QUALITY CHOICE AND MARKET STRUCTURE: A DYNAMIC ANALYSIS OF NURSING HOME OLIGOPOLIES

• Date: 01 November 2015
• DOI: http://doi.org/10.1111/iere.12137
• Author: Haizhen Lin
• Published date: 01 November 2015

INTERNATIONAL ECONOMIC REVIEW

Vol. 56, No. 4, November 2015

QUALITY CHOICE AND MARKET STRUCTURE: A DYNAMIC ANALYSIS

OF NURSING HOME OLIGOPOLIES∗

BYHAIZHEN LIN 1

Indiana University, U.S.A.

This article develops a dynamic model of entry and exit to analyze quality choice and oligopoly market structure in

the nursing home industry. I find significant heterogeneity in the competitive effects across market structures: Firms of

similar quality levels compete more strongly than dissimilar firms. Sunken entry costs are extremely large, and quality

adjustment behavior is governed by significant fixed adjustment costs. A proposal to eliminate low-quality nursing

homes is found to cause a large supply-side shortage, and another proposal to lower entry costs has offered a perverse

incentive to provide low quality of care.

1. INTRODUCTION

Poor quality of patient care plagues the nursing home industry. Statistics from the U.S.

General Accounting Office (GAO) show that more than 25% of nursing facilities nationwide

have serious quality problems that could either harm residents or place them at the risk of

death (GAO, 1999). Recognizing this concern, federal and state governments have increased

their scrutiny of nursing home operations and have initiated stringent regulations as a result.

For example, a nursing facility has to pass strict licensing requirements to be certified to accept

Medicare and Medicaid patients. In order to retain this certification, a facility must be surveyed

annually to verify its compliance with all federal and state regulatory requirements. If a facility

fails to meet these requirements, it may be penalized, denied payment for patients, and even

subjected to immediate termination of operation. Increased enforcement has also resulted in

more frequent issuance of civil money penalties for instances of noncompliance (OIG, 2005).

Despite these tremendous efforts, poor quality persists within the industry. Quality of care has

been the second most common violation of federal regulations on nursing homes; between 2000

and 2006, the percentage of U.S. nursing homes that received violations increased from 22% to

30% (Harrington et al., 2007).

The persisting quality issue has prompted the following research questions for this study:

First, how do firms make product-type decisions? To be more specific, how do nursing homes

choose their quality levels? Understanding a firm’s choice of product type is crucial to examining

differences in market structure, which have profound implications for regulations on market

power, entry, and quality. Second, what happens if low-quality nursing facilities are eliminated?

Heterogeneity in product type shields firms from competing directly with each other. Markets

with one homogeneous product type may not support as many firms as those with more than one

product type. If the market shrinks by a considerable amount, a large welfare loss will ensue, as

∗Manuscript received July 2012; revised November 2013.

1I am deeply indebted to Marc Rysman for his guidance and support. I have also greatly benefited from conversations

with Victor Aguirregabiria, Allan Collard-Wexler, Randy Ellis, and Ginger Jin. I thank the editor, three anonymous

referees, Calixte Ahokpossi, Jacob Glazer, Chun-Yu Ho, Kevin Lang, Ching-To Albert Ma, Michael Manove, Jeffrey

Prince, Kosali Simon, and seminar participants at Boston University, Drexel University, University of Rochester,

University of Toronto, and the 6th International Industrial Organization Conference for their helpful suggestions and

comments. All remaining errors are my own. Please address correspondence to: Haizhen Lin, Department of Business

Economics and Public Policy, Kelley School of Business, Indiana University, 1309 E. 10th Street, Bloomington, IN

47405. Phone: 812-855-3535. Fax: 812-855-3354. E-mail: hzlin@indiana.edu.

1261

C

(2015) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social

and Economic Research Association

1262 LIN

(i) patients may not find nursing homes nearby2and (ii) a large number of patients may not get

access to or will have to pay a high price for nursing home care. In this regard, I am interested

in how many high-quality nursing facilities will remain in the market if low-quality facilities are

not allowed to operate and how long it will take for the industry structure to reach a new steady

state. Third, will an increase in competition lead to improved quality of care? If competition is

associated with quality of care, then mechanisms for enhancing competition will have a positive

impact. Otherwise, it is unlikely that increased competition will achieve its intended influence

on the providers’ quality decisions.

To answer the above questions, I develop a dynamic model of entry, exit, and quality choice

using the Markov-perfect equilibrium (MPE) framework provided by Maskin and Tirole (1988)

and Ericson and Pakes (1995). The model incorporates two features regarding quality decisions:

strategic interaction and intertemporal dependence. Both features are especially important if

one is interested in examining how different policies would affect firms’ quality choice in an

oligopoly setting.3The model also allows for different cost structures and flexible competi-

tion patterns between low- and high-quality firms. Specifically, I allow the entry and quality

adjustment costs to differ depending on a nursing home’s quality choice. I also examine the

competitive effects within the same quality type and across different quality types. Finally,

the model employs a two-stage fixed effects approach to control for unobserved heterogeneity

across markets. This approach allows me to assign markets to different groups according to their

estimated profitability levels. I then add a pair of group-specific dummies (fixed low-effect and

fixed high-effect) to the per-period payoff function as controls for unobserved heterogeneity in

profitability levels for either quality type across markets. Failing to correct for unobserved het-

erogeneity in profitability across markets and across different quality types may lead to biased

estimation.

The model is estimated for quality choices in the U.S. nursing home industry using a hybrid

two-step estimator as presented by Aguirregabiria and Mira (2007) and Pakes et al. (2007).

I use nurse staffing levels as a measure of quality, with more details covered in Subsection

3.2. The main results are summarized as follows: First, a great variation exists in regard to

profitability levels across markets and across different quality types. Second, there is significant

heterogeneity in the competitive effects across market structures; firms of similar quality levels

compete more strongly than dissimilar firms. Third, it is more profitable to provide low quality

of care in larger markets. Finally, entry costs are large, and quality adjustment behavior is

governed by significant fixed adjustment costs. For example, in order to switch to high quality,

a low-type firm must pay an adjustment cost of roughly 2.5 times its monopoly profits.

Based on the estimation of the structural model, I study three counterfactuals. The first

predicts that the overall quality of care deteriorates given a positive growth in the elderly

population. The next two examine the impact of two policy proposals that intend to improve

quality of care in the nursing home industry. The proposal to eliminate low-quality nursing

homes causes a large supply shortage: The total number of nursing homes decreases by more

than 25%, leaving about 65% of the markets either without a nursing home or dominated by

monopolies. The proposal to lower entry costs offers a perverse incentive to provide low quality

of care. This outcome is driven by increased competition, which narrows the gap in pay rates

between private-pay and Medicaid patients and therefore dilutes the profits for providing high

quality of care.

This article contributes to the existing literature in several ways. First, it is part of the literature

on market structure and endogenous product-type decisions. Different from previous studies

such as Mazzeo (2002) and Seim (2006), this article takes a dynamic approach and allows firms

to adjust their quality levels in each period. To the extent that this article has quantified the fixed

costs of adjustment in a dynamic setting, it is related to papers such as Collard-Wexler (2013)

2Previous literature, such as Mehta (2006), has found evidence of patients’ inclinations to stay in nursing homes close

to their homes.

3Given that entry and quality adjustment costs play an important role in this industry, ignoring either feature might

lead researchers to false conclusions regarding policy impact on entry, quality, and industry dynamics.