Regulation and welfare: evidence from paragraph IV generic entry in the pharmaceutical industry

• Published date: 01 November 2016
• Date: 01 November 2016
• DOI: http://doi.org/10.1111/1756-2171.12157
• Author: Chirantan Chatterjee,Matthew J. Higgins,Lee Branstetter

RAND Journal of Economics

Vol.47, No. 4, Winter 2016

pp. 857–890

Regulation and welfare: evidence from

paragraph IV generic entry in the

pharmaceutical industry

Lee Branstetter∗

Chirantan Chatterjee∗∗

and

Matthew J. Higgins∗∗∗

This article estimates welfare effects of accelerated generic entry via Paragraph IV challenges.

Using data from 2000–2008 for hypertension drugs in the United States, we estimate demand

using a random-coefficients logit model. We find consumers gain $42 billion whereas producers

lose $32.5 billion from entry. This modest $9.5 billion gain in social welfare is consistent with

our observation that overall consumption does not increase after entry—generic sales displace

branded sales, shifting surplus downstream from producers to consumers, insurance companies,

∗Carnegie Mellon University & NBER; branstet@cmu.edu.

∗∗India Institute of Management, Bangalore; chirantan.chatterjee@iimb.ernet.in.

∗∗∗Georgia Institute of Technology & NBER; matt.higgins@scheller.gatech.edu.

We thank Tamer Abdelgawad, Serguey Braguinsky, Iain Cockburn, Marty Gaynor, David Greenstreet, Bart Hamilton,

Lowell Taylor, Jerry Thursby,Kenneth Train, Scott Stern, Dietmar Harhoff, Carolin Haeussler, Murray Aitken, Kensuke

Kubo, Vineet Kumar, Jian Ni, and Ellerie Weber, as well as seminar participants at Emory, Cornell, Carnegie Mellon,

IIM Bangalore, IIM Ahmedabad,Passau, Charles River Associates, Precision Health Economics, and the NBER Summer

Institute (Cambridge, 2011) for valuable comments and discussions. We thank Bhaven Sampat for helpful comments

and for sharing patent data. We are grateful to Judith Chevalierand two anonymous referees for detailed comments and

suggestions that significantly improved an earlier draft. We thank Susan Jack, Margaret Warner, and Robert Anderson

for guidance in using the National Health Interview Survey, and we thank Antara Dutta for sharing her code with us

as well as for extensive discussions. Programming assistance by Avadhoot Jathar, Nachiket Sahoo, Anubrata Banerjee,

and Trupti Natu is appreciated. We are grateful for the time Dr.Terry Simon and Dr. Ronald Severtis spent explaining

to us the treatment of hypertension. We also thank Murray Aitken and IMS Health Incorporated for their generous

support and access to their data. The statements, findings, conclusions, views, and opinions contained and expressed

herein are not necessarily those of IMS Health Incorporated or any of its affiliated or subsidiary entities. The statements,

findings, conclusions, views, and opinions contained and expressed in this article are based in part on data obtained

under license from the following IMS Health Incorporated or affiliate information service(s): IMS MidasTM,IMS

LifecycleTM, IMS National Disease and Therapeutic IndexTM, IMS National Prescription AuditTM , January 2000 to

December 2008, IMS Health Incorporated or its affiliates. All Rights Reserved. Higgins acknowledges funding from

The Imlay Professorship and Pfizer, Inc. Chatterjee and Branstetter acknowledge funding from NSF SCISIP grant no.

#0830233 and from the Young Faculty Research Chair at IIM Bangalore. Authors are listed alphabetically and the usual

disclaimers apply.

C2016, The RAND Corporation. 857

858 / THE RAND JOURNAL OF ECONOMICS

and retailers. We demonstrate significant cross-molecular substitution and discuss challenges in

determining what fraction of downstream surplus actually goes to consumers.

1. Introduction

The Drug Price Competition and Patent Term Restoration Act, informally known as the

“Hatch-Waxman Act,” was passed in 1984 and designed to balance access to pharmaceuti-

cal products while incentivizing pharmaceutical innovation in the United States. An important

provision in this legislation allows generic manufacturers to attempt to enter the market be-

fore patents protecting original branded products have expired. Using this mechanism, known

as a Paragraph IV (Para-IV) certification or challenge, generic manufacturers seek to enter

patent-protected markets either by claiming noninfringement or invalidity of the branded prod-

uct’s patent. Even though the Para-IV challenge has been available to generic manufacturers

since the passage of Hatch-Waxman, for reasons we will discuss, the number of challenges

did not grow significantly until the late 1990s. Using unique and novel data we quantify,

for the first time, the welfare effects of this accelerated generic entry as a result of Para-IV

challenges.

We focus on the hypertension market in the United States, which is relatively large in terms

of pharmaceutical revenues and disease prevalence. As a benchmark, we estimate the welfare

effects of generic entry using the nested multinomial-logit demand model popularized by Berry

(1994). However, given the restrictive assumptions and shortcomings associated with the nested-

logit model, we focus our analysis on the results obtained from a full random-coefficients logit

model (Nevo, 2000a, 2000b). Estimation of this model on quarterly data for the 2000–2008 period

suggests consumer welfare gains from Para-IV entry of $42 billion. These are significant gains,

but they are largely offset by declines in producer surplus. Our estimates suggest that Para-IV

entry reduces producer surplus by $32.5 billion, leading to net social gains of approximately $9.5

billion. Importantly, the net social gains do not arise from an expansion in the total quantity of

drug consumption as prices fall—for reasons we will discuss later, even large price declines do

not appear to bring large numbers of inframarginal consumers into the US market for prescription

drugs. Instead, the net social gains appear to be driven, in part, byan expansion in product variety

associated with Para-IV entry. As we will see, branded producers often respond to Para-IV entry

by launching reformulations of existing products that offer additional benefits to at least some

patients; this does not appear to expand the total quantity of drug consumption but expands variety

in a way that generates additional consumer surplus.

Our use of the term “consumer surplus” suggests that the benefits of entry and lower prices

accrue directly to consumers, but we acknowledge in this article the reality that consumers

interact with drug manufacturers through a complicated chain of intermediaries that include

prescribing physicians, drug retailers, and insurance companies. It is really this nexus of parties

that chooses drugs and collectively appropriates the gains from lowerdrug prices. It may therefore

be more accurate to speak of “downstream” surplus rather than consumer surplus, though we

use the latter term to avoid confusion and maintain continuity with the literature. Determining

exactly what component of the aggregate downstream surplus actually goes to consumers is not

possible, even with the rich data at our disposal. We return to these issues later in the article

and describe future work that might help us disaggregate the downstream surplus into its various

components.

In addition to welfare gains, we also document a high degree of cross-molecular substitution

in this market. Cross-molecular substitution occurs when patients shift their hypertension drug

consumption from a branded product to the generic version of a different branded product, based

on a different molecule. Wepresent anecdotal evidence suggesting that insurance companies en-

courage this shift among their customers and show that the scope for cross-molecular substitution

C

The RAND Corporation 2016.

BRANSTETTER, CHATTERJEEAND HIGGINS / 859

in hypertension appears to be substantial.1The implications of cross-molecular substitution are

profound; a branded product’s intellectual property protection, within a market, is only as strong

as their weakest patent of its branded competitors. A high degree of cross-molecular substitution

thus amplifies the positive impact of generic entry on consumer welfare and the negative impact

of entry on producer profits.

Hatch-Waxman was originally designed to balance access to cheaper drugs on the one hand

while preserving appropriate incentives for innovationon the other. We demonstrate in this article

that Para-IV challenges have been an effective mechanism in terms of providing accelerated

access to inexpensive generic drugs. Has this rising generic entry had an impact on the rate

and direction of drug research and development activity? A full consideration of that question

is beyond the scope of this article. However, ongoing research by the authors seeks to examine

and quantify this effect (Branstetter, Chatterjee, and Higgins, 2014). Preliminary results suggest

that the rising generic entry in US markets has significantly reshaped the nature of global drug

development activity.

The article proceeds as follows. Section 2 offers a brief discussion of the regulatory environ-

ment in which pharmaceutical firms operate in the United States. Section 3 discusses the relevant

prior literature. Our data and methodology are presented in Sections 4 and 5. Section 6 presents

results, and we discuss the implications of our work and conclude in Section 7.

2. Regulatory environment and early generic entry

Hatch-Waxman and paragraph IV challenges. The cur rent regulatory environment

faced by pharmaceutical companies in the United States can be traced to the passage of the

Hatch-Waxman Act in 1984. This legislation expedites Food and Drug Administration (FDA)

approval for generic entry while extending the life of pharmaceutical patents in order to com-

pensate innovators who lost time on their “patent clocks” waiting for FDA approval (Grabowski,

2007). This balance was deemed necessary to equalize two conflicting policy objectives:

giving pharmaceutical firms incentives to conduct drug research while improving consumer

welfare by enabling generic firms to quickly bring copies to market (Federal Trade Commission

(FTC), 2002).

When a pharmaceutical company submits a New Drug Application (NDA) to the FDA for

approval they are required, by law, to identify all relevant patented technologies necessary to

create the drug; these patents are subsequently listed in the FDA OrangeBook.2Upon approval of

a drug, the FDA will restore patent term to the pharmaceutical firm for time used by the FDA in

the approval process (Grabowski,2007).3In addition, the FDA will also grant each new approved

product regulatory protection lasting for five years (“data exclusivity”)that r uns concurrently with

patent protection.4During this data exclusivity period, regardless of the status of the underlying

patent(s), no generic entry may occur. At the conclusion of data exclusivity, only patents protect

branded products. The period running from the cessation of data exclusivity to the expiration of

a drug’s patents is commonly referred to as “market exclusivity” (see Figure 1).

Prior to the passage of Hatch-Waxman, generic manufacturers seeking to sell their products

in the US market had to demonstrate the safety and efficacy of their products by putting them

through clinical trials. Although the outcome of these trials lacked the uncertainty involved in the

1See Aitken, Berndt, and Cutler (2009), who also draw attention to this phenomenon.

2During our sample period, there was no regulatory pathwaythrough which generic producers could certify their

products as “bioequivalent” to existing biologic drugs. The AffordableCare Act created the legal basis for entry, but the

FDA did not approvesuch a product until March 2015. We thank an anonymous referee for urging us to clarify this point.

3There are limits to this. Pharmaceutical firms cannot receive a patent extension of more than five years, nor are

they entitled to patent extensions that give them effectivepatent life (post approval) of greater than 14 years.

4There are exceptions, for example, orphan drugs receive seven years and reformulations receive three years of

data exclusivity.A pediatric indication can receive an additional six months of data exclusivity. With the recent passage

of the GAIN Act, certain antibiotics are eligible for five additional years of data exclusivity.

C

The RAND Corporation 2016.