The Economic Burden of Providing Health Insurance

The Economic Burden of Providing Health Insurance: How Much Worse Off Are Small Firms?

Christine Eibner
Copyright Date: 2008
Edition: 1
Published by: RAND Corporation
Pages: 80
https://www.jstor.org/stable/10.7249/tr559emkf
  • Cite this Item
  • Book Info
    The Economic Burden of Providing Health Insurance
    Book Description:

    More than 60 percent of nonelderly Americans receive health-insurance (HI) coverage through employers. However, rising health-care costs may threaten the long-term viability of the employer-based insurance system. This report explores trends in the economic burden associated with HI provision for small and large businesses, as well as the quality of plans that small and large firms offer.

    eISBN: 978-0-8330-4782-3
    Subjects: Health Sciences, Management & Organizational Behavior

Table of Contents

  1. Front Matter
    (pp. i-ii)
  2. Preface
    (pp. iii-iv)
  3. Table of Contents
    (pp. v-vi)
  4. Figures
    (pp. vii-viii)
  5. Tables
    (pp. ix-x)
  6. Summary
    (pp. xi-xiv)
  7. Acknowledgments
    (pp. xv-xvi)
  8. Abbreviations
    (pp. xvii-xviii)
  9. CHAPTER ONE Introduction
    (pp. 1-6)

    More than 60 percent of nonelderly Americans receive health-insurance coverage through employers, either as policy holders or as dependents (Fronstin, 2006). However, rising health-care costs are leading many to question the long-term viability of the employer-based insurance system. Health spending per privately insured person increased by an average of 8.6 percent per year between 2000 and 2005, more than double the annual average increase in per-capita GDP (Ginsburg et al., 2006). The Agency for Healthcare Research and Quality (AHRQ) reports that the average total premium for an individual employer-sponsored insurance policy was $3,991 in 2005, an increase of 32.5 percent...

  10. CHAPTER TWO Data
    (pp. 7-10)

    Our data come from two BLS data files—the ECI and the EBS. The ECI measures quarterly changes in total wage and nonwage compensation per hour for U.S. employees and is fielded to a nationally representative sample of establishments. Compensation costs for wage and nonwage benefits are collected for a randomly selected group of occupations within each establishment. We converted the ECI data to an establishment-level file by computing average compensation costs per employee across all sampled occupations. We also converted the weights, originally designed to represent all U.S. employees, to establishment-level weights by summing the weights over all occupations...

  11. CHAPTER THREE Methods
    (pp. 11-14)

    We analyzed the ECI data using logistic regression, ordinary least squares (OLS) regression, and quantile regression. In contrast to OLS regression, which allows the researcher to predict the mean of the outcome holding other characteristics constant, quantile regression (Buchinsky, 1998; Koenker and Hallock, 2001) allows for predictions at various percentiles across the distribution, such as the median, the 25th percentile, and the 75th percentile. Understanding differences in health-insurance costs relative to payroll across the distribution is important both because means can be heavily influenced by outliers and because quantiles can shed light on dispersion of health insurance–cost ratios across...

  12. CHAPTER FOUR Results
    (pp. 15-34)

    Table 4.1 shows the means of key variables in the ECI for the years 2000, 2003, and 2005 for each firm-size category. For all firm-size categories, we see a slight downward trend in offer rates and a clear upward trend and health-insurance expenditures per employee. Our measure of burden, theHI/Pratio, also increased over time. The smallest firms have substantially lower offer rates and lower per-capita payroll than larger firms have. However, conditional on offering, health-insurance costs per employee are not necessarily lower at smaller firms. Annual payroll per employee, an approximation of the average annual salary within the...

  13. CHAPTER FIVE Limitations
    (pp. 35-36)

    Our analysis is limited to single-unit establishments, but costs and time trends at multiunit establishments may have been different. To the extent that multiunit establishments are larger, have dissimilar compensation, and have better risk-pooling and bargaining ability than single-unit establishments, we might expect that health-insurance burdens would be lower at these firms. Thus, our results may understate the true difference in health-insurance burdens between large and small firms. Another limitation is that, while the data allow us to observe information on both health-insurance costs and plan quality, we do not have data on enrollment at firms of different sizes. Previous...

  14. CHAPTER SIX Discussion
    (pp. 37-40)

    We find substantial increases over time in health-insurance burdens for all firm-size categories, holding constant other factors, such as unionization and industry mix. Health-insurance burdens increased at the mean and at the 25th, 50th, and 75th percentiles of the distribution. By 2005, a typical offering firm spent between 7 and 10 percent of payroll on health insurance, depending on size. Because these percentages exceed expenditure limits suggested in recent employer-mandate proposals (such as the 4-percent payroll tax discussed in the California governor’s plan), employer mandates could have the unintended consequence of reducing employer coverage. In some cases, firms may prefer...

  15. CHAPTER SEVEN Conclusion
    (pp. 41-42)

    Health-insurance burdens borne by employers grew substantially between 2000 and 2005, regardless of firm size. While growth was highest for small firms (<25 employees), medium-sized firms (25 to 99 employees) did not face higher growth in health-insurance burdens than larger employers did. Our results suggest that targeting health reforms that build on the employer-based system solely to large firms may be unwarranted, given that large firms do not face lower health-insurance burdens than moderately small firms do. However, our results indicate that the growth in health-insurance burdens has been substantial regardless of firm size. As a result, it seems that...

  16. APPENDIX Supporting Data
    (pp. 43-58)
  17. References
    (pp. 59-62)