Quality information and consumer health plan choices

Abstract

This paper presents an empirical analysis of the effects of providing information about plan quality on consumers’ health plan choices in a private employment setting. Analysis of plan switching behavior suggests that the provision of quality information had a small, but significant effect on consumer plan choices. Employees were more likely to switch from plans with lower reported quality. Cross-sectional analyses of plan choice indicate that reported quality played a role in plan selection even after controlling for other health plan characteristics commonly associated with plan choice. The age of the policyholder and the type of policy purchased moderated the effects of plan characteristics on plan choice in ways that may be consequential for adverse selection.

Introduction

In the United States, most consumers under the age of 65 obtain health insurance through an employer. Many, though not all of these, consumers have the opportunity to choose between two or more health plans. Recent surveys indicate that the percentage of private-sector employees with a choice of health plans has grown and that large employers are more likely to offer a set of health plans to their employees (Jensen et al., 1997). The federal government and most state governments offer current employees and retirees a choice of health plan. In addition, 64% of Medicare beneficiaries in 1998 had a choice between two or more Medicare Managed Care plans (Schneider et al., 2000).

Increasingly, consumers have available information on health plan performance to facilitate their choice. Employers frequently gather or purchase health plan performance data to distribute to their employees during the health plan selection period. The federal government now collects and distributes health plan quality data (HEDIS measures and patient reports of quality through the CAHPS survey) on Medicare managed care plans to Medicare beneficiaries. The Consumers Checkbook publishes a guide to health insurance plans for federal employees that lists enrollee plan ratings on multiple dimensions of quality of care and customer service (Washington Consumers’ Checkbook, 1997). The National Committee on Quality Assurance (NCQA) posts its accreditation of managed care plans on its world wide web site (www.ncqa.org) and markets a health plan performance database (Quality Compass) to private employers and other institutions. Furthermore, many national (US News and World Report, 1996; Consumer Reports, 1996; Newsweek, 1996) and local publications report performance data on health care providers and insurers.

Despite the availability of this information, there have been few published studies of the effects of health plan performance data on consumers’ health plan choices. The existing literature on this topic does not provide convincing evidence that health plan performance data affect consumers’ health plan choices (Schneider and Epstein, 1998, Chernew and Scanlon, 1998; see Mukamel and Mushlin, 1998 and Scanlon et al., 2001 for a different set of findings). These findings are troublesome for practitioners and policymakers implementing managed competition. Embedded in the theory of managed competition is an assumption that health plan quality can be measured and that all else equal consumers will enroll in plans with higher measured quality. If these assumptions hold and if consumers have a choice of health plan, then market forces will generate incentives for health plans to improve the quality of care. The managed competition model has been implemented in several settings; however, the underlying assumptions relating to the effects of health plan quality information have not been rigorously tested using data on individual health plan choices.

This paper presents a case study analysis of the effects of providing health plan performance data on consumers’ enrollment decisions. The analyses are based on data from a natural experiment in which a large employer began providing health plan performance data to its employees to facilitate their health plan selections. A rich panel data set facilitates analysis of plan switching behavior, the relative importance of different health plan characteristics in consumers’ choices, and differences in health plan choices attributable to employee characteristics.

There exist a variety of analytical techniques to empirically study the determinants of consumer health plan choices. Theoretically these models are straightforward to apply; in practice the researcher encounters numerous problems with unobserved characteristics of the consumer and the alternatives. Unfortunately, different empirical methods frequently yield different results. In this paper, I employ two different empirical methodologies to illustrate this problem and to triangulate on some interesting results. Analysis of plan switching behavior suggests that the provision of quality information had a small, but significant effect on consumer plan choices. Employees were more likely to switch from plans with lower reported quality. Cross-sectional analyses of plan choice indicate that reported quality played a role in plan selection even after controlling for other health plan characteristics frequently associated with plan choice. The age of the policyholder and the type of policy purchased moderated the effects of plan characteristics on plan choice in interesting and intuitive ways.

The paper is organized as follows. Section 2 outlines the choice model that forms the basis for the econometric analyses. Section 3 reviews previously published empirical studies of consumers’ health plan choices. Section 4 gives a brief history of the health insurance benefit at Harvard University and describes the data. Section 5 presents the analytic results. Section 6 concludes with a discussion of the implications of the findings.