THE GROWING SHARE OF UNINSURED WORKERS EMPLOYED BY LARGE FIRMS Sherry Glied Columbia University Jeanne M. Lambrew George Washington University Sarah Little Columbia University October 2003 The authors would like to thank Doug Gould at Columbia University for research assistance and data analysis on this report. Support for this research was provided by The Commonwealth Fund. The views presented here are those of the authors and should not be attributed to The Commonwealth Fund or its directors, officers, or staff, or to members of the Task Force on the Future of Health Insurance. Copies of this report are available from The Commonwealth Fund by calling its toll-free publications line at 1-888-777-2744 and ordering publication number 672. The report can also be found on the Fund’s website at www.cmwf.org. CONTENTS List of Figures and Tables................................................................................................ iv About the Authors........................................................................................................... v Executive Summary....................................................................................................... vii Introduction .................................................................................................................... 1 Uninsured Workers in Large Firms: Scope of the Problem............................................... 2 Access to Health Coverage at Large Firms........................................................................ 7 Labor Market Changes and Large Firms ........................................................................... 9 Conclusions and Recommendations .............................................................................. 13 Tables............................................................................................................................ 17 Appendix. Methodology................................................................................................ 21 References..................................................................................................................... 24 iii LIST OF FIGURES AND TABLES Figure ES-1 Share of Uninsured Workers by Firm Size, 1987–2001 .......................... viii Figure ES-2 Insurance Coverage of Workers in Large Firms by Household Income, Percent Uninsured During 1998.......................... ix Figure 1 Share of Uninsured Workers by Firm Size, 1987–2001 ............................. 3 Figure 2 Trend in Uninsured Rate by Firm Size, 1987–2001.................................. 4 Figure 3 Insurance Coverage of Workers in Large Firms by Household Income, Percent Uninsured During 1998........................... 5 Figure 4 Trend in Workers with Insurance Through Own Job by Firm Size, 1987–2001 .......................................................................... 6 Figure 5 Health Insurance Offer Rate by Firm and Establishment Size .................... 9 Figure 6 Share of Workers with Incomes Below 200% FPL by Firm Size, 1987–2001 ........................................................................ 10 Figure 7 Percent of Workers in Unions by Firm Size, 1987–2001......................... 11 Figure 8 Percent of Workers in Manufacturing Jobs by Firm Size, 1987–2001 ...... 12 Figure 9 Factors Affecting Change in Coverage in Large Firms, 1987–2001 .......... 13 Table 1 Distribution of the Nonelderly Uninsured by Firm Size, 2001 ................ 17 Table 2 Comparison of Nonelderly, Uninsured Workers in Large Firms to Other Firm Workers and Other Uninsured, 2001............................... 18 Table 3 Insurance Status During the Year, Adults, 1998 ...................................... 19 Table 4 Access to Employer-Based Insurance of Uninsured Adults by Firm Size, 1998.................................................................................. 19 Table 5 Percentage Offered Health Insurance by Establishment Size, 1987 vs. 1998 ......................................................................................... 20 Table 6 Factors Affecting Change in Coverage in Large Firms, 1987–2001.............................................................................................. 20 Appendix Table A-1 Factors Predicting Establishments’ Offer of Health Insurance ............................................................... 23 Appendix Table A-2 Predictors of Percent of Uninsured Workers, 1987 2001 ............................................................................ 23 iv ABOUT THE AUTHORS Sherry Glied, Ph.D., is professor and chair of the Department of Health Policy and Management at Columbia University’s Mailman School of Public Health. Her research on health policy has focused on the financing of health care services in the United States. She is an author of recently published articles and reports on managed care, women’s health, child health, and health insurance expansions. She is currently conducting research sponsored by The Commonwealth Fund on the characteristics of uninsured Americans and on strategies to expand health insurance coverage. Professor Glied holds a B.A. in economics from Yale University, an M.A. in economics from the University of Toronto, and a Ph.D. in economics from Harvard University. Jeanne M. Lambrew, Ph.D., is an associate professor of health policy at George Washington University. She conducts policy-relevant research on Medicare, Medicaid and the uninsured, and long-term care. Dr. Lambrew worked on health policy at the White House from 1997 through 2001 as the program associate director for health at the Office of Management and Budget and as the senior health analyst at the National Economic Council. In these positions, she worked on the creation and implementation of the Children’s Health Insurance Program, development of the president’s Medicare reform plan and long-term care initiative, and implementation and oversight of Medicaid and disability policies. Prior to serving at the White House, Dr. Lambrew was an assistant professor of public policy at Georgetown University and a special assistant coordinating Medicaid and state studies at the Department of Health and Human Services. Dr. Lambrew has her master’s degree and Ph.D. from the Department of Health Policy, School of Public Health, at the University of North Carolina at Chapel Hill. Sarah Little graduated from Harvard University in 2002 with a B.A. in economics. She worked as a research assistant for Professor Sherry Glied at Columbia University’s Mailman School of Public Health in 2002–03. She currently is a first-year medical student at the University of California, San Francisco, School of Medicine. v EXECUTIVE SUMMARY A generally unreported phenomenon in recent years is the increasing number of U.S. workers in large firms who lack health insurance. Although large employers are much more likely than small ones to offer health coverage, recent evidence suggests that large-firm workers and their dependents comprise a significant and growing share of the working uninsured. A combination of factors seems to be responsible for falling health coverage rates within large firms: the decline in manufacturing jobs and unionization rates, restrictions placed on benefit eligibility and higher employee premium contributions, as well as service industry trends and the changing structure of large corporations. Because it is likely to persist, this trend has important implications for policies designed to reduce the number of uninsured Americans. This report profiles uninsured workers in large firms, compares their characteristics with other groups of uninsured, and assesses health coverage trends in small, mid-sized, and large firms. It also analyzes labor market changes that could affect this pattern. Major findings are summarized below. Uninsured Workers in Large Firms: A Growing Group Significant proportion of the uninsured. In 2001, about 9.6 million, or 26 percent, of the nation’s uninsured worked in, or had a family member working in, a large firm. This number exceeded the number of low-income uninsured children targeted by the State Children’s Health Insurance Program (CHIP) (6.3 million), the number of unemployed and uninsured adults (3.9 million), and the number of older adults ages 55 to 64 lacking insurance (3.2 million) in that year. Growing share of uninsured workers. From 1987 to 2001, the proportion of uninsured workers who were employed by firms with 500 or more employees grew from 25 percent to 32 percent (Figure ES-1). The proportion working in small businesses (fewer than 100 employees) or mid-sized ones (100 to 499 employees) declined. The rate of uninsured workers in large firms increased by 57 percent (from 7% to 11%), compared with a 25 percent increase for small firms and a 27 percent increase for mid-sized firms. vii Figure ES-1. Share of Uninsured Workers by Firm Size, 1987 2001 Large (500+) Medium (100–499) Percent Small (<100) 100 25 32 80 14 60 12 40 61 57 20 0 1987 2001 Source: Authors’ analysis of March Current Population Survey, 1988–2002. Similarities and Differences with Small-Firm Workers Similar low income profile as small-firm workers. Generally, uninsured workers working for large employers, much like their counterparts in small employers, have low incomes and are less likely to be married than insured workers. Over two of five (46%) low-income workers working for large employers had a time uninsured during the year (Figure ES-2). Compared with insured workers in large employers, uninsured workers are more likely to work part-time and be employed in industries that are less likely to offer coverage. viii Figure ES-2. Insurance Coverage of Workers in Large Firms by Household Income, Percent Uninsured During 1998 Percent Uninsured All Year Uninsured Part Year 50 46 40 18 30 20 14 28 5 8 10 3 9 5 0 Less than 200% FPL 200%–599% FPL 600%+ FPL Source: Authors’ analysis of 1998 Medical Expenditure Panel Survey Full-Year File. Greater access to job-based coverage. Among private sector businesses, large firms are more likely than small ones to offer their workers coverage. The largest firms (1,000 or more employees) are 9.7 times more likely than the smallest firms (fewer than 25 employees) to offer health insurance. The relationship between size and likelihood of offering coverage was even stronger when looking at the establishment level (an establishment is a whole or part of a firm at a single location). Large establishments (500 to 999 employees) were 11.4 times more likely than small establishments (fewer than 25 employees) to offer health insurance. Among large businesses, those with multiple, smaller establishments were less likely to offer coverage than those with a single establishment. Fewer getting coverage through own jobs. The decline in coverage among workers working for large employers has been concentrated among private sector firms. From 1987 to 2002, the proportion of private sector workers in large firms who were insured through their own job fell—from 71 percent to 66 percent. The rate for small-firm workers remained stable at approximately 46 percent, while it fell slightly, from 64 to 63 percent, for workers in mid-sized firms. Access to job-based coverage matters. More than 70 percent of uninsured workers in large firms reported they lacked access to job-based health coverage. Forty-four percent of large-firm workers who were not offered coverage by their ix employers were uninsured. Similarly, 47 percent of workers in small firms who were not offered coverage were also uninsured. Impact of a Changing Labor Market for the Private Sector Workforce Stagnant wage growth in large firms. While the proportion of workers with low income (less than 200 percent of poverty) declined in small and mid-sized firms from 1987 to 2001, the proportion remained the same—about 20 percent— in large private sector firms. Low income is strongly tied to lack of health insurance coverage. Decline in unionization rates. The rate of workers who are union members declined by one-third in large firms between 1987 and 2001—a greater decline than in small and mid-sized firms. In fact, in 2001, the proportion of large-firm workers who were union members was lower than the proportion in mid-sized firms. Decline in manufacturing jobs. Between 1987 and 2001, the proportion of workers in manufacturing jobs declined by 2 percentage points in small firms, 8 percentage points in mid-sized firms, and 11 percentage points in large firms. Increase in the ratio of establishments per firm. The number of establishments per firm increased from 1.21 in 1988 to 1.25 in 1999. Over the same period, the rates at which health insurance was offered and the rates at which workers were insured in small establishments fell. Given evidence that large firms are increasingly made up of a number of small establishments, the decline in access to coverage in small establishments could be contributing to the increase in uninsured workers in large firms. Which factors matter most? About 60 percent of the rise in both the proportion and rate of uninsured workers nationally who are employed by large firms can be attributed to the decline in manufacturing jobs and unionization rates. These findings suggest that to reduce the number of uninsured, policymakers will have to address growing gaps in employee health coverage at large firms. Reforms to achieve this goal include removal of barriers to coverage in firms that generally offer health benefits (e.g., waiting periods and restrictions for part-time workers); allowing large- business workers to participate in alternative group health coverage options (e.g., the Federal Employees Health Benefits Program); and requiring that all large firms offer and possibly contribute to coverage of employees. The study also suggests that policies to assist low-income people in affording health insurance should not exclude those who work in large firms. Doing so could leave millions of Americans without health coverage. x THE GROWING SHARE OF UNINSURED WORKERS EMPLOYED BY LARGE FIRMS INTRODUCTION Efforts to reduce the number of uninsured Americans have typically focused on small firms because they employ the majority of uninsured workers (Ginsburg, 1998; Yegian, 2002; Nichols, 1997). In 2001, approximately 68 percent of uninsured workers were in firms with fewer than 500 employees, 57 percent in firms with fewer than 100 employees, and 41 percent in firms with fewer than 25 employees.1 Small firms are less likely to offer health insurance to their workers: in 2003, about 65 percent of businesses with three to 199 workers offered health insurance, compared with 98 percent of those with 200 or more employees (Kaiser/HRET, 2003). As a consequence, federal and state policymakers have rallied around efforts to promote insurance among small-firm workers (107th Congresses, 2002a and 2002b; Silow-Carroll, 2001; Rosenberg, 2002; Yegian, 1998). Less attention has been devoted to uninsured workers in large firms. While fewer in number than those in small and mid-sized firms, large-firm employees without health coverage comprise a significant and growing proportion of the uninsured population. Approximately 9.6 million uninsured Americans are associated with businesses with 500 or more employees, and evidence presented in this study suggests that growth in this group of uninsured has been rapid and may continue to be so. The recession that began in 2001 disproportionately affected large firms—not just by reducing jobs, but also by reducing benefits. Businesses with 200 or more employees were significantly more likely to report that health insurance was their greatest concern (65 percent reported this, versus 53 percent of small firms) (Kaiser/HRET, 2002). Large firms were three times as likely as small firms to report that they were very likely to increase employees’ share of premiums as a way to control costs (Kaiser/HRET, 2003). Understanding the characteristics of uninsured workers in large firms, as well as trends in their health insurance coverage, is critical for assessing gaps in our coverage system and designing policies to fill them. Both incremental and comprehensive reform proposals often concentrate on employees of small firms and those without access to job- based coverage. Such policies may miss an important gap in coverage and forgo opportunities to possibly reduce the number of uninsured in large firms. 1 These numbers include workers who are not self-employed and who earned at least $3 an hour (in 2000 dollars). 1 This report examines the characteristics of the uninsured in large firms and why they comprise a growing proportion of the uninsured. (See Appendix for methodology.) It profiles the uninsured in large firms to identify similarities and differences between this group and other groups of uninsured. In addition, it assesses coverage trends in small, mid- sized, and large private sector firms from 1987 to 2001, as well as labor market trends that could affect this pattern. Firms are generally divided into three groups: small (fewer than 100 workers), mid-sized (100 to 499 workers), and large (500 or more workers). Some of the analyses were conducted both at the firm and establishment level or, because of data limitations, at the establishment level only. Specific trends in the U.S. large-firm labor market that could affect health coverage include: (1) workers’ poverty levels; (2) the decline in rates of unionization; (3) the shift in U.S. industries away from manufacturing and toward the service sector; and (4) the relationship between firms and their component establishments (an establishment is a whole or part of a firm at a single location). UNINSURED WORKERS IN LARGE FIRMS: SCOPE OF THE PROBLEM Number of Uninsured in Large Firms In 2001, about 9.6 million, or 26 percent, of the nation’s uninsured worked in or had a family member working for a large employer (Table 1). This includes 2.7 million of the nation’s 9 million uninsured children. These 9.6 million uninsured Americans represent nearly one of 11 workers and dependents in large firms—a smaller proportion than that in small firms (24%), but surprisingly high given that large firms are typically generous in their employee benefits. The number of uninsured associated with large employers is equal to or exceeds the number of uninsured targeted by recent policy initiatives. It is more than the 6.3 million low-income, uninsured children who are the focus of the State Children’s Health Insurance Program (CHIP) (U.S. Census Bureau, 2002). It exceeds the number of unemployed people who are uninsured—3.9 million, a number that has increased during this recession (Etheridge and Dorn, 2003). It is more than twice the number of uninsured adult ages 55 to 64 (3.2 million) who are the focus of Medicare buy-in policies (U.S. Census Bureau, 2002b). And it is higher than the Bush Administration’s estimates of the number of uninsured who would benefit from the president’s individual tax credit proposal (4 million) (Executive Office of the President, 2003). As such, the uninsured in large firms seem to be a group worthy of policy attention. Proportion of Uninsured in Large Firms From 1987 to 2001, the proportion of uninsured workers nationwide who were in firms with 500 or more employees increased from 25 percent to 32 percent (Figure 1). 2 Meanwhile, the proportion of uninsured in small and mid-sized firms declined. This trend is a reflection of two factors. First, the proportion of workers employed by large firms rose from 46 percent in 1987 to 48 percent in 2001. Second, the uninsured rate increased across the board. However, even if the share of workers across firm sizes had remained at 1987 levels, the share of the uninsured employed by large firms would still have increased over this period. This increase in the proportion of uninsured workers working for large employers was concentrated in large, private sector firms. Figure 1. Share of Uninsured Workers by Firm Size, 1987 2001 Large (500+) Medium (100–499) Percent Small (<100) 100 25 32 80 14 60 12 40 61 57 20 0 1987 2001 Source: Authors’ analysis of March Current Population Survey, 1988–2002. Uninsured Rate in Large Firms Although uninsured rates have risen at firms of all sizes, they appear to have risen fastest among large-employer workers (Figure 2). Including public and private sector workers, from 1987 to 2001, the uninsured rate increased from 20 percent to 25 percent in small firms, from 11 percent to 14 percent in mid-sized firms, and from 7 percent to 11 percent in large firms, according to the March Current Population Survey (CPS). Small firms experienced the largest percentage-point increase in their proportion of uninsured, but the rate of increase was greatest in large firms. The uninsured rate among large firms increased by 57 percent, compared with 25 percent in small firms and 27 percent in mid-sized firms. 3 Figure 2. Trend in Uninsured Rate by Firm Size, 1987 2001 Percent 1987 1991 1996 2001 30 24 25 23 20 20 13 14 14 11 10 11 10 9 7 0 Small (<100) Medium (100–499) Large (500+) Source: Authors’ analysis of March Current Population Survey, 1988–2002. Profile of Uninsured in Large Firms To get a sense of the characteristics of uninsured employees in large firms, they were compared with uninsured workers in small firms (Table 2). Like other workers without health coverage, the uninsured in large firms have disproportionately low incomes. In 2001, 20 percent were poor and 53 percent had income below 200 percent of the poverty threshold (about $38,800 for a family of four in 2003). A slightly higher proportion of uninsured workers in small and mid-sized firms had income below 200 percent of poverty (57%). Indeed, the problem of uninsurance in large firms is concentrated among low- income workers. While 46 percent of low-income workers in large firms are uninsured for some time during the year, only 14 percent of middle- and 8 percent of high-income workers in large firms are uninsured at any point during the year (Figure 3). Uninsured workers in large firms differed somewhat from other uninsured workers in terms of their family status: they were slightly more likely to be single than uninsured workers in small firms (65% vs. 59%). By contrast, only 40 percent of workers who were insured through large firms were single. 4 Figure 3. Insurance Coverage of Workers in Large Firms by Household Income, Percent Uninsured During 1998 Percent Uninsured All Year Uninsured Part Year 50 46 40 18 30 20 14 28 5 8 10 3 9 5 0 Less than 200% FPL 200%–599% FPL 600%+ FPL Source: Authors’ analysis of 1998 Medical Expenditure Panel Survey Full-Year File. While the sociodemographic profile of uninsured workers in large firms did not differ markedly from that of other uninsured workers, work patterns did. Uninsured employees of large firms were slightly more likely to work part-time than were those working in small firms (32% vs. 29%). Large-firm workers without insurance were nearly twice as likely to work part-time as were their insured counterparts (32% vs. 17%). Type of employment varied across these groups as well. Compared with those in small firms, uninsured workers in large firms were more concentrated in retail jobs (34% vs. 23%). A greater proportion of the uninsured who worked in small or mid-sized businesses were in agricultural or construction jobs. Duration of Uninsured Periods Lack of health insurance is more likely to be a temporary problem for workers in large businesses than it is for workers in small businesses, mainly because large-firm employees have greater access to job-based coverage. In 1998, about 43 percent of full-time workers in large establishments (defined here as 100 or more employees) who were uninsured in January gained some type of health insurance during the course of the year.2 By contrast, only 27 percent of workers in small establishments gained coverage during the year. 2 Note that, due to data limitations, we switched from using “firms” as the unit of analysis to “establishments,” and that a large firm is defined as having 100 or more employees rather than 500. 5 In both small and large establishments that offered their employees health insurance, about two-thirds of workers who were uninsured in January gained job-based coverage at some point that year. In firms not offering coverage, only 12 percent of workers in small establishments and 15 percent in large establishments gained coverage during the year. Income level does not appear to be a significant reason for the longer uninsured periods experienced by small-firm workers: about half of low-income, uninsured workers (below 200 percent of poverty) in both large and small businesses that offer health benefits enrolled in coverage during the year (Table 3). This suggests that poorer access to job- based insurance is a principal reason why small-firm employees have longer gaps without health coverage. Coverage Obtained Through Employee’s Own Firm—Private Sector Workforce Among all large employers, most of the change in coverage rates over time reflects shifts in coverage patterns for large private sector firms. Over the fifteen years 1987 to 2001, the rate at which large, private sector workers actually got coverage through their employment has declined, with large-firm workers faring much worse than workers at small or mid-sized firms. The rate at which workers were insured through their own job remained stable for small firms (about 40%), fell slightly for mid-sized firms (64% to 63%), and fell more dramatically for large firms (from 71% to 66%) (Figure 4). Figure 4. Trend in Workers with Insurance Through Own Job by Firm Size, 1987 2001 Percent 1987 1991 1996 2001 80 71 68 66 66 64 61 63 63 60 40 38 39 40 40 20 0 Small (<100) Medium (100–499) Large (500+) Source: Authors’ analysis of March Current Population Survey, 1988–2002. Private Sector Workers. 6 ACCESS TO HEALTH COVERAGE AT LARGE FIRMS Access to Health Coverage in Large Firms There are three basic explanations for why workers may not have health insurance: (1) they declined coverage when it was offered; (2) they are not eligible for the coverage offered by their employer; or (3) their firm does not offer health benefits. Surveys of employers show that the overwhelming percentage of large firms offer coverage (Kaiser/HRET, 2003). Surveys of workers, however, suggest that the majority of those at large firms who are uninsured were not offered health coverage. In 1998, 71 percent of uninsured workers in large firms were not offered health benefits by their employer, lower than the 84 percent of uninsured workers in small to mid-sized firms who were not offered coverage (Table 4). An additional 15 percent of uninsured workers in large firms were not eligible for the health benefits available to other employees. The remaining 14 percent of the uninsured in large firms were eligible for their employer’s health benefits but declined them. There appears to be little association between firm size and take-up of coverage among those employees who are offered health insurance. Of the 20 percent of all workers in large firms not offered health coverage through their job in 1998, 44 percent were uninsured. Similarly, of the 40 percent of workers in small firms not offered coverage, 47 percent were uninsured (data not shown). Low-income workers in large firms are less likely to be offered employer sponsored insurance than are higher income workers in these firms. Only 67 percent of low-income workers in large firms are offered employer sponsored insurance whereas 89 percent of mid- and high-income workers in large firms are offered health insurance by their employer. The following discussion examines the factors contributing to the increase in the uninsured rate for employees in large private sector firms. Why Do Some Large Firms Offer Coverage While Others Do Not? A firm’s size is linked to its likelihood of offering health insurance to its employees. With health insurance, economies of scale matter. Studies suggest that in addition to large groups’ ability to spread risk, their administrative costs are lower (Pauly, Percy, and Herring, 1999). But features other than size also affect whether an employer offers coverage; in fact, recent research suggests that workers’ income is a better predictor of lack of health benefits than firm size (Ferry et al. 2001). In addition, the likelihood of an establishment offering health coverage is 20 percent higher in manufacturing than in 7 service jobs, and 60 percent higher if some workers are union members, according to the National Center for Health Statistics (1997). Using regression analysis to assess the relative importance of factors affecting access to job-based coverage, we found that both firm and establishment size are significantly associated with insurance offer rates—even after controlling for key worker and employer characteristics (see Appendix Table A-1). The largest firms (1,000 or more employees) had 9.7 times greater odds of offering health insurance than the smallest firms (fewer than 25 employees). Large establishments (500 to 999 employees), meanwhile, were found to have 11.4 times greater odds of offering coverage than small establishments (fewer than 25 employees). Employee income, unionization, and type of job mattered as well. Compared with establishments whose average payroll per employee was less than $12,000, those with an average payroll of $12,000 to $25,000 had more than twice the odds of offering health insurance. Establishments whose average payroll was $25,000 to $50,000 had more than three times the odds of offering health insurance. Establishments where any employees were members of labor unions had 1.4 times greater odds of offering health benefits compared with establishments where no unionization had occurred. Relative to establishments in other industries, those in the mining/manufacturing sector had 1.4 times the odds of offering coverage, while those in the retail sector had only 0.63 times the odds. The complex relationship between firm size, establishment size, and insurance coverage has not been well addressed by researchers (Zarkin et al., 1995). Some surveys, such as one by the National Center for Health Statistics (1997), have shown that firms with multiple component establishments are more likely to offer health insurance than those that comprise a single establishment, but the effect of establishment size independent of firm size has not been investigated. The analysis we conducted suggests that focusing solely on firm size ignores workers in multi-establishment firms who are at greater risk of lacking health coverage. By predicting insurance offer rates for all the various combinations of establishment and firm size, we found that within the large-firm sector, large firms that have multiple, smaller establishments are less likely to offer health insurance coverage than those that have a single establishment (Figure 5). For example, single-establishment firms with 25 to 99 employees are more likely to offer coverage than firms that have 500 to 999 employees but multiple establishments with fewer than 25 workers. 8 Figure 5. Health Insurance Offer Rate by Firm and Establishment Size Establishment Size <25 25–99 100–499 500–999 1,000+ Offer Rate 95 98 98 99 98 100 92 96 89 91 95 86 85 81 80 75 60 50 40 20 0 <25 25–99 100–499 500–999 1,000+ Firm Size Source: Authors’ analysis of Robert Wood Johnson Foundation, Community Tracking Establishment Survey, 1997. LABOR MARKET CHANGES AND LARGE FIRMS The uninsured rate for workers in large firms has risen, even when taking into account the changes in the size of the workforce employed by such firms. This suggests that other factors may be at work. Prior research suggests four potential reasons for this growing problem: an increase in low-income workers; a decrease in unionization rates; a shift away from manufacturing jobs; or an increase in the number of smaller establishments within large firms. This study explores the trends and roles of each of these factors. Increase in Low-Income Workers It is well established that workers with low income are less likely to have access to job- based health insurance and to be insured. Although the poverty rate decreased in the last decade (prior to the recent recession), research suggests that disparities in employer- sponsored coverage by wage level have been growing (Medoff, 2001). This phenomenon could have contributed to the growth in the uninsured rate in large firms if the proportion of low-income workers in these firms increased as well. Indeed, evidence suggests that large firms have been shifting toward a lower-income workforce. While the percentage of 9 workers in small and mid-sized firms with incomes at or below 200 percent of the poverty level declined from 1987 to 2001, the percentage in large firms increased slightly, although it decreased in the last five years (Figure 6). Figure 6. Share of Workers with Incomes Below 200% FPL by Firm Size, 1987 2001 Percent 1987 1991 1996 2001 40 34 32 32 30 28 27 25 26 22 22 22 20 20 20 10 0 Small (<100) Medium (100–499) Large (500+) Source: Authors’ analysis of March Current Population Survey, 1988–2002. Private Sector Workers. Decline in Unionization Rates The steady decline in unionization rates among private employers means that fewer workers today have access to the health benefits that many labor unions have traditionally secured for their members. In 1987, 14 percent of workers reported that they were a member of a labor union. By 2001, the proportion had fallen to 10 percent. Decline in unionization rates was sharpest for employees in large firms. In 1987, 20 percent of large- business employees reported they were union members, compared with 15 percent in mid-sized firms and 6 percent in small firms. By 2001, unionization levels fell 7 percentage points in large firms, 4 percentage points in mid-sized firms, and were basically stable in small firms (Figure 7). In other words, there has been a one-third decline in the rate of unionized workers in large firms. 10 Figure 7. Percent of Workers in Unions by Firm Size, 1987 2001 Percent 1987 1991 1996 2001 30 20 20 18 14 15 15 12 11 12 11 11 13 10 10 6 5 6 6 0 All Small (<100) Medium Large (500+) (100–499) Source: Authors’ analysis of March Current Population Survey, 1988–2002. Private Sector Workers. Decline in Manufacturing Jobs Perhaps due to a greater historical presence of unionized workers, the manufacturing sector has been the most likely to offer jobs with health coverage. Medoff and colleagues (2001) found that employees in manufacturing jobs are approximately 1.5 times as likely to have health insurance through their own employer compared with workers in other sectors. Yet these jobs, too, are on the decline as the U.S. labor market shifts away from manufacturing. In 1987, 23 percent of private workers were employed in manufacturing jobs; that proportion had fallen to 17 percent by 2001. Manufacturing firms tend to be large; thus the decline in manufacturing was most acute in large firms. In 1987, 31 percent of employees in large firms, 32 percent in mid-sized firms, and 13 percent in small firms were in the manufacturing sector. Between 1987 and 2001, there was an 11 percentage- point decline in the proportion of large-business workers employed in manufacturing, an 8 percentage-point decline in mid-sized firms, and a 2 percentage-point decline in small firms (Figure 8). Thus, large firms experienced both the largest percentage-point and percent decline in manufacturing jobs, with the rate of such jobs dropping by one-third. 11 Figure 8. Percent of Workers in Manufacturing Jobs by Firm Size, 1987 2001 Percent 1987 1991 1996 2001 40 32 31 29 29 28 30 25 23 22 24 20 20 20 17 13 13 12 11 10 0 All Small (<100) Medium Large (500+) (100–499) Source: Authors’ analysis of March Current Population Survey, 1988–2002. Private Sector Workers. More Establishments per Firm The little-investigated trend among large firms toward having a greater number of small, component establishments could also be related to the decline in workers’ access to coverage in large companies. As described above, small establishments within large firms are less likely to offer coverage than single-establishment firms of the same size. Moreover, between 1988 and 1999, the growth rate for establishments (16%) exceeded that for firms (13%) (U.S. Census Bureau, 1999). Consequently, the number of establishments has increased from 1.21 per firm in 1988 to 1.25 per firm in 1999. In addition, the uninsured rate in small establishments rose more rapidly than that in large establishments between 1987 and 1998, as did the share of uninsured working in small establishments. Health insurance offer rates, meanwhile, declined in both large and small establishments, by about the same percentage (see Table 5). Given evidence that large firms are increasingly made up of a greater proportion of small establishments, the decline in health coverage and access to coverage in small establishments could be behind some of the increase in uninsured workers in large firms. Which Factor Matters Most? Income, unionization, and manufacturing are all positively correlated with changes in health insurance coverage—and large-firm employees have fared the worst on all three variables. Indeed, in our regression analysis, we found that the percentage of large-firm 12 workers in manufacturing, in unions, and earning less than 200 percent of the poverty level all have statistically significant effects on the proportion who are uninsured in large firms (see Appendix Table A-2). But what is the relative impact of each of these variables on the uninsured rate in large firms? Our analysis found that the shift away from manufacturing from 1987 to 2001 accounted for 18 percent of the increase in the rate of uninsured workers and 20 percent of the increased share of uninsured in large firms (Figure 9 and Table 6). This is in line with, though somewhat lower than, prior estimates that approximately 30 percent of the decline in employee coverage can be accounted for by shifts in industry and occupation (Medoff et al. 2001). The reduced rate of unionization among workers in large firms accounted for 38 percent of the rise in the uninsured rate and 43 percent of the increase in the proportion of uninsured workers in large firms—again, consistent with prior estimates that 20 to 35 percent of the decline in employee health coverage can be attributed to decline in unionization rates (Buchmueller 2002). The change in the proportion of workers below 200 percent of poverty in large firms, however, had a negligible impact on the uninsured rate in large firms, mainly because of the very low percentage increase in low-wage workers in large firms (0.1%). Figure 9. Factors Affecting Change in Coverage in Large Firms, 1987 2001 Other Percent Manufacturing Decline Unionization Decline 100 80 44 37 60 20 18 40 20 38 43 0 Increase in Uninsured Rate Increased Share of in Large Firms All Uninsured Workers Employed in Large Firms Source: Authors’ analysis of March Current Population Survey, 1988 and 2002. Private Sector Workers. Based on Regression Results in Appendix Table A-2. CONCLUSIONS AND RECOMMENDATIONS While much attention has been devoted to uninsured workers in small firms, this study finds workers in large firms represent a faster-growing segment of the uninsured labor 13 force. The 9.6 million uninsured large-firm workers and their dependents are not dissimilar to other groups of the uninsured. But compared with other workers in large firms, they tend to work in firms that are less likely to offer employee health benefits. Labor market changes—lower unionization rates, the disappearance of manufacturing and other jobs that have historically included comprehensive health coverage, and the growing number of smaller establishments at large firms—appear to contribute to the erosion of health coverage for those employed by large businesses. Of these trends, the declines in unionization rates and in manufacturing jobs are responsible for nearly two-thirds of the increased proportion of uninsured at large firms. These results deserve further examination in at least two respects. The first is the relationship between the size of a firm and the size of its component establishments. Our study suggests that a small establishment (fewer than 25 workers) in a large firm (500 to 999 workers) is less likely to offer coverage than a mid-sized firm (25 to 99 workers) comprising a single establishment. Although policies often target small-business employees for insurance expansion initiatives, small-establishment workers should also be considered. Second, our analysis suggests that, contrary to previous studies, the uninsured problem in large firms stems from employers’ failure to offer coverage to any workers, not from these workers’ ineligibility for an existing company health plan. Given the small percentage of large firms that do not offer coverage and the difference from other studies (e.g., Brown et al. 2002), this is surprising. It is possible that this finding is a reporting problem—that is, some survey respondents may think their employer does not offer health benefits when, in fact, they do, although they themselves are not eligible for them. Although still an access problem, it is one that would require different solutions. From a policy perspective, the growing uninsured problem in large firms is a serious one. The causes of the problem are largely intractable, since they relate to broad- based changes in the labor market. In addition to the structural changes described in this study, other research has found that the increased use of contingent workers, stricter policies with regard to access to health insurance for part-time workers, and increased waiting periods, among other practices, could be contributing to this problem. These changing employment policies and work relationships are difficult to modify through public policy. And while they are a meaningful fraction of the uninsured, the uninsured in large firms nevertheless represent a small fraction of all workers in large firms. Identifying and assisting these workers is akin to finding the proverbial needle in a haystack. The problem of “crowd out”—which occurs when a large proportion of funding for public programs goes to people who already have private health coverage—would likely create hurdles for any incremental reform effort. 14 Despite these challenges, policymakers could pursue a number of reforms to improve access to health insurance for workers in large firms. At the incremental level, policies could be designed to limit gaps in coverage that occur when uninsured workers have to wait a defined period before their health benefits are activated. In 1999, half of firms with 5,000 or more employees and over two-thirds of those with 200 to 4,999 employees had to endure a waiting period before coverage began (Gabel et al., 2001). Other research shows that eliminating coverage gaps could protect workers in transition: studies find that workers who experience even short spells without health insurance have problems accessing health care and paying medical bills that are very similar to those of the long-term uninsured (Duchon et al. 2001; Schoen and DesRoches 2000; Hoffman 2001). Policymakers could seek to limit or eliminate waiting periods in large firms, or provide incentives or regulations that would make more part-time workers eligible for coverage. An even larger proportion of uninsured workers in large firms have reported that their firms do not offer coverage. In tackling this problem among small firms, policymakers have considered and, in some cases, implemented tax breaks, exemptions from regulation, and other incentives to encourage large firms that do not offer coverage to do so. However, experience suggests that such programs achieve only limited success (Rosenberg, 2002; Mitchell and Osber, 2002). This leaves two options: create an alternative source of health coverage or require large firms to offer coverage. Several of the major reform proposals (Davis and Schoen, 2003; Collins et al., 2003) allow individuals or large firms themselves to buy into a nationwide purchasing pool similar to the Federal Employees Health Benefits Program. To avoid adverse selection, such a plan would have to prevent the use of this pool by only sicker and lower-income workers in large firms (e.g., allow individual workers in large firms to participate only if their employer makes it an option for all workers). Even so, allowing large firms to participate could stabilize such a pool, given their size and number of employees. Others reform plans, as well as a new California law, would require that all large firms offer health coverage. This idea, a component of the least comprehensive health reform proposals in 1993–94, includes requirements that most employers at least offer, if not contribute to, health benefits for workers.3 If such a policy were restricted to firms with 500 or more employees, only a small fraction of the 16,700 firms of this size would be affected (U.S. Census, 1999). That said, a mandate raises concerns, including whether those firms affected by the policy are competing with small firms that do not offer coverage, and whether a mandate provides an incentive for firms to contract work out to reduce payroll size, among other labor market distortions (Steurle, 1994; Long and 3 Affordable Health Care Now Act of 1994, 103rd Cong., 2nd session, H.R. 5300. 15 Marquis, 1993). Despite the political challenge of requiring firms to offer coverage, the federal cost of such a mandate would be low and its pay-off could be high. Finally, comprehensive health reform plans may not be able to achieve universal coverage if they do not recognize the need to provide access to affordable coverage for workers in large firms. A number of previous proposals would have limited financial assistance to low-income people who do not work for large firms or firms that offer coverage to some of their workers. Yet, uninsured workers in large businesses—like all the uninsured—are disproportionately low-income and may not be able to afford coverage. A meaningful percentage of these workers, moreover, cannot access health coverage because of eligibility restrictions or waiting periods, or because their firm simply does not provide health benefits. Our study suggests that plans for universal coverage must address coverage gaps in large firms in order to achieve their goal of insuring all Americans. 16 TABLES Table 1. Distribution of the Nonelderly Uninsured by Firm Size, 2001 Working Adults Non-Working Adults Children Total Number Number Number Number Firm Size (millions) Percent (millions) Percent (millions) Percent (millions) Percent Large Firm (500+) 6.1 30 0.8 10 2.7 29 9.6 26 Medium Firm (100 499) 2.0 10 0.3 4 0.9 10 3.2 9 Small Firm (1 99) 9.2 45 1.3 17 3.1 34 13.6 36 Nonworker/Unknown 3.2 16 5.4 69 2.6 27 11.2 30 Total 20.5 100 7.8 100 9.3 100 37.6 100 Note: Individuals were classified as being associated with a large firm if any individual in the family worked in a large firm. Universe is all nonelderly people excluding working adults with hourly wages $3 and lower. Source: March CPS, 2002. 17 Table 2. Comparison of Nonelderly, Uninsured Workers in Large Firms to Other Firm Workers and Other Uninsured, 2001 Small and Large Firms (500+ workers) Mid-Sized Firms Own ESI Uninsured Uninsured Distribution (percent of total) (41.3 million) (5.0 million) (13.2 million) Income (as percent of federal poverty level) Less than 100 percent 2% 20% 24%* 100 percent to 199 percent 10 33 33 200 percent to 599 percent 57 40 37 600 percent and more 31 7 6 Family status Single, no children 32 56 50* Single, children 8 9 9 Married, no children 28 17 20 Married, children 32 18 21 Work status Full time 83 68 71* Part time 17 32 29 Industry Agriculture, forestry, fisheries, minerals 1 1 7* Construction industry 2 3 17 Manufacturing industries 19 11 10 Transportation, communication & utilities 12 8 5 Wholesale trade 3 2 3 Retail trade 11 34 23 Finance, insurance, real estate 8 4 4 Service industries 35 34 32 Public administration 10 3 <1 * Indicates a statistically significant difference across categories in a chi-squared test (p<0.01). Note: The number of uninsured, adult workers in the large-firm category in Table 1 is greater than the number of uninsured large-firm workers in Table 2 because the amount in Table 1 includes all workers living in a household with a large-firm worker. So if a small-firm worker is living with a large-firm worker, they are both considered adult workers in the large-firm category. Table 2 shows only workers who work for large firms, excluding family members. Source: March CPS, 2002. All workers. 18 Table 3. Insurance Status During the Year, Adults, 1998 Uninsured Insurance Status of All Adults in January (percent of row) Who Gained Insured Insured Uninsured Coverage Full Year Part Year Full Year Large establishments 43% 85% 10% 6% Offered 64 88 10 2 Income (as percent of federal poverty level) Less than 200 percent 54 60 30 10 200 percent to 599 percent 63 89 9 2 600 percent or more 83 94 6 0 Not offered 15 56 11 33 Income (as percent of federal poverty level) Less than 200 percent 26 41 21 38 200 percent to 599 percent 7 61 8 31 600 percent or more 0 68 0 32 Small establishments 27% 71% 13% 16% Offered 64 83 14 3 Income (as percent of federal poverty level) Less than 200 percent 54 64 27 10 200 percent to 599 percent 70 85 13 2 600 percent or more 70 92 7 1 Not offered 12 50 12 39 Income (as percent of federal poverty level) Less than 200 percent 11 29 16 55 200 percent to 599 percent 12 54 10 36 600 percent or more 15 75 7 18 Source: 1998 Medical Expenditure Panel Survey Full-Year File. Nonelderly working adults. Table 4. Access to Employer-Based Insurance of Uninsured Adults by Firm Size, 1998 Distribution Work Status (of row) of Uninsured Full-Time Part-Time Large firms Eligible, not enrolled 14% 66% 34% Firm offers, not eligible 15 43 57 Firm does not offer 71 72 28 Small firms Eligible, not enrolled 9% 74% 26% Firm offers, not eligible 7 52 48 Firm does not offer 84 69 31 Source: Matched February–March 1999 CPS. All workers. 19 Table 5. Percentage Offered Health Insurance by Establishment Size, 1987 vs. 1998 Establishment Size <100 100+ Percent offered 1987 68.8% 97.7% Percent offered 1998 56.6% 85.0% Percentage point change –12.2% –12.7% Universe: All workers not in public administration. Source: Authors’ calculations from the 1987 National Medical Expenditure Survey and 1998 Medical Expenditure Panel Survey. Table 6. Factors Affecting Change in Coverage in Large Firms, 1987–2001 Percent Uninsured Share of the Among Workers in Uninsured Employed in 500+ Employee Firms 500+ Employee Firms 1987 Data 7.8% 22.0% 2001 Data 12.0 28.6 If there had been no decline in large firm manufacturing 11.3 27.3 Contribution of manufacturing 0.8 1.3 to change (17.7% of total change) (19.5% of total change) If there had been no decline in large-firm unionization 10.4 25.8 Contribution of de-unionization 1.6 2.9 to change (38.3% of total change) (43.1% of total change) If there had been no change in manufacturing or unionization 9.7 24.3 Contribution of both factors 2.4 4.3 to change (56.0% of total change) (64.3% of total change) Source: Authors’ calculations from the March Current Population Survey based on regression results in Appendix Table A-2. Universe is all private sector workers who earned at least $3 an hour (in 2000 dollars). 20 APPENDIX. METHODOLOGY No single source of data contains information on individuals, firms, and establishments over time to enable a consistent set of analyses to answer the questions addressed in this study. Because this paper aims to produce a broad understanding of the characteristics and trends of health coverage in large firms, it draws upon four different sources of data. First, the March supplements of the Current Population Survey (CPS) were used for most population-based data on the characteristics and trends in coverage of workers in large firms. This is the data set most widely used to assess the nation’s health insurance coverage and has sufficient sample size to make relevant comparisons. Second, the February supplement to the CPS was merged with the March supplement to enable analysis of access to job-based health coverage. Third, the 1998 Medical Expenditure Panel Survey (MEPS) Household Component Full Year Consolidated data file was used to gauge access to job-based coverage and take-up rates among workers. Fourth, the establishment survey of the 1997 Robert Wood Johnson Community Tracking Survey (RWJ CTS) was examined to understand the relative roles of unionization, manufacturing, income, and establishment size on health insurance coverage among private sector firms. In the analysis of factors contributing to changes over time, we restrict the population to non-government, private sector workers. All analyses of workers define workers as people who are not self-employed and who earned at least $3 an hour (in 2000 dollars). Firm size in this paper is generally broken into three groups: small (fewer than 100 employees), medium (100 to 499 employees), and large (500 or more employees). Although many studies define a large firm as having 100 or more employees (see Brown et al. 1990), we chose the higher threshold to ensure that we were clearly capturing trends in large versus mid-sized firms. Firms with 500 or more workers employ nearly half of the nation’s workers (U.S. Census Bureau 1999). However, when using MEPS and RWJ CTS, different ranges were used due to data limitations. The MEPS categorized employers by establishment size rather than by firm size, so that it counts employees at a particular site rather than at an entire firm, which may operate multiple sites. As such, some individuals who would be classified as “large-firm workers” in the CPS analysis may be classified as “small-establishment workers” in the MEPS analysis. Generally, the analyses focus on the size of the firm where nonelderly adults work; the estimate of the total number of uninsured associated with large firms includes any uninsured person who has in the health insurance unit at least one worker employed by a large firm. 21 We used the CPS revised methodology for measuring the uninsured. This defines health insurance hierarchically, so that each individual is assigned one health insurance category, even when there is more than one reported source of coverage during the year. “Adults” are defined as individuals ages 19 through 64 inclusive. A family is defined as a health insurance unit—a smaller family unit than that used by the Census Bureau, which includes more distant relatives and others living in the household—so less income is counted and thus slightly more low-income uninsured people are reported here. This report includes analyses to isolate the factors associated with changes in and offers of health coverage in large firms. Two-tailed t-tests were used to determine whether differences between the uninsured in large and small firms were statistically significant. Using the 1997 RWJ CTS, we were able to isolate the impact of unionization, manufacturing, income, and establishment size on health insurance coverage. We ran a multi-variable regression on the log likelihood that an establishment would offer health insurance, controlling for age, gender, payroll, years in business, proportion of permanent workers, firm size, establishment size, industry, unionization, and worker poverty levels. To decompose the effect of each variable on the increase in uninsurance in large firms, a multi-variable time-series regression using March CPS data was conducted. We regressed the percent of large-firms workers who were uninsured (by state and year) against the percent of small-firm workers who were uninsured, the percent of medium firm workers who were uninsured, the percent of large-firm workers who were employed in the manufacturing sector, the percent of large-firm workers who were in labor unions, and the percent of large-firm workers who were at or below 200 percent of the federal poverty level. The percent of small- and medium-firm workers who are uninsured captures economy-wide trends that would be expected to affect all firms in the same way. 22 Appendix Table A-1. Factors Predicting Establishments’ Offer of Health Insurance Odds Ratio Variable (Standard Error) Establishment size 25–99 1.982506 (.1540721) Establishment size 100–499 4.120082 (.6428094) Establishment size 500–999 11.41969 (6.835156) Establishment size 1,000+ 4.463251 (2.078081) Firm size 25–99 2.973043 (.2152067) Firm size 100–499 5.651469 (.5412479) Firm size 500–999 4.117962 (.6526941) Firm size 1,000+ 9.736934 (.9998436) Industry = Manufacturing/Mining 1.410977 (.0816848) Industry = Retail .6348515 (.0284489) Establishment has employees who are labor union members 1.408841 (.1461532) Percentage of permanent workers under age 30 1.003189 (.0008067) Percentage of permanent workers between 30 and 39 years of age 1.00365 (.0007882) Percentage of permanent workers between 40 and 49 years of age 1.004191 (.0008353) Number of years the establishment has been in business 1.011977 (.0008537) New establishment (in business a year or less) .5676975 (.0708407) Average payroll per employee 1.000009 (8.96e-07) Average payroll is $12,000–$25,000 2.050729 (.0886526) Average payroll is $25,000–$50,000 3.375436 (.1678068) Average payroll is $50,000–$100,000 2.561801 (.2324831) Percent of workers that are permanent 1.008401 (.0009407) Percent of permanent workers that are female .9987435 (.0005332) Source: 1997 Robert Wood Johnson Community Tracking Survey. Establishment-Level Data: Dependent variable is whether or not the establishment offers health insurance. Appendix Table A-2. Predictors of Percent of Uninsured Workers, 1987 2001 Odds Ratio Variable (Standard Error) Constant .03 (.01) Percent uninsured in small firms (less than 100 employees) .33 (.02) Percent uninsured in medium firms (100–499 employees) .07 (.02) Percent of large firm employees in the manufacturing sector –.07 (.01) Percent of large firm employees in labor unions –.24 (.05) Percent of large firm employees with incomes less than 200% of the federal poverty level .03 (.01) Source: Regression analysis using the 1988–2002 March Current Population Surveys. Universe is all private- sector workers who earned at least $3 per hour (2000 dollars). Dependent variable is the percent of uninsured large-firm workers (workers employed in firms with 500 or more employees) by year and by state. Errors clustered around a term combining both state and year. 23 REFERENCES 107th Congress, S. 2035 (2002a). “Health Plan Purchasing Alliance Act of 2002,” Senators Jeffords (I-VT), Clinton (D-NY), Snowe (R-ME). 107th Congress, S. 2679 (2002b). “Health Insurance Access Act of 2002,” Senators Baucus (D-MT), Smith (R-WA), Torricelli (D-NJ). Brown, C., J. Hamilton et al. (1990). Employers Large and Small (Cambridge, Mass.: Harvard University Press). Brown E. R. et al. (2001). The State of Health Insurance in California: Findings from the 2001 California Health Interview Survey (Los Angeles: UCLA Center for Health Policy Research). Buchmueller, T., J. DiNardo et al. (2002). “Union Effects on Health Insurance Provision and Coverage in the United States,” Industrial and Labor Relations Review 55 (4): 610–27. Collins, S., K. Davis, and J. Lambrew (September 2003). Health Care Reform Returns to the National Agenda: The 2004 Presidential Candidates’ Proposals (New York: The Commonwealth Fund). Current Population Survey (March 1988–2002). Duchon, L. et al. (2001). Security Matters: How Instability in Health Insurance Puts U.S. Workers at Risk (New York: The Commonwealth Fund). Etheridge, Lynn, and Stan Dorn (2003). Health Insurance for Laid-Off Workers: A Time for Action (Washington, D.C.: Economic and Social Research Institute). Available at http://www.esresearch.org/newsletter/january03/stan_lynn.pdf. Executive Office of the President (2003). Budget of the United States Government, Fiscal Year 2004 (Washington, D.C.: U.S. Government Printing Office). Note: This estimate represents the cumulative number who would be helped throughout the year. Thus, it is not directly comparable to the other estimates which are point-in-time estimates. Ferry, D. H., B. Garrett et al. (2002). “Health Insurance Expansions for Working Families: A Comparison of Targeting Strategies,” Health Affairs 21 (4): 246–54. Gabel, J. R. et al. (2001). “Embraceable You: How Employers Influence Health Plan Enrollment,” Health Affairs 20 (4): 196–208. Ginsburg, P. B., J. R. Gabel et al. (1998). “Tracking Small-Firm Coverage, 1989–1996,” Health Affairs 17 (1): 167–71. Hoffman, C. et al. (2001). “Gaps in Health Coverage Among Working-Age Americans and Its Consequences,” Journal of Health Care for the Poor and Underserved 12 (3): 272–89. Henry J. Kaiser Family Foundation/Health Research and Educational Trust (2002). Employer Health Benefits 2002 Annual Survey (Menlo Park, Calif., and Chicago, Ill.: Kaiser Family Foundation and HRET). Henry J. Kaiser Family Foundation/Health Research and Educational Trust (2003). Employer Health Benefits 2003 Annual Survey (Menlo Park, Calif., and Chicago, Ill.: Kaiser Family Foundation and HRET). Medical Expenditure Panel Survey (1987, 1998). Medoff, J. L., H. B. Shapiro et al. (2001). How the New Labor Market Is Squeezing Workforce Health Benefits (New York: The Commonwealth Fund). Mitchell, J. B., and D. S. Osber (2002). “Using Medicaid/SCHIP to Insure Working Families: The Massachusetts Experience,” Health Care Financing Review 23 (3): 35–45. 24 National Center for Health Statistics (1997). “Employer-Sponsored Health Insurance: State and National Estimates” (Hyattsville, Md.: NCHS). Nichols, L., L. Blumberg et al. (1997). Small Employers: Their Diversity and Health Insurance (Washington, D.C.: Urban Institute). Pauly, M., A. Percy, and B. Herring (1999). “Individual Versus Job-Based Health Insurance: Weighing the Pros and Cons,” Health Affairs 18 (6): 28–44. Robert Wood Johnson Foundation (1997). Community Tracking Survey. Establishment Survey (Princeton, N.J.: RWJF). Rosenberg, S. N. (2002). Lessons from a Small Business Health Insurance Demonstration Project (New York: The Commonwealth Fund). Schoen, C., and C. DesRoches (2000). “Uninsured and Unstably Insured: The Importance of Continuous Insurance Coverage,” Health Services Research 35 (1, Part II): 187–206. Silow-Carroll, S., E. K. Waldman et al. (2001). Expanding Employment-Based Health Coverage: Lessons from Six State and Local Programs (New York: The Commonwealth Fund). Steuerle, C. E. (1994). “Implementing Employer and Individual Mandates,” Health Affairs 13 (2): 54–68; and S. H. Long and M. S. Marquis (1993). “Gaps in Employer Coverage: Lack of Supply or Lack of Demand?” Health Affairs 12 (Suppl.): 282–93. U.S. Census Bureau. Statistics of U.S. Businesses, 1988–1999. U.S. Census Bureau, Department of Commerce (1999). “Statistics of U.S. Businesses, 1999: All Industries by Employment Size of Enterprise” (Washington, D.C.: U.S. Census Bureau). Available at http://www.census.gov/epcd/susb/1999/us/US--.htm. U.S. Census Bureau, Department of Commerce (2000). “Low Income Uninsured Children by State: 1998, 1999 and 2000” (Washington, D.C.: U.S. Census Bureau). Available at http://www.census.gov/hhes/hlthins/liuc00.html. U.S. Congressional Budget Office, H.R. 3090 (2001). “Economic Security and Assistance for American Workers Act of 2001” (Washington, D.C.: CBO). U.S. Census Bureau (2001). “Table HI-2: Health Insurance Coverage Status and Type of Coverage—All People by Age and Sex, 1987 to 2001” (Washington, D.C.: U.S. Census Bureau). Available at http://www.census.gov/hhes/hlthins/historic/hihistt2.html. Yegian, J. M., T. C. Buchmueller et al. (1998). Health Insurance Purchasing Alliances for Small Firms: Lessons from the California Experience (Oakland: California HealthCare Foundation). Yegian, J. M. (2002). Why Don’t More Small Businesses Offer Health Insurance? (Oakland: California HealthCare Foundation). Zarkin, G. A., S. A. Garfinkel et al. (1995). “Employment-Based Health Insurance: Implications of the Sampling Unit for Policy Analysis,” Inquiry 32 (3): 310–19. 25 RELATED PUBLICATIONS In the list below, items that begin with a publication number are available from The Commonwealth Fund by calling our toll-free publications line at 1-888-777-2744 and ordering by number. These items also can be found on the Fund’s website at www.cmwf.org. Other items are available from the authors and/or publishers. #671 Health Care Reform Returns to the National Agenda: The 2004 Presidential Candidates’ Proposals (September 2003). Sara R. Collins, Karen Davis, and Jeanne M. Lambrew. This analysis reviews the health reform proposals of the candidates for the 2004 presidential election, comparing the numbers of uninsured who would be covered under each plan as well as the estimated costs. #654 American Health Care: Why So Costly? (June 11, 2003). Karen Davis, The Commonwealth Fund. In invited testimony before a Senate Appropriations subcommittee hearing on rising health care costs, the Fund’s president outlined a number of steps that need to be taken to achieve a high- performing, accessible health system, including: public reporting of health care cost and quality data, establishment of quality standards, broad-scale demonstrations of new approaches to insurance coverage, investment in modern information technology and improved care processes, provider performance incentives, and elimination of waste and ineffective care. #657 Creating Consensus on Coverage Choices (April 23, 2003). Karen Davis and Cathy Schoen, The Commonwealth Fund. Health Affairs Web Exclusive. In this article, the authors propose an innovative framework to provide automatic, affordable health insurance to nearly all Americans. The approach would combine tax credits for private insurance with public program expansions. It would also promote insurance efficiencies through automatic enrollment, use of information technology, and group coverage. The framework could be phased in over time and modified along the way. Available online at http://www.healthaffairs.org/WebExclusives/2203Davis.pdf/. #622 Time for Change: The Hidden Cost of a Fragmented Health Insurance System (March 2003). Karen Davis, The Commonwealth Fund. In invited testimony before the Senate Special Committee on Aging, Fund president Karen Davis detailed the failure of the U.S. health care system to meet the objectives of ensuring access to needed medical care and protecting Americans from the financial burden of costly medical bills. Calling the system “costly, complex, and confusing,” Davis said the solution requires automatic and affordable health insurance coverage for all Americans and shared responsibility for financing coverage. #596 Expanding Health Insurance Coverage: Creative State Solutions for Challenging Times (January 2003). Sharon Silow-Carroll, Emily K. Waldman, Heather Sacks, and Jack A. Meyer, Economic and Social Research Institute. The authors summarize lessons from 10 states that have innovative strategies in place for health insurance expansion or have a history of successful coverage expansion. The report concludes with recommendations for federal action that could help states maintain any gains in coverage made and possibly extend coverage to currently uninsured populations. #585 Small But Significant Steps to Help the Uninsured (January 2003). Jeanne M. Lambrew and Arthur Garson, Jr. A number of low-cost policies could ensure health coverage for at least some Americans who currently lack access to affordable insurance, this report finds. Included among the dozen proposals outlined is one that would make COBRA continuation coverage available to all 26 workers who lose their job, including employees of small businesses that are not currently eligible under federal rules. #527 Are Tax Credits Alone the Solution to Affordable Health Insurance? Comparing Individual and Group Insurance Costs in 17 U.S. Markets (May 2002). Jon R. Gabel, Kelley Dhont, and Jeremy Pickreign, Health Research and Educational Trust. This report identifies solutions that might make tax credits and the individual insurance market work. These include raising the amount of the tax credits; adjusting the credit according to age, sex, and health status; and combining tax credits with new access to health coverage through existing public or private group insurance programs. #512 Security Matters: How Instability in Health Insurance Puts U.S. Workers at Risk (December 2001). Lisa Duchon, Cathy Schoen, Michelle M. Doty, Karen Davis, Erin Strumpf, and Stephanie Bruegman. This report, based on The Commonwealth Fund 2001 Health Insurance Survey, finds that in the past year one of four Americans ages 19 to 64—some 38 million adults—was uninsured for all or part of the time. Lapses in coverage often restrict people’s access to medical care, cause problems in paying medical bills, and even make it difficult to afford basic living costs such as food and rent. #478 Universal Coverage in the United States: Lessons from Experience of the 20th Century (December 2001). Karen Davis. This issue brief, adapted from an article in the March 2001 Journal of Urban Health: Bulletin of the New York Academy of Medicine, traces how the current U.S. health care system came to be, how various proposals for universal health coverage gained and lost political support, and what the pros and cons are of existing alternatives for expanding coverage. #415 Challenges and Options for Increasing the Number of Americans with Health Insurance (January 2001). Sherry A. Glied, Joseph A. Mailman School of Public Health, Columbia University. This overview paper summarizes the 10 option papers written as part of the series Strategies to Expand Health Insurance for Working Americans. #438 A 2020 Vision for American Health Care (December 11/25, 2000). Karen Davis, Cathy Schoen, and Stephen Schoenbaum. Archives of Internal Medicine, vol. 160, no. 22. The problem of nearly 43 million Americans without health insurance could be virtually eliminated in a single generation through a health plan based on universal, automatic coverage that allows choice of plan and provider. The proposal could be paid for, according to Fund President Davis and coauthors, by using the quarter of the federal budget surplus which results from savings in Medicare and Medicaid. 27