Introduction Directory UMM :Journals:Journal of Health Economics:Vol19.Issue1.Jan2000:

Ž . Journal of Health Economics 19 2000 61–91 www.elsevier.nlrlocatereconbase Medicaid and crowding out of private insurance: a re-examination using firm level data Lara Shore-Sheppard a , Thomas C. Buchmueller b, , Gail A. Jensen c a Department of Economics, UniÕersity of Pittsburgh, Pittsburgh, PA, USA b Graduate School of Management, UniÕersity of California, IrÕine, CA and National Bureau of Economic Research, USA c Institute of Gerontology and Department of Economics, Wayne State UniÕersity, Wayne, NE, USA Abstract While previous research has identified a relationship between expanded Medicaid eligibility and falling private health insurance coverage, the exact mechanism by which this ‘‘crowding out’’ occurs is largely unexplained. We combine individual and firm-level data to investigate possible responses to the Medicaid expansions by firms and workers. We find no evidence that the expansions affected employer offers of insurance to workers. However, we find some evidence of an effect on the probability that a firm offers family coverage, and on the percentage of full-time workers accepting employer-sponsored coverage offered to them. q 2000 Elsevier Science B.V. All rights reserved. JEL classification: G22; I18; I38 Keywords: Medicaid; Private insurance; Crowd-out

1. Introduction

In recent years, private health insurance coverage levels have declined, leading to considerable speculation as to the cause of this decline. To date, however, little Corresponding author. Graduate School of Management, University of California, Irvine, CA 92697-3125, USA; Tel.: q1-949-824-5247; fax: q1-949-824-8469; E-mail: tcbuchmuuci.edu 0167-6296r00r - see front matter q 2000 Elsevier Science B.V. All rights reserved. Ž . PII: S 0 1 6 7 - 6 2 9 6 9 9 0 0 0 1 9 - 3 consensus has been reached. One trend that has been noted is that the decline in private coverage occurred contemporaneously with significant increases in eligibil- ity for Medicaid, the public insurance program for the poor. Starting in the late 1980s a series of laws dramatically expanded Medicaid eligibility, particularly among children of the working poor. More recently, federal legislation was passed to provide states with US24 billion over 5 years to provide health insurance coverage for additional poor children, the largest increase in public health insur- ance for children in three decades. As concern about uninsured children continues to spur public action, a better understanding of the impact of expanding public health insurance is critical for designing future policy. An important question is whether expanding public insurance has the effect of reducing private insurance coverage. While movement from private to publicly-provided coverage is not necessarily a policy failure, to the extent that such movement takes place, the overall rate of health insurance coverage will not increase as much as expected. Several recent studies have addressed the question of whether the recent Ž Medicaid expansions ‘‘crowded out’’ private insurance Cutler and Gruber, 1996; Dubay and Kenney, 1997; Shore-Sheppard, 1996, 1997; Yazici and Kaestner, . 1998; Blumberg et al., 1999 . While the estimated magnitudes of crowding out vary greatly, all of the studies find that at least some crowd-out occurred. However, an important piece of the puzzle remains unexplained: the mechanism by which crowding out may have occurred. Declines in private insurance among low income workers may have resulted from actions taken by firms, such as reductions in offers of insurance or increases in cost sharing that encourage employees with eligible family members to drop employer-sponsored coverage. Another possibility is that when public insurance becomes available, employees may drop employer-sponsored coverage even in the absence of action by employ- ers. The mechanism by which movement from private to public coverage occurs has implications for the welfare impacts of the expansions. For example, if crowding out occurs as a result of firm actions, other employees not eligible for the expansions may be affected. Unfortunately, the household survey data sets used in previous studies provide very limited information on health insurance and therefore provide little insight on the roles played by firms and workers in movements from private to public coverage. Ž . In this paper, we combine Current Population Survey CPS data on Medicaid eligibility with data from several national establishment surveys to investigate in greater detail the response of employers and employees to the Medicaid expan- sions. We attempt to answer several questions. First, to what extent did firms Ž employing a large percentage of newly Medicaid-eligible workers or workers . with newly eligible family members stop offering health insurance to their employees? Second, is there any evidence that affected employers continued to offer insurance, but in some way reduced the generosity of their benefits to induce Medicaid-eligible employees to drop private coverage? Potential responses of this type include dropping coverage for dependents, or increasing the amount that employees must contribute directly for coverage. Third, did the Medicaid expan- sions affect the percentage of workers accepting offers of employer-sponsored insurance?

2. Background and previous literature