of uninsured workers contrast with very low rates of uninsured workers in other major industry groups, especially in government (7 percent) and in finance, insurance, and real estate (8 percent). In general, industries with low rates of uninsured workers have relatively high rates of employer coverage, either sponsored by the workers' own employers or received as dependents of workers at other firms. In no industry is a low rate of employer coverage fully offset by coverage from other sources, such as individually purchased insurance or public programs. In firms that do not contribute to health coverage, coverage received by workers as dependents of workers in other firms represents a real subsidy. This subsidy is collected in the form of higher wages by the dependent workers; by the firm as higher profit (perhaps supporting an otherwise

3 DATAWATCH 317 Exhibit 1 Uninsured And Employer-Insured Workers, By Industry, 1991 Industry group Percent uninsured workers Employer-insured workers Total per all workers Covered by own employer per all workers Ratio of workers covered by other employer per 100 workers covered by own employer High uninsured Agriculture Construction Personal/entertainment services 41.6% % % Moderate uninsured Self-employed Retail trade Business/repair services Low uninsured Wholesale trade Mining Transportation, communications, utilities Manufacturing Professional services Finance, insurance, real estate Government Source: Tabulations from the March 1992 Current Population Survey. Note: Tabulations include only civilian workers ages eighteen to sixty-four. unprofitable enterprise); or by consumers as lower product prices. The cost of coverage provided to dependent workers employed by another firm represents a transfer among firms and industry groups in effect, a tax on firms that export coverage and a subsidy to firms that import coverage for workers that they do not otherwise insure- In general, industry groups with relatively high proportions of uninsured workers are also relatively large importers of health insurance coverage from other employers' plans and from Medicaid (Exhibit 2). Variation by size of firm. Differences in coverage also occur systematically among firms of different sizes, regardless of industry group. Among wage and salary workers who identified firms with fewer than ten employees as their principal job in 1991, 29 percent reported having no insurance from any source throughout the year (Exhibit 3). This rate was systematically lower as firm size increased; in firms of 1,000 employees or more, only 9 percent of workers reported having no coverage from any source during the year. Workers in large firms are substantially more likely than workers in small

4 318 HEALTH AFFAIRS I Spring (I) 1994 Exhibit 2 Imported Coverage From Other Employer Or Medicaid, By Industry, 1991 Source: Tabulations from the March 1992 Current Population Survey. Exhibit 3 Percentage Of Workers Without Insurance And Percentage With Employer-Based Coverage, By Firm Size, 1991 Size of firm Self-employed Percent uninsured workers 25.8% Employer-insured workers Covered by own Total per employer per all worker s all workers 42.8% 15.6% Ratio of workers covered by otber employer per 100 workers covered by own employer workers workers workers workers workers 1,000 workers or more Source: Tabulations from the March 1992 Current Population Survey. Note: Tabulations include only civilian workers ages eighteen to sixty-four.

5 DATAWATCH 319 firms to have coverage sponsored by their own employer (regardless of full-time job status) and less likely to have employer-based coverage only as a dependent of another worker. Among workers in firms of a thousand or more employees, just seventeen workers are covered only as a dependent for every hundred workers covered by their own employer (Exhibit 3). In contrast, for every hundred workers covered by their own employer in firms with fewer than ten employees, another ninety-five are covered only as the dependent of another worker. Medicaid. Medicaid is a relatively small but growing source of coverage among low-income workers in the United States, mainly as a result of federal expansions of Medicaid eligibility effective since 1990 and expansions in many states in excess of the federally required minimum. All states must cover pregnant women and infants (to age one) with family incomes below 133 percent of the federal poverty level States may extend eligibility to pregnant women and infants up to 185 percent of poverty with full federal matching, and many have done so. Some states have extended Medicaid to these groups with income as high as 200 percent of poverty. These provisions have resulted in a small but growing number of workers and families who qualify for Medicaid. While only about 2 percent of American workers reported receiving coverage from Medicaid during 1991, in some industry groups agriculture, retail trade, business and repair services, and personal and entertainment services reliance on Medicaid as an exclusive source of health coverage ranged between 4 percent and 6 percent of total industry employment. Similarly, reliance on Medicaid by workers employed in small firms (4 percent) is greater than that among workers in large firms (2 percent). However, nearly 29 percent of Medicaid-insured workers identified firms with 1,000 employees or more as their principal employer; only 20 percent identified firms with fewer than ten employees (Exhibit 4). As defined here, these workers reported no coverage from an employer (either from their own employer or as a dependent) that year. Growth And Loss Of Employer-Insured Jobs Since 1985 U.S. net employment has grown 7.4 percent. This includes rapid growth in some industries (professional services, personal/entertainment services, and self-employment) and declining employment in others (manufacturing and mining). As employment has changed among industry groups, so has the proportion of workers covered by their own employer (Exhibit 5). Several aspects of the relative change in employer-covered workers versus the change in industry employment since 1985 are significant. First, industries that have grown most rapidly since 1985 also have gained em-

6 320 HEALTH AFFAIRS Spring (I) 1994 Exhibit 4 Percentage Of Medicaid-Insured Workers, By Firm Size, 1991 Source: Tabulations from the March 1992 Current Population Survey. ployer-insured jobs. However, in each of the fast-growing industries, employment growth has substantially exceeded the growth of employerinsured jobs. For example, for every hundred jobs gained in professional services (the highest-growth industry), only forty included coverage (Exhibit 5). As a result, while the number of own-employer-insured workers in high-growth industries has risen, the percentage of workers in the industry with coverage from their own employer has declined. Second, while own-employer coverage has trailed employment growth in all industries, the erosion of coverage has been particularly pronounced in slow-growth industries. In very-low-growth industries construction and agriculture the decline in employer-insured jobs has been dramatic, with losses of 270 to 344 employer-insured jobs per 100 new jobs gained overall. Finally, in both manufacturing and mining industry groups that historically have had among the highest rates of employer-insured workers the rate of loss of employer-insured jobs has exceeded the rate of decline in total employment (Exhibit 6). For every 100 manufacturing jobs that have been lost since 1985, 224 employer-insured manufacturing jobs have been lost (see Exhibit 5). These changes in the number and rate of employer-insured jobs across sectors suggest that further employment growth is unlikely to stem the dwindling rate of employer-based coverage. Furthermore, the rapid decline of insured jobs in major sectors of the economy may be problematic for other industries. Both mining and manufacturing are relatively small importers of coverage from other industries. To the degree that they are also exporters of health coverage to dependent workers in other industries, the

7 DATAWATCH 321 Exhibit 5 Changes In Total Employment And Employer-Sponsored Coverage, By Industry, Industry/ employment growth Total Percent change in employment 7.4% Percent change in own-employer coverage -1.7% Change in ownemployer coverage per 100-worker growth in employment -14.3% High growth Professional services Personal/entertainment services Self-employed Moderate growth Finance, insurance, real estate Transportation, communications, utilities Government Wholesale trade Business/repair services Retail trade Low growth or declining Construction Agriculture Manufacturing Mining Source: Tabulations from the March 1986 and March 1992 Current Population Surveys. Note: Dashes indicate values of less than Tabulations include only civilian workers ages eighteen to sixty-four. faster loss of insured jobs in mining and manufacturing withdraws a real subsidy from other industries and contributes to an ongoing decline in employer-based coverage relative to employment in all sectors. Effects Of Changes In Employer-Insured Jobs Across Sectors Changes in the structure of employment and primary employer coverage may exert a secondary effect across sectors by changing available sources of insurance coverage for workers whose only coverage is received as a dependent of another worker. In particular, declining own-employer coverage in industries that export coverage suggests loss of dependents' coverage and, therefore, growing noncoverage in other industry groups as well The distribution of health insurance exported to dependent workers in other industry groups (Exhibit 7) is significant in several ways. First, industry groups that have gained employment the fastest since 1985 are also among the strongest net importers of dependent coverage from other indus-

8 322 HEALTH AFFAIRS Spring (I) 1994 Exhibit 6 Percentage Change In Total Jobs And Health-Insured Jobs, Source: Tabulations from the March 1986 and March 1992 Current Population Surveys. try groups, both absolutely and relative to own-industry employment. In 1991 the professional services industry "imported" coverage for 1.2 million workers net of its export of coverage to workers in other industry groups. Second, the two industry groups that absolutely lost employer-insured jobs between 1985 and 1991 manufacturing and mining are major exporters of insurance coverage to workers in other industries. In 1991 manufacturing exported coverage to 3.2 million dependent workers in other industries net of imported coverage; these dependent workers were not otherwise covered by an employer plan of their own. On average, for every hundred manufacturing jobs that are lost, a net fifteen workers in other industry groups also lose employer-sponsored coverage. Thus, the net loss of more than one million manufacturing jobs since 1985 has produced an estimated net loss of dependent workers' coverage for at least 157,000 workers employed in other industries. The slow growth of employer-insured jobs in import industries, moreover, has failed to offset the interindustry loss of dependent workers' coverage. Finally, within industry groups, large firms are the greatest net exporters

9 DATAWATCH 323 Exhibit 7 Net Export Of Health Insurance To Dependent Workers In Other Industries, 1991 Industry/ employment growth High growth Professional services Personal/entertainment services Self-employed Percent change in employment 29.6% Percent change in own-employer coverage 19.7% Net number of other-industry workers covered as dependents (millions) Moderate growth Finance, insurance, real estate Transportation, communications, utilities Government Wholesale trade Business/repair services Retail trade Low growth or declining Construction Agriculture Manufacturing Mining Source: Tabulations from the March 1986 and March 1992 Current Population Surveys. Note: Dashes indicate values of less than Tabulations include only civilian workers ages eighteen to sixty-four. of dependent workers' coverage both absolutely and as a proportion of large-firm workers (Exhibit 8). In 1991 firms with 1,000 employees or more exported coverage to 4.3 million dependent workers in firms of other sizes, net of imported coverage. For every hundred large-firm workers in 1991, nearly ten workers in smaller firms were covered as dependents. Similar to the net loss of exported coverage associated with the loss of employer-insured jobs in some industries, the downsizing of large U.S. firms apparently has also contributed to the erosion of coverage among workers and dependents since Further growth in small firms is unlikely to produce offsetting increases in coverage for workers and their families, absent government action requiring significant employer contributions to coverage. In 1991 small firms and particularly those with fewer than twenty-five workers were strong net importers of insurance coverage. Export Of Health Insurance Benefits As Taxation Across Sectors The export of health insurance to dependent workers represents in effect a self-imposed tax on firms that offer benefits, collected as a subsidy to

10 324 HEALTH AFFAIRS Spring (I) 1994 Exhibit 8 Net Export Of Health Insurance To Dependent Workers, By Firm Size, 1991 Source: Tabulations from the March 1992 Current Population Survey. workers and their employers in firms that do not provide benefits (Exhibit 9). Net tax subsidy rates are calculated in each sector as the number of workers in another sector that employers cover as dependents, net of imported coverage, per the number of workers covered within the sector as the primary beneficiary. Obviously, sectors with both a high rate of covered workers and a high self-imposed net tax bear the largest total burden to support dependent workers employed in other sectors. Among the industry groups, the highest rates of effective taxation on health insurance benefits offered to workers occur in mining (35 percent) and manufacturing (20 percent), as well as in transportation, communications, and public utilities (23 percent). In addition, these industry groups have relatively high rates of covered workers, resulting in a disproportionate cost burden for them. Compared with all other industry groups, mining and manufacturing also have shown absolute loss of employment as well as the fastest loss of insured jobs. Conversely, the greatest rates of subsidy are being collected by firms and workers in personal and entertainment services (72 percent) and by self-employed workers (141 percent) both fast-growing sectors as well as in retail trade (35 percent). The magnitude of self-imposed taxation to support dependent workers also is positively related to firm size. In 1991 all firms with 100 employees or

11 DATAWATCH 325 Exhibit 9 Estimated Rate Of Taxation Or Subsidy, By Industry And Firm Size, 1991 Industry/ employment growth High growth Professional services Personal/entertainment services Self-employed Moderate growth Finance, insurance, real estate Transportation, communications, utilities Government Wholesale trade Business/repair services Retail trade Low growth or declining Construction Agriculture Manufacturing Mining Firm size Self-employed workers 1-9 workers workers workers workers workers 1,000 workers or more Percent of workers with own-employer coverage 55.9% Estimated tax (+) or subsidy (-) rate a -12.5% Source: Tabulations from the March 1986 and March 1992 Current Population Surveys. Note: Tabulations include only civilian workers ages eighteen to sixty-four. a Calculated as the net number of other-industry workers covered as dependents per own-covered workers. Positive values indicate a tax; negative values indicate a subsidy. more paid a significant, self-imposed net tax per covered worker to support coverage of dependent workers in smaller firms (Exhibit 9). This tax varied from 4.5 percent among firms that employed 100 to 499 workers to nearly 13 percent among firms that employed 1,000 workers or more. Conversely, firms with fewer than 100 employees received an average net subsidy of 3.4 percent to 50 percent per covered worker in 1991 in the form of dependents' coverage provided to their workers by larger firms. These taxes and subsidies across sectors are likely to affect the number of jobs that offer coverage as well as job growth within sectors. That is, the additional cost imposed on insured mining and manufacturing jobs, as well as on insured jobs in large firms (sectors that have taxed themselves heavily to support dependent workers in other industries) discourages the offering

12 326 HEALTH AFFAIRS Spring (I) 1994 of benefits in those industries and may contribute to total job loss as well Conversely, high rates of subsidy encourage employment in some sectors especially professional services and self-employment, as well as retail trade. In addition, the availability of these subsidies to some employers probably deters other employers from providing health benefits to their workers. That is, in industries that import substantial amounts of coverage, firms that do not receive subsidies are best able to compete with firms that are subsidized only if they, too, do not offer coverage to their workers. Competition does not allow employers in such an industry to consider whether their workers are covered at all. Concluding Remarks The erosion of employer-sponsored health insurance among American workers and their families is significant. The loss of employer-insured jobs in one sector affects coverage in others as dependents' coverage also is lost. Major industry groups and small and large firms are strongly interdependent, in terms of the dependents' coverage imported from and exported to other sectors. It is notable that the industry groups that have actually lost jobs since 1985, and at an even faster pace have lost employer-insured jobs, are among the strongest net exporters of dependent coverage to other industry groups. Conversely, the rate of dependents' coverage imported by high-growth industries suggests that their growth has been encouraged by the real subsidies provided to their workers in the form of such coverage. These patterns offer little to support the belief that erosion of employerbased coverage will reverse itself. The extraordinary growth of health care costs persistently and by wide margins exceeding the growth of prices and wages in all other sectors implies an increasing burden on businesses that provide health insurance benefits to their workers. As the subsidies provided to other-industry workers in the form of dependents' coverage also become more burdensome, the propensity to offer family coverage or any coverage at all to one's own workers declines. Aggressive government action to curb the growth of health care costs may stem the further erosion of employer-based coverage among workers and their families. However, it is unclear whether that action alone would be sufficient to reverse the loss of insured jobs that is being driven by competition. Nevertheless, failure to act seems virtually certain to encourage continued loss of employer-based coverage among American workers, as well as the loss of total employment in sectors that tax themselves to support coverage of dependent workers employed eslewhere. The opinions expressed in this paper are those of the author and do not necessarily reflect those of the Center for Risk Management and Insurance Research, its associates, or its sponsors.