Industry Shock Expectations, Interindustry Linkages, and Merger Waves: Evidence from the Hospital Industry

• Author: Robert Town,Minjung Park
• DOI: http://doi.org/10.1111/jems.12067
• Published date: 01 September 2014
• Date: 01 September 2014

Industry Shock Expectations, Interindustry

Linkages, and Merger Waves: Evidence from the

Hospital Industry

MINJUNG PARK

Haas School of Business

University of California

Berkeley, CA, USA

minjungp@gmail.com

ROBERT TOWN

The Wharton School

University of Pennsylvania

Philadelphia, PA,USA

rjxtown@gmail.com

It is well known that mergers often occur in waves, and this paper develops a new mechanism for

merger waves: expectations over industry shocks. We develop a simple test of this explanation and

use it to explore the role of expectations in the context of the 1990s hospital merger wave. Managed

care such as Health Maintenance Organizations (HMOs) started to become popular in the late

1980s and ultimately became an important player in the health insurance market. Our empirical

analysis shows that the expected increase in the popularity of HMOs was partly responsible for

the hospital merger wave of the 1990s: hospitals feared that the “innovation” of managed care in

the downstream insurance market would penetrate the upstream hospital market and responded

to this belief by merging. Our results show the importance of incorporating expectations and

interindustry linkages into the understanding of merger waves.

1. Introduction

It is well known that mergers often occur in waves and during these waves large amounts

of capital are reallocated across the economy (e.g., Town, 1992; Holmstrom and Kaplan,

2001). Many explanations for merger waves have been offered in the literature. One

of the leading hypotheses is that merger waves are a response to industry shocks.

Changes in economic environments such as deregulation, input costs movements, or

technological developments could necessitate a large-scale reallocation of assets in an

industry, resulting in merger waves clustered by industry (Stigler, 1950; Gort, 1969;

Mitchell and Mulherin, 1996; Andrade et al., 2001). Others emphasize the role of capital

liquidity and the relaxation of financing constraints (Harford,2005; Eisfeldt and Rampini,

2006). For instance, Harford (2005) argues that both industry-level shocks and sufficient

capital liquidity are required for a merger wave to occur. Yet others posit that merger

waves might occur when managers take advantage of temporary market misvaluation

(Rhodes-Kropf and Viswanathan, 2003; Shleifer and Vishny, 2003).

The current consensus is that all of these theories have some relevance for explain-

ing merger waves—different theories apply to different periods and differentindustries

Wethank Martin Gaynor, an anonymous reviewer, and thecoeditor for their insightful comments. All remain-

ing errors are our own.

C2014 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume23, Number 3, Fall 2014, 548–567

Merger Waves in the Hospital Industry 549

(Martynova and Renneboog, 2008). Yet, despite this belief, there is no broad agreement

on the correct theory to apply to a specific merger wave context. For example, several

papers have argued that the merger wave of the late 1990s was a consequence of stock

market misvaluation (Rhodes-Kropf and Viswanathan, 2003; Shleifer and Vishny, 2003)

whereas Holmstrom and Kaplan (2001) contend that it was driven, in large part, by

deregulation. Stigler (1950) famously characterized the turn of the 20th century merger

wave as motivated by monopoly objectives in response to the Supreme Court ruling on

the illegality of price fixing. Jovanovic and Rousseau (2008) emphasize the causal role of

new, general purpose technologies in merger waves. According to their view, the 1900s

wave was caused by the diffusion of electricity and the 1990s wave was a consequence

of improvements in information technology. If economists cannot agree on the causal

factors underlying a given merger wave, it suggests that our understanding of merger

waves is still incomplete.

The empirical evidence used to test these theories generally emanates from popular

press and trade publication assessments combined with across industry correlations

between some hypothesized factor (which is generally proxied by an industry-level

variable) and merger activity. Although the use of cross-industry variations is a useful

first step toward understanding merger waves, this approach suffers from standard

threats to causal inference. In particular, such an approach does not control for omitted

industry characteristics that are correlated with both the hypothesized factor and merger

activity.1

In this paper, we study the large, transformative hospital merger and acquisi-

tions wave of the 1990s. The mean, population weighted, Herfindahl–Hirschman index

(HHI) in Health Services Areas (HSAs)—a common definition of market in the hospital

industry—increased from 1,975 to 2,694 between 1990 and 2000. This period overlaps

with the rise of managed care health insurance and the policy debate over the ultimately

unsuccessful Clinton health reform.2These two events are widely credited as igniting

the storm of hospital consolidation.

Prior to the rise of managed care, the dominant form of health insurance was

indemnity insurance. In an indemnity plan, the enrollee can see any provider and the

insurer generally pays “usual and customary” rates. The rise of Health Maintenance

Organizations (HMOs; a strong form of managed care) significantly altered the market

environment for hospitals. For a typical hospital, approximately half of its patients

are covered by private health insurance and managed care may reduce those patients’

demand for inpatient services by tightly controlling health care utilization. Such negative

demand shocks can prompt mergers(e.g., Qiu and Zhou, 2007). More importantly, HMOs

have greater bargaining power with hospitals than indemnity insurers because HMOs

can direct a large pool of patients to (or away from) hospitals through the selective

contracting mechanism (Altman et al., 2003).3Thus, merger could be a logical strategic

response of hospitals when faced with decreased bargaining leverage against HMOs

(Gowrisankaran et al., 2013).

Our goal in this paper is to empirically investigate whether the HMO “revolution”

in the downstream health insurance market prompted the hospital merger wave of

the 1990s in the upstream market. The merger waves literature generally ignores the

1. The weaknesses of interindustry studies are well known in the industrial organization literature (see

Schmalensee, 1989).

2. The promotion of HMOs was central to the Clinton health reform proposal.

3. Health insurers construct networks of providers that their enrollees can access. Selective contracting

refers to the practice of including a subset of the available providers in the health insurer’s network.