Paper to be presented at the

35th DRUID Celebration Conference 2013, Barcelona, Spain, June 17-19

THE IMPACT OF PATENT WARS ON FIRM STRATEGY: EVIDENCE FROM

THE GLOBAL SMARTPHONE MARKETYongwook Paik

University of Southern CaliforniaStrategy

yongwook.paik@marshall.usc.edu

Feng ZhuUniversity of Southern California

Strategyfzhu@marshall.usc.edu

AbstractWe examine how patent wars affect firm strategy. We hypothesize that, as patent wars intensify, firms reduce the shareof their business in markets with strong intellectual property (IP) protections and increase the share of their business inmarkets with weak IP protections due to increased litigation risks. This effect is attenuated for firms with strongertechnological capabilities and is more pronounced for firms from countries with weak IP systems. Using data from theglobal smartphone market, we find strong support for our theoretical argument. This study sheds new light on theemerging patent enforcement strategy literature by highlighting the significance of heterogeneity in the efficacy ofnational patent systems in explaining firm strategy in platform-based markets.

Jelcodes:L10,L19

The Impact of Patent Wars on Firm Strategy:

Evidence from the Global Smartphone Market*

Yongwook Paik

Marshall School of Business University of Southern California

Los Angeles, CA 90089-0808 Phone: 213-740-7646

Email: Yongwook.Paik@marshall.usc.edu

Feng Zhu Marshall School of Business

University of Southern California Los Angeles, CA 90089-0808

Phone: 213-740-8469 Email: fzhu@marshall.usc.edu

May 24, 2013

* We thank Rahul Kapoor, Brian Wu, PK Toh, Richard Wang, Deepak Hegde, Gabriel Natividad, Juan Alcacer, Jeffrey York, Donghyung Shin and seminar participants at the Darden & Cambridge Entrepreneurship and Innovation Research Conference, Atlanta Competitive Advantage Conference, and the USC Organization and Strategy Workshop for valuable comments and suggestions. The authors acknowledge excellent research assistance from Yongzhi (Alex) Wang and Heejin Woo.

1

The Impact of Patent Wars on Firm Strategy:

Evidence from the Global Smartphone Market

Abstract

We examine how patent wars affect firm strategy. We hypothesize that, as patent wars

intensify, firms reduce the share of their business in markets with strong intellectual

property (IP) protections and increase the share of their business in markets with weak IP

protections due to increased litigation risks. This effect is attenuated for firms with

stronger technological capabilities and is more pronounced for firms from countries with

weak IP systems. Using data from the global smartphone market, we find strong support

for our theoretical argument. This study sheds new light on the emerging patent

enforcement strategy literature by highlighting the significance of heterogeneity in the

efficacy of national patent systems in explaining firm strategy in platform-based markets.

2

INTRODUCTION

Patents and patent enforcement strategies have become an essential part of a firm’s competitive

strategy in today’s world. Firms strategically use patents as isolating mechanisms to protect their

intellectual property or as defense mechanisms to help them obtain access to external innovations through

mutual hold-up (Somaya, 2003). For example, the smartphone industry has spent an estimated $20 billion

on patent litigation and patent purchases since 2010, and the number of district court patent filings has

tripled in the last 20 years (Cohan, 2012). In 2011, for the first time, spending by Apple and Google on

patent lawsuits and unusually big-dollar patent purchases exceeded their spending on research and

development of new products (Duhigg and Lohr, 2012).

While patents, and intellectual property issues in general, have been studied extensively by legal

scholars and economists (Bessen and Meurer, 2008b; Lanjouw and Schankerman, 2001; Merges et al.,

1997)1, only recently have strategy scholars started to pay more attention to patents as an element of firm

strategy (Somaya, 2012). For example, recent studies have shown that patents have deterrence effects on

rivals (Agarwal, Ganco, and Ziedonis, 2009; Clarkson and Toh, 2010), that settlements and specialized

tribunals can function as patent enforcement strategies as an alternative to costly litigation (Somaya, 2003;

Somaya and McDaniel, 2012), and that aggressive patenting strategies can be used to avoid being "fenced

in" by owners of complementary technologies (Ziedonis, 2004). These recent studies have led strategy

scholars to examine how patent enforcement strategies are used by firms in dealing with direct rivals.

Less attention, however, has been paid to how patent litigation affects market participants who

are not themselves directly involved in the litigation. This question has become increasingly important, as

an increasing number of today’s markets have evolved into highly interconnected networks of

organizations. In such networks, firms co-evolve around shared sets of knowledge or technologies,

working both cooperatively and competitively to develop products and services (Iansiti and Levien, 2004a;

1 The main debate among legal scholars and economists focuses on whether the current intellectual property system, especially the patent system, promotes innovation and dissemination of new knowledge. For a review of this literature, see, for example, Nordhaus (1969), Gilbert and Shapiro (1990), Klemperer (1990), Merges and Nelson (1990), Scotchmer (1991), Lemley (1997), Sakakibara and Branstetter (2001), Moser (2005), etc.

3

Moore, 1996). Firm strategies are often interdependent when they share such sets of technologies (Adner,

2006, 2012; Adner and Kapoor, 2010). It is thus natural to expect that the impact of patent litigation will

propagate to interdependent market participants, even if such parties are not directly involved in litigation.

Consequently, patent litigation may affect the overall performance of a broader network of firms.

Moreover, extant studies often use data from the US market only. But, since patent litigation, as a

value appropriation mechanism, are effective only when supported by strong institutional environments,

given the increasingly global and significant levels of heterogeneity of institutional environments across

different countries and businesses today, we need a better understanding of how firms react to potential

patent litigation on the global scale (Somaya, 2012).

In this paper, we investigate the effect of escalated patent litigation - the so-called ‘patent wars’ -

on firm strategy using data from the global smartphone market. We draw on the law, economics, and

strategy literatures on patent enforcement (Agarwal et al., 2009; Lanjouw and Lerner, 1997; Shaver, 2012;

Somaya, 2003; Somaya and McDaniel, 2012) to form testable hypotheses. Echoing some recent studies

using cross-country variation in IP systems (Delgado, Kyle, and McGahan, 2013; Kyle and McGahan,

2012), we show that the efficacy of patent enforcement systems across markets plays a significant role in

firm strategy during patent wars and ultimately shapes the global competitive landscape,. In our empirical

context, as patent wars intensify, smartphone vendors2, even those not directly involved in patent

litigation, gradually shift their business foci to markets with weaker intellectual property (IP) rights

protection. This shift, however, is attenuated for vendors with stronger technological capabilities and is

more pronounced for vendors whose home markets have weak IP systems. Together, these changes shape

the competitive landscape for platform competition.

This paper contributes to the emerging patent enforcement strategy literature (Agarwal et al.,

2009; Somaya, 2003; Somaya, 2012; Somaya and McDaniel, 2012; Ziedonis, 2004) that focuses on how

firms leverage their patents as business assets for competitive advantage. The extant literature suggests

2 Throughout this paper, we use the term “vendor” to refer to firms that manufacture and sell smartphone mobile devices.

4

that a firm’s patent strategy does not work in vacuum, but must take the institutional environment into

account, especially the effectiveness of the patent enforcement system, which is an important topic but

has yet to be considered in the literature (Somaya, 2012). Our paper is the first study to examine the

relationship between the effectiveness of the patent enforcement system and firm strategy in an

international context as well as the consequences of this relationship. We investigate this important

relationship in an international context because the effectiveness of the patent enforcement system does

not change over a short period of time but rather displays ample variation across countries. We show that

patent litigation is prevalent and effective as a value appropriation strategy in markets with strong IP

protections, which leads competitors to shift their business focus to markets with weak IP protection.

THEORETICAL FRAMEWORK

Patents are an exclusionary form of government-granted monopoly rights for a limited period of

time, usually 20 years from the filing date (Lerner, 2002). The patent system is predicated on the theory

that inventors will be incentivized to innovate and disclose new knowledge to the public in exchange for a

limited period of exclusivity in commercializing their inventions (Mazzoleni and Nelson, 1998). With

globalization, and in spite of efforts such as TRIPS (Trade Related Aspects of Intellectual Property

Rights)3 to standardize IP protection internationally, national patent systems still operate largely

independently and exhibit significant differences across countries (Kyle and McGahan, 2012; Park, 2008;

Somaya, 2000). These differences include coverage, membership in international treaties, restrictions on

patent rights, enforcement (e.g., availability of preliminary injunctions), and duration of protection

(Ginarte and Park, 1997). Accordingly, the efficacy of patent enforcement as an appropriability

mechanism varies substantially across countries. Hence, firms strategically choose to rely more on patent

litigation in markets with strong IP regimes, while they choose to rely on other means, such as secrecy,

3 TRIPS is an international agreement, administered by the World Trade Organization (WTO), that sets down minimum standards for many forms of intellectual property (IP) regulation, as applied to nationals of other WTO members. It was negotiated at the end of the Uruguay Round of the General Agreement on Tariffs and Trade (GATT) in 1994.

5

lead time advantages and the use of complementary marketing and manufacturing capabilities in weak IP

regimes (Cohen, Nelson, and Walsh, 2000; Levin et al., 1987).

The effectiveness of patents as an appropriability mechanism varies substantially across

industries as well. Traditionally, patents were relatively well-defined and effective in "discrete" product

industries, as with chemical or pharmaceutical formulas (Grabowski and Vernon, 1982; Mansfield,

Schwartz, and Wagner, 1981), but were quite porous, imperfect, and unclear in other industries (Cohen et

al., 2000; Lemley and Shapiro, 2005). For example, systems product industries, such as the computer

industry, make products that include numerous patented inventions, which may in turn be owned by many

different firms, creating a "thicket" of patents. Thus, as Somaya (2003) notes, firms in such industries

devise cooperative mechanisms, such as patent pooling (Bittlingmayer, 1988; Joshi and Nerkar, 2011;

Merges, 1996), cross-licensing (Grindley and Teece, 1997), and even tacit truces on patent litigation

between firms (Von Hippel, 1988), to obtain access to each other’s patents and thereby retain their

abilities to commercialize products. While these industries have historically tended to rely more heavily

on lead time, secrecy, and manufacturing or design capabilities than patents to recoup their R&D

investments (Hall and Ziedonis, 2001), in recent years, they have increasingly come to rely more on the

strategic management – and enforcement - of their patents.

But the recent explosion in patent enforcement litigation in these industries (Meurer and Bessen,

2005) are not an entirely new phenomenon. More than a hundred years ago, in the early days of the

incandescent light bulb market, which eventually replaced gas lights, patent warfare was crucial to

securing Thomas Edison’s control of the light-bulb market. In the first year of litigation, Edison’s

company was involved in nearly one hundred patent disputes (Bright, 1972; Shaver, 2012)4. Still earlier,

the 1850s witnessed the ‘sewing machine war’(Lampe and Moser, 2010; Mossoff, 2011)5. In the spirit of

4 By 1901, Edison’s side had spent about $2,000,000 on well over two hundred infringement suits under its lamp and lighting patents at a time when the company’s market capitalization was $12,000,000. The litigation was not confined to the American shores. Foreign patents and litigation, especially in Great Britain and Germany, formed an important part of Edison’s transnational litigation strategy (Bright, 1972). 5 In recent years, there have been fights over diapers, air fresheners, oil drilling equipment, and one over heart devices that has lasted more than a decade. However, none of those received the attention that is being given to the

6

Somaya (2003), the "strategic stakes" were high in these markets because the expected payoff from

litigating and driving out competition became greater than that of a cooperative mechanism. As a result,

firms actively pursued patent enforcement litigation rather than settle or cooperate. The strategic stakes

are particularly high in platform-based or standard-based markets, which often exhibit ‘winner takes all’

dynamics (Shapiro and Varian, 1999; Zhu and Iansiti, 2012). Hence, despite the wide availability of

cross-licensing and patent pools to achieve cooperation in such industries, todays’ explosion of patent

enforcement litigation seems more consistent with the “patent warfare” view suggested by Shaver (2012)6.

According to the patent warfare view, technology firms race to assemble patent portfolios—

initially for defensive purposes in the context of a dynamic and competitive field—but then, as the

industry matures, convert their shields into weapons to eliminate their competitors in pursuit of market

dominance with its platform. It is important to note that such firms need not themselves innovate to obtain

these shields and weapons but can instead hoard shields and weapons by acquiring patents. The patent

warfare view implies that patent wars, rather than cooperative outcomes, are likely to result when the

technology involved in a platform or a standard is in dispute or, more generally, when the strategic stakes

in a market are sufficiently high. This view characterizes today’s patent strategy dynamics as less

collaborative and more destructive than in the view taken by Feldman (2012)7.

On the other hand, patent enforcement is a costly competitive weapon. Direct legal and trial costs

alone can run in the millions of dollars, and indirect costs can include managerial time and other resource-

intensive distractions of capital and human resources (Agarwal et al., 2009; Landes and Posner, 2003).

Therefore, the expected payoff of patent litigation must be sufficiently high to justify the extremely costly,

disruptive, and time-consuming nature of this method of enforcement (Lemley and Shapiro, 2005; Shane

smartphone wars. http://www.bloomberg.com/news/2012-10-08/apple-phone-patent-war-like-sewing-machine-minus-violence.html, accessed February 2013. 6 The patent warfare view (Shaver, 2012) resembles the board game Risk!, in which the board resembles a map of the world and competing parties assemble strategic assets, and then battle with their rivals for world domination. 7 The view by Shaver (2012) also conforms to industry practice today. “… in practice, patents are weapons. Technology companies load up on patents like Cold War nations stockpiling nuclear bombs, hoarding them for use when an important market is at stake.” Robert McMillan, “How Apple and Microsoft Armed 4,000 Patent Warheads,” http://www.wired.com/wiredenterprise/2012/05/rockstar/, last accessed January 2013.

7

and Somaya, 2007). Moreover, there should be a reasonable expectation that the enforcement of patent

rights will be effectively honored in an effective, predictable and consistent manner. Thus, for various

reasons, markets with strong IP systems have become natural ‘battlegrounds’ for international patent

litigation. Such markets are usually member countries of international treaties, so that legal procedures are

similar across these member countries and relatively quick and efficient, which lowers litigation costs.

Such countries are also likely to offer ‘primary’ enforcement mechanisms such as preliminary injunctions,

which offer a quick and effective means of deterring imitators (Lanjouw and Lerner, 1996; Lanjouw and

Lerner, 2001). When patent infringement litigation is successful, it is extremely costly to the infringer.

Large damages or royalties can be a financial burden to the infringing firm (Bhagat, Brickley, and Coles,

1994), in addition to potential reputation damages (Lerner, 1995). But more importantly, injunctions can

be particularly devastating to businesses characterized by rapid technological change and short product

life cycles (Lanjouw and Lerner, 2001). All told, patent enforcement strategy has become an effective

competitive weapon and is likely to be used by firms in markets with strong IP protections rather than in

markets with weak IP protections, especially when the stakes in the market are sufficiently high relative

to the cost of enforcement.

The perceived danger of patent enforcement strategies employed by rivals can affect the strategic

decisions of firms (Jaffe, 2000; Lee and Mansfield, 1996). To avoid the cost and uncertainty associated

with patent wars, firms will shift their business foci to weaker IP markets for risk management purposes.

Thus, we propose the following hypothesis:

H1: As patent wars intensify, the share of a firm’s business in weak IP markets will increase.

Intensified patent wars may have heterogeneous effects on firms with different capabilities,

especially in the international context (Alcacer, Dezso, and Zhao, 2013; Shaver and Flyer, 2000). When

patents are enforced, some alleged infringers settle and agree to pay royalties, while others may both

challenge the validity of the plaintiff’s patent – thus potentially committing themselves to an even costlier

8

litigation process – and countersue from the basis of their own patent portfolios. According to the patent

warfare view, these firms should have strong technological capabilities, enabling them to produce large

patent portfolios of their own (or obtain one as a result of patent acquisitions) and effectively lower their

litigation costs (Bessen and Meurer, 2008a; Lerner, 1995; Meurer and Bessen, 2005). Such firms can even

become aggressive litigators and initiate patent enforcements themselves. In contrast, firms lacking such

technological capabilities are at a significant disadvantage, as they cannot offset their greater litigation

risk by trading patents from their portfolios to achieve a more rapid resolution of their suits (Lanjouw and

Schankerman, 2004).

Hence, in the case of patent wars, firms with greater technological capabilities can defend

themselves more effectively or even actively initiate litigation in any potential battleground, resulting in

less incentive to avoid high-stakes markets characterized by patent disputes (Jaffe, 2000). Such firms will,

therefore, be less susceptible to the effect proposed in H1. Thus, we propose the following hypothesis:

H2: The effect in H1 will be attenuated for firms with stronger technological capabilities.

The effect in H1 may also differ for firms with different institutional backgrounds. In an

international context, the "imprinting" of firms by their home-country institutional environments may

influence their capability and willingness to engage in costly litigation (Stinchcombe, 1965). Such

capabilities can carry over into international operations, as firms face new challenges in foreign

institutional environments (Zaheer, 1995).

As Somaya and McDaniel (2012) note, in the context of patent enforcement, firms from countries

with weak IP systems may develop and carry over beliefs and strategic orientations toward patents that

are deeply misaligned with a strong patent environment. Due to different beliefs shaped by institutional

differences between firms from countries with strong IP systems and firms from countries with weak IP

systems (Barkema, Bell, and Pennings, 1996), the two parties may be less likely to settle when a patent

lawsuit is filed, thereby increasing the risk of litigation (Lanjouw and Lerner, 1997; Priest and Klein,

9

1984). Differences in belief are not eliminated by hiring outside service providers who are familiar with

host-country institutions, as external advice is inherently limited in its influence on firm behavior and

learning, especially when it is impeded by deeply entrenched institutional conditioning (Somaya and

McDaniel, 2012).

Furthermore, countries with weak IP protections are less frequently a battleground for patent

litigation, owing to their ineffective enforcement systems. Thus, firms from such countries will have less

experience (and fewer benchmark cases) with patent litigation than firms from countries with strong IP

systems. Such experience asymmetry places firms from countries with weak IP protections at a

disproportionate disadvantage and increases the risk of litigation (Lanjouw and Schankerman, 2004;

Lerner, 1995). Therefore, firms from countries with weak IP systems will be more susceptible to the

effect proposed in H1, due to increased risk and uncertainty. Thus, we propose the following hypothesis:

H3: The effect in H1 will be more pronounced for firms whose home country has a weak IP system than

for firms whose home country has a strong IP system.

EMPIRICAL CONTEXT: THE GLOBAL SMARTPHONE MARKET AND THE PATENT

WARS

"The smartphone industry today is characterized by a thicket of patents and wars based on those

patents. Every day brings a new lawsuit or development between Apple, HTC, Microsoft, Motorola

Mobility (MMI), Nokia, and Samsung." (Carrier, 2012)

We test our hypotheses in the context of the global smartphone market, which has grown rapidly

in recent years. A smartphone, as distinct from a feature phone, is a mobile phone with advanced

functionalities and connectivity, incorporating such attributes as media players, compact digital cameras,

video cameras, and GPS navigation. It is typically built on a mobile operating system or platform such as

Google’s Android, Apple’s iOS, Nokia’s Symbian, and Microsoft’s Windows Phone.

10

Although the first smartphone can be traced back to early 2001, the smartphone industry did not

start to grow rapidly until Apple’s release of the iPhone in 2007. The iPhone, which is based on Apple’s

proprietary operating system, iOS, is one of the first mobile phones to use a multi-touch interface. In the

same year, Android, an open source smartphone operating system owned by Google, was unveiled along

with the founding of the Open Handset Alliance: a consortium of hardware, software, and

telecommunication companies devoted to advancing open standards for mobile devices. The first

Android-powered phone, the HTC Dream, was released in October 2008. Since then, Google’s Android

platform has gained popularity and gradually taken over the market, becoming the most popular

smartphone operating system today.

Other platforms are either declining in popularity (Blackberry, Linux, PlamOS and Symbian) or

are very small niche players (e.g., WebOS, Maemo and Bada). One exception is Microsoft’s Windows

Phone. Windows Phone is a proprietary mobile operating system developed by Microsoft and the

successor to its Windows Mobile platform, although it is incompatible with it. The new platform was

released in November 2010. Although the Windows Phone’s market share is still very small, Microsoft

increased its sales recently, due to its partnership with Nokia, which ceased to use its own smartphone

platform, Symbian, in 2011.

For some platforms, their developers are the sole hardware suppliers. For example, Apple and

BlackBerry keep their smartphone platforms closed to other smartphone vendors and build smartphone

devices themselves. On the other hand, many platform providers, such as Google, Microsoft and Nokia,

choose to make their platforms accessible to third-party vendors, enabling these vendors to build mobile

devices powered by these platforms. Some vendors use only one platform for all of their smartphones,

while others use multiple operating system platforms across their smartphones (e.g., Samsung

simultaneously sold Android-based, Microsoft-based, Symbian-based, and Bada-based smartphones).

Although vendors in this industry started to enforce patents against rivals as early as late 2009,

the industry has witnessed an explosion of patent litigation since 2011, involving many smartphone

vendors and almost every platform provider. The frequency of new patent dispute filings peaked in the

11

summer of 2011, with many firms involved in multiple lawsuits. For example, Apple has been the

plaintiff or the defendant in more than 100 patent lawsuits. Table 1 provides a list of high-profile patent

lawsuits.8 Patent lawsuits are not limited to the U.S. courts and have also been launched in various

countries. Although platform providers may not target each other directly in these patent litigation, these

patent wars are often viewed as a proxy for platform competition in this industry: as the fate of a platform

provider and its device vendors are linked, a platform provider can sue a device manufacturer of a

competing platform to increase the cost of bringing the device to market and thus slow the diffusion of the

competing platform.

'Insert Table 1 about here'

While patent disputes over hardware technology vary widely across vendors, most of the patent

disputes relating to software have concerned the Android operating system, which Google distributes free

of charge to vendors manufacturing and selling Android-powered smartphones9. Apple’s former CEO,

Steve Jobs, once said that he is willing to go to thermonuclear war over Android because, he claims, it is a

stolen product10. Thus, Apple has been aggressively pursuing patent litigation against Android mobile

phone vendors such as Samsung, HTC, and Motorola, to name a few. A series of ongoing lawsuits

between Apple and Samsung regarding the design of smartphones and the Android operating system have

become one of the most publicized patent litigation in recent history.

In the meantime, these platform providers have sought to expand their patent portfolios rapidly,

both to protect themselves and to attack their competitors. For example, Google purchased Motorola

Mobility for US$12.5 billion in 2011, primarily to acquire the rights to over 17,000 patents owned by

Motorola. In the same year, a consortium of companies, including Apple, Ericsson, Microsoft, RIM and

8 Source: http://en.wikipedia.org/wiki/Smartphone_wars, accessed March 2013. 9 In fact, most patent disputes in this industry include some sort of software-related claim. The U.S. Patent Office recently examined patents in some of the high-profile litigation cases among four major firms in the smartphone industry – Samsung, Apple, Motorola, and Microsoft – and found that, as of November 2012, 65 out of 73 patents that remained in controversy included at least one software-related claim (Graham and Vishnubhakat, 2013). 10 Steve Jobs famously invoked the metaphor of nuclear war in threatening a competing phone system, a statement that became popular in the media: “I will spend my last dying breadth if I need to, and I will spend every penny of Apple’s $40 billion in the bank, to right this wrong. I’m going to destroy Android, because it’s a stolen product. I’m willing to go to thermonuclear war on this” (Isaacson, 2011).

12

Sony, outbid Google in an auction of over 6,000 Nortel mobile-related telecommunications patents, at

US$4.5 billion. Today, industry experts constantly criticize the now out-of-control escalation of patent

wars, as they are mutually destructive and costly, resulting in a prisoners' dilemma. For example, Apple

and Google have paid their lawyers more than $400 million in their patent battles over recent years

(Radovsky, 2012), and when recently asked about his overall feelings on patent lawsuits, Apple CEO,

Tim Cook, said "It’s a waste, it’s a time suck" (Haselton, 2012).

The patent wars in the smartphone industry offer an ideal setting in which to examine the linkage

between patent litigation and the impact of the efficacy of counties’ patent systems on firm strategy. First,

most smartphone vendors have sales in multiple countries with varying degrees of patent system efficacy.

Given their limited resources and the industry characteristic of rapid technological change and short

product life cycles, vendors often must decide which countries will serve as their primary markets. The

effectiveness of the patent system in enforcing their patent rights through litigation certainly should play a

role in forming strategy and consequently shaping the competitive landscape in this industry. By

gathering data on their operations in multiple countries over time, we can test how patent wars influence

their business foci across countries. Second, the patent wars started in late 2009 and intensified

subsequently, as indicated in Table 1. Furthermore, the rapid escalation of the patent wars was largely

unexpected for industry participants and caught many firms off-guard. We can thus take advantage of this

exogenous shock to examine how non-focal firms changed their behavior as a result of the war’s

emergence.

METHODS

Data description

We construct our dataset of smartphone vendors using data from Gartner Inc.

(www.gartner.com).11 Our dataset includes the population of smartphone vendors from the first quarter

11 Gartner is a public company (NYSE: IT) founded in 1979 and is one of the world’s leading information technology research and advisory companies. Gartner research products and data have been extensively used in the

13

(1Q) of 2008 to the fourth quarter (4Q) of 2012. This period represents an early stage of the global

smartphone market as well as a period that includes the intensified patent wars among vendors. Our

dataset details vendors’ smartphone and smartphone platform (i.e., operating system) sales to users in

Asia/Pacific, Eastern Europe, the Middle East and Africa, Latin America, North America, Western

Europe, and in selected countries within those regions for each quarter during 1Q2008-4Q2012. Many

vendors have entered and exited the industry in various periods; hence, we have an unbalanced panel

dataset. While many vendors have sales in most major countries, there remains considerable variation in

the number of countries in which vendors operate and in sales levels achieved by vendors in different

countries.

Table 2 lists smartphone sales to end-users in different countries in 4Q2012, the last period in our

dataset. It shows that the two most important markets are China and the United States, which together

account for more than 45% of global smartphone sales.

'Insert Table 2 about here'

Table 3 provides sales information for the top 10 vendors in 4Q2012. The top 10 list comprises

vendors from several countries, including South Korea, the U.S., Finland, Taiwan, Canada, Japan, and

China, suggesting that competition in this industry is indeed global. Samsung and Apple are the two most

significant vendors in the global market and, together, account for more than 50% of total smartphone

sales.

'Insert Table 3 about here'

Variables

Dependent variable

industry and in prior academic studies (Alcacer and Oxley, 2012; Lechner and Floyd, 2012; Schroth and Janner, 2007).

14

To test our hypotheses, we track how many smartphone units a vendor (i) sells in a given country

or region (j) in each quarter (t). We then divide this number by the total number of units the vendor sells

in that quarter across all countries and regions. Thus,

ݎݕݎݐݑܥ ௧ ൌ ௧σ ௧

where ௧ indicates the number of smartphone units sold by vendor i, in country j, in quarter t, and ݎݕݎݐݑܥ௧ indicates the "country share" in vendor i’s business in quarter t (i.e., the share of

vendor i’s sales, in country j, in quarter t). In contrast to sales measures, this measure reflects the strategic

importance of a country to a vendor in a given period and is independent of the growth of the vendor’s

overall sales.

Because our dependent variable is continuous and constrained between 0 and 1, we follow

previous studies (Hausman and Leonard, 1997; Liu et al., 1999; Loudermilk, 2007; Wagner, 2001, 2003)

and use a fractional logit model (Papke and Wooldridge, 1996, 2008) in our analysis, which uses simple

quasi-likelihood estimation methods for regression models with a fractional dependent variable.

Compared with log-odds type procedures, the fractional logit model has the advantage of handling data at

the extreme values of 0 and 1 (Papke and Wooldridge, 1996, 2008). The unit of analysis for testing our

hypotheses is the vendor-country-quarter.

Independent variables

PatentWarIntensity. We construct a proxy measure for patent war intensity in the smartphone

industry by using LexisNexis Academic to count the number of media articles related to smartphone patent

litigation covered in all ‘major world publications’ (e.g., the New York Times, the Wall Street Journal,

the Korea Times, The Australian and so forth) on the grounds that, as patent wars intensify, so does the

frequency of media coverage of smartphone patent litigation and knowledge of it among vendors. We

counted the number of articles using the two keywords "smartphone" and "patent" and their use in various

other combinations. These articles may include information on which vendor sued which vendor in which

15

country, major licensing and settlement announcements, and major patent acquisition announcements

during our sample period. Figure 1 depicts our measure of patent war intensity graphically. It shows that

there was a noticeable escalation of patent litigation in the later part of our sample period.

IP_index. This variable is a measure of the strength of the IP system in a given country. It is a

measure based on the Intellectual Property Rights Index (IPRI), published annually since 2008 by the

Property Rights Alliance through the Hernando de Soto Fellowship Program. This index combines the

methodology of Ginarte and Park (1997)12, which has been extensively used in prior studies (Oxley, 1999;

Somaya and McDaniel, 2012; Zhao, 2006), and an opinion-based measure of the protection of IP (World

Economic Forum’s Global Competitiveness Index), where expert participants in each country are asked to

rate their nation’s IP protection, scoring it from "weak and not enforced" to "strong and enforced." An

increase in the IP_index reflects increased IP protection in a given country. Therefore, to test H1, we

expect the interaction term PatentWarIntensity * IP_index to have a negative and statistically significant

effect on our dependent variable.

TechCapability. This is a vendor-specific measure of technological capability. Consistent with

prior studies (Ahuja, 2000; Narin, Noma, and Perry, 1987; Song and Shin, 2008), we count the number of

all smartphone-related patents13 each vendor has at the end of 2008, using the United States Patent and

Trademark Office patent database. Following prior studies (Ahuja, 2000; Stuart and Podolny, 1996), we

use patent data from the United States Patent and Trademark Office for all firms, including firms

headquartered outside the United States, to maintain consistency, reliability, and comparability, as

patenting systems across nations differ in their applications of standards, systems for granting patents, and

the value of the protections granted (Basberg, 1987; Griliches, 1990). The United States represents one of

12 Ginarte and Park (1997) construct an index that reflects the strength of a country’s patent laws based on five extensive criteria: coverage, membership in international treaties, restrictions on patent rights, enforcement, and duration of protection. 13 These include patents under the international patent classification code H04 (Electric communication technique), which covers the majority of smartphone-related patents. As a robustness check, we then further include H01 (Basic electric elements), H03 (Basic electronic circuitry), G02 (Optics), G06 (Computing, calculating, counting), and G11 (Information storage) and find virtually no change in our results.

16

the largest markets for smartphones, and firms commercializing their inventions would most likely patent

in the United States, if they were to patent anywhere at all.

We choose to count the number of patents for each vendor at the end of 2008 for two reasons.

First, given that our sample period ends in 2012 and that a typical patent application takes approximately

3-4 years to be granted, the number of patents granted preceding our sample period reflects the vendor’s

technological capability free of any right-censoring concerns. Second, and more importantly, after the

patent war starts in 2009, vendors may start to engage in strategic patenting behavior. Our measure is thus

free from this endogeneity concern.

We then take the log of 1 plus the total patent counts to account for skewness. To test H2, we

expect the interaction term PatentWarIntensity * TechCapability * IP_index to be positive and

statistically significant.

WeakOrigin. This is a vendor-specific dummy variable equal to 1 if a vendor comes from a

country with weak intellectual property protections and 0 otherwise. We locate a vendor’s country of

origin based on its headquarters (Zhang et al., 2010) and determine whether the home country has strong

or weak IP protections using our IP_index. If its average IP index is below the median of all countries’

average IP indices across our whole study period, WeakOrigin = 1, and 0 otherwise. To test H3, we

expect the interaction term PatentWarIntensity * WeakOrigin * IP_index to be negative and statistically

significant.

Vendors’ business foci may also be affected by many other factors, so we first construct a few

measures to control for demand-side drivers. One may be concerned that vendors might shift their

business foci to emerging markets (which may be correlated with weak IP countries) because their

income levels or populations are rising. To control for the relative market size, we use variations in

countries' shares of world GDP total (GDPshare), which is constructed using both GDP per capita and

population level data from the International Monetary Fund (IMF) World Economic Outlook database.

Vendors may shift their business foci because of different levels of interests in smartphones across

different countries. To control variations in such levels of interest, we construct a measure of smartphone-

17

specific market share by computing a country’s share of world total smartphone sales in a given quarter

(CountrySalesShare).

We then construct a few measures to control for supply-side drivers. One would expect that

vendors might decide on their business foci based on competitive dynamics in each country, so we count

the number of vendors competing in each country in each quarter (NumFirms) and compute a country-

level Herfindahl index (HHI) to account for market concentration (Xia, Tan, and Tan, 2008).

To facilitate a causal interpretation, we lag both our demand- and supply-side measures by one

quarter. Because some of our variables only exist in the form of annual data (e.g., GDP and population),

we convert all of our annual measures into quarterly measures through linear interpolation, assuming four

equal increments between years. Table 4 shows summary statistics of our variables.

'Insert Table 4 about here'

RESULTS

Figure 2a shows the share of smartphone sales in strong and weak IP countries, where “strong”

and “weak” designations are based on whether a country's IP_index is above or below the median, while

Figure 2b presents the trend in sales volume of smartphones in strong and weak IP countries. Both figures

show that sales of smartphones in strong and weak IP countries were approximately 50-50 until 4Q2010,

after which there are more sales in weak IP countries than in strong IP countries. While suggestive, this

trend is expected and consistent with H1.

We now turn to our regression analysis results. To address endogeneity concerns stemming from

unobserved vendor heterogeneity, we include vendor fixed effects in our regression analysis, and include

country fixed effects to control for time-invariant country characteristics. Intertemporal trends and macro-

environmental shocks are controlled for by quarter fixed effects, which thus absorb the main effect of

PatentWarIntensity. Table 5 reports our regression results. The dependent variable is each vendor’s sales

in a given country-quarter divided by that vendor’s total sales in that quarter, ݎݕݎݐݑܥ௧. Model

(1) uses all vendors in our sample. The interaction term PatentWarIntensity * IP_index is negative and

18

statistically significant, suggesting that - consistent with H1 - smartphone vendors shift their business foci

to weak IP countries rather than strong IP countries as the patent war intensifies, all else being equal.

We then split the sample into vendors who were and were not involved in any patent litigation

during the sample period, and repeat the analysis separately for each group. We consider a vendor to be

involved in patent litigation if it was a plaintiff or defendant in any outstanding smartphone patent lawsuit

during the sample period. Models (2) and (3) report the results. Interestingly, we find that the interaction

effect is much more pronounced and statistically significant only for those vendors not involved in patent

litigation, suggesting that the shift in business foci is not driven by any direct effects of patent litigation

(e.g., some vendors are banned from selling their phones as a result of patent litigation). The result in

Model (2) also increases our confidence for a causal interpretation, as the vendors included in Model (2)

are not involved in the patent war. In this case, the marginal effect of one standard deviation increase in

the PatentWarIntensity computed at the mean value of all other variables leads to a reduction in the

dependent variable by 0.007. Given that the mean value of our dependent variable is 0.051 and the mean

value of IP_index (6.45) is greater than the median (6.32), this implies a 13.7% reduction in sales in a

given strong IP country.

In Model (4), instead of using the interaction term PatentWarIntensity * IP_index, we interact

quarter dummies with the IP_index to break down the time trend and see when the shift in business foci

effect is taking place during the sample period14. As it takes time for vendors to adjust their business foci,

we expect the effect to show up well after the patent war emerges – and, indeed, our result suggests that

4Q2011-3Q2012 is the salient period. In sum, the results in Table 5 support H1.

'Insert Table 5 about here'

In Table 6, to test H2, we examine the moderating effect of vendors’ technological capabilities.

While the inclusion of a three-way interaction term in a full-sample analysis produces unbiased consistent

14 In model (4), because we have a long list of quarter interaction terms, for brevity, we only indicate statistical significance of the estimated coefficients and do not report standard errors. For the same reason, we do not report quarter interaction terms as an alternative model specification in subsequent tables, although all such analyses are consistent with our hypotheses.

19

estimates of the parameters of interest, the interpretation of a three-way interaction term may nevertheless

be challenging. Hence, we first present a split-sample analysis in Models (1) and (2), based on whether a

vendor has strong technological capabilities, as measured by whether the TechCapability variable is

above or below the median-level. The interaction term, PatentWarIntensity * IP_index, in Models (1) and

(2), suggests that vendors with strong technological capabilities may not be affected by the escalation of

patent wars, while vendors with weak technological capabilities are affected and shift their business foci

to weak IP countries. We then include the three-way interaction term PatentWarIntensity *

TechCapability * IP_index, and repeat the analysis for all vendors, for only vendors not involved in

litigation, and for only vendors involved in litigation, respectively - the results are reported in Models (3)-

(5). We find that the coefficient of the three-way interaction term is positive and statistically significant in

Models (3) and (4), implying that the effect in H1 is attenuated for vendors with stronger technological

capabilities. The results in Models (4) and (5) also imply that these effects apply only to those vendors

that are not involved in any patent litigation, which facilitates a causal interpretation. Overall, the results

in Table 6 support H2.

'Insert Table 6 about here'

In Table 7, we examine the moderating effect of vendors’ countries of origin. Mirroring the

results presented in Table 6, we first present a split-sample analysis in Models (1) and (2), based on

whether a vendor's country of origin is a weak or a strong IP country. The interaction term,

PatentWarIntensity * IP_index, in Models (1) and (2), suggests that vendors from weak IP countries shift

their business foci to weak IP countries as patent wars intensify. In models (3) and (4), the coefficient of

the three-way interaction term, PatentWarIntensity * WeakOrigin * IP_index, is negative and statistically

significant. This implies that the effect in H1 is more pronounced for vendors whose home country has a

weak IP system than for firms whose home country has a strong IP system. The results in Models (4) and

(5) also show that these effects apply only to vendors not involved in any patent litigation, which

continues to facilitate a causal interpretation.

20

While WeakOrigin captures the "imprinting" of firms by their home-country institutional

environments, which may influence their capability and willingness to engage in costly litigation, a

possible concern is that our measure is also strongly correlated with poor manufacturing quality. That is,

vendors from weak IP countries may have limited capabilities to produce competitive smartphones and

therefore must sell their products in weak IP countries. However, if it is poor capabilities to make phones

that are driving our results, we would expect these vendors to sell their products in weak IP countries all

the time and that any change in business focus should not be picked up by a time trend. In an unreported

analysis, we investigate the time trend in vendors' changes in strategy for those from weak IP countries,

mimicking Model (4) in Table 5, by interacting quarter dummies with the IP_index. Similar to the result

in Model (4) of Table 5, we find that vendors from weak IP countries start shifting their sales to weak IP

countries starting from 3Q2011 (but not before) and increasingly thereafter, which strongly rejects this

alternative explanation. For the alternative explanation to hold, we would have to believe that these

vendors suddenly lose their ability to make phones relative to vendors from strong IP countries. In reality,

we believe the opposite is true – that the gap is decreasing due to imitation. All in all, the results in Table

7 are not consistent with this alternative explanation, but rather support H3.

'Insert Table 7 about here'

Consequences of Vendors’ Shift in Business Foci

In this section, we explore the consequences of vendors’ strategies resulting from the escalation

of the patent wars. As noted above, the intensified patent litigation war is deemed a proxy war for

operating system platform wars. In particular, closed systems, such as Apple's iOS, Microsoft Windows

Phone, and Blackberry, intensively attacked Google's Android-powered smartphones. As Android was the

center of most attacks, we expect vendors manufacturing Android-powered smartphones to shift their

business foci more toward weak IP countries. As a result, we expect the market share of Android

operating system to grow more in weak IP countries relative to strong IP countries.

21

To test this prediction, we track which operating system has been adopted in the smartphones that

vendor i sells in country j in quarter t. We then compute each operating system's market share in each

country for each quarter.

Figure 3 depicts sales of Android smartphones in strong and weak IP countries. It shows that

sales of Android phones in weak IP countries exhibit exponential growth, while sales of Android phones

in strong IP countries show more linear growth. Again, while only suggestive, the trend in Figure 3 is

consistent with our above prediction. In addition, Figure 4 shows the market shares of Android, iOS, and

all other platforms. We can see that over time the platform battle became a two-way battle between

Google’s Android and Apple’s iOS, with all other operating systems collectively managing only a small

market share globally.

‘Insert Figure 3 about here’

‘Insert Figure 4 about here’

These findings are borne out by Model (1) of Table 8, which investigates the market share of

Android smartphones in strong IP and weak IP systems, based on our IP_index. The dependent variable is

the Android’s market share in a given country, and the unit of analysis is the platform-country-quarter.

We include the lag of Android’s market share to control for the interest level of Android devices in each

country and the lag of GDPshare to control for relative market size. The interaction term,

PatentWarIntensity * IP_index, is negative and statistically significant, which implies that Android

platform sales grow faster in weak IP countries than in strong IP countries as the patent war intensifies,

consistent with our prediction. Thus, as Android phones faced increasing litigation risks in strong IP

countries, along with the fact that more vendors were adopting the open-system Android free of charge as

the primary operating system for their phones, Android proliferated in a very short period of time in weak

IP countries. Interestingly, it appears that the patent wars intended to hamper the proliferation of Android

phones have merely shifted its sales to weak IP countries, where it has flourished.

'Insert Table 8 about here'

22

Robustness Checks

Following previous studies (Tanriverd 砦 and ee, 200 ), our analysis uses vendors’ sales shares in

different countries to measure their business foci. As these can be measured in many different ways, we

use new product introductions as an alternative measure, and test whether our results hold with this new

measure as a robustness check. We expect vendors to introduce more new smartphones in countries or

regions that are important to their businesses, whereas, if they avoid markets with strong IP markets

because of increased litigation risk, then we can expect them to also minimize their exposure to such risk

by being selective and minimizing the number of product models released. A recent study by Toh and

Kim (2012) echoes such firm strategy in the context of technological uncertainty and a firm's

technological specialization (i.e. a firm's technological specialization increases as it faces highly litigious

rivals). For this analysis, we use alternative data sources – GfK and GSMArena.com – and an alternative

dependent variable, the number of smartphone models released in each country. GfK (www.gfk.com) is

one of the world's leading market research companies based in Germany: for each country, our GfK data

include vendor name, product name, product ID number, product's earliest marketing date (month & year),

and its non-subsidized price for each quarter. GSMArena.com (www.gsmarena.com) provides information

to consumers – including vendor name, product name, product release date, and other technical product

features – to help them choose a mobile device. We scrape data from this source using a web crawler, and

then use it to cross-check with our GfK data and fill in missing data, to form our final dataset for the

robustness check analysis.

Models (2)-(4) of Table 8 report the results. We use negative binomial models, because our

dependent variable is a count variable, and find that, as predicted, vendors launch a smaller number of

product models in strong IP countries as the patent war intensifies, and that this effect is more pronounced

for vendors not involved in patent litigation. We are aware that our results might be driven by systematic

changes in vendors’ pricing strategies across countries. As a robustness check, using data from GfK and

GSMArena.com, we compute the average price level of a vendor’s smartphones in a given country and

quarter. In a series of (unreported) fixed-effects panel regressions, we repeat the analysis in Table 5 after

23

replacing our dependent variable with the price level (in logarithm), but find no evidence of shifts in

pricing strategies as the patent war intensifies.

Finally, we are concerned that vendors’ shift toward weak IP countries might be driven by the

deployment of infrastructures to support smartphone technologies (e.g. 3G/4G technology) in those

countries in recent years, although this explanation does not explain the systematic differences between

vendors with different levels of technological capabilities. Nevertheless, we conduct a robustness check

by restricting countries to those which already had infrastructures that supported smartphones at the

beginning of our sample period, which we identified by whether it registered smartphone sales in

1Q2008,– i.e., Brazil, Canada, China, France, Germany, Italy, Japan, Mexico, Russia, Spain, U.K., and

the U.S. Repeating the analyses only with this set of countries, we find similar results (unreported).

DISCUSSION AND CONCLUSION

This paper views firm’s patent as a competitive weapon and shows how patent litigation affects

market participants who are not themselves directly involved in the litigation in the global market for

smartphones. In recent years, suing over patents has become another powerful competitive weapon,

similar to marketing and pricing. Thus, patent strategy and its enforcement are becoming increasingly

important for value appropriation by innovators (Teece, 1986), and strategy scholars can benefit from a

better understanding of patent enforcement strategies and their consequences.

This study sheds new light on firms' patent enforcement strategies by highlighting the

significance of heterogeneity in the effectiveness of national patent systems in global competition. Firms

use markets with strong IP protection as a natural battleground for the deployment of their patent

enforcement strategies, which leads to increased litigation risk for other participants in those markets.

Consistent with our theoretical argument, we find that, as patent wars intensify, firms focus their

businesses increasingly on markets with weak IP protection than on markets with strong IP protection,

even when such firms are not involved in patent litigation. This effect is attenuated for firms with stronger

technological capabilities and is more pronounced for firms from countries with weak IP systems. The

24

main driver of our theoretical argument and empirical analyses is the disproportionate increase in

litigation risk in strong IP countries versus weak IP countries. Our results show that vendors that are more

confident in their litigation capability are less susceptible to the increased litigation risk.

Our study contributes to the emerging patent enforcement strategy literature (Agarwal et al., 2009;

Clarkson and Toh, 2010; Polidoro and Toh, 2011; Somaya, 2003; Somaya, 2012; Somaya and McDaniel,

2012; Ziedonis, 2004) that focuses on how firms leverage their patents as business assets for competitive

advantage. We argue that firms race to assemble patent portfolios and eventually convert their shields into

weapons to eliminate their competitors in pursuit of market dominance. This study also contributes to the

business ecosystem literature (Adner, 2006, 2012; Adner and Kapoor, 2010; Iansiti and Levien, 2004a,

2004b; Moore, 1996; Zhu and Iansiti, 2012). An increasing number of today’s markets have evolved into

highly interconnected organizational networks - often referred to as business ecosystems (Iansiti and

Levien, 2004a; Moore, 1996) – and this interconnectivity means that effects on some ecosystem

participants are likely to extend to affect the rest. In our setting, the deterrence effect of patent litigation

affects vendors not directly involved in patent litigation, as well as influencing the market shares of

different smartphone platforms.

Our study offers some practical managerial implications. First, our findings suggest that rivals in

global industries may not be deterred from competing in high-stake markets by patent enforcement –

instead they may simply shift their business foci to markets with weak IP systems where such

enforcement is ineffective. In our setting, some smartphones vendors - such as Apple - launch a series of

intensified patent litigation with the intention of gaining competitive advantage in specific markets, only

to find themselves competitively disadvantaged in the global platform war. Hence, it may not always be

in a firm's best interest to deter imitation using patent litigation (Polidoro and Toh, 2011). While currently

pending patent litigation may be years away from a conclusive outcome, it seems likely that a strong

Android ecosystem has emerged in the majority of the world, partly because of heterogeneity in the

efficacy of different IP systems, and partly because Android, as an ‘open’ software system, is available to

vendors free of charge. Hence, managers need to be mindful of the broader consequences of using patent

25

litigation as a deterrence mechanism when competing on a global scale in heterogeneous markets.

Second, digital convergence implies that many future technological innovations will grow

increasingly complex and are likely to rely on many patents. Our results highlight the importance of

holding valuable patents as a defense mechanism: competitors who do not have patents with which to

countersue are likely to be forced to shift their business foci, even if they are not directly involved in

litigation. Because the fates of participants in one ecosystem are interlinked, competition between

platform providers can take the form of indirect wars between any two participants from different

ecosystems. Thus, in addition to paying close attention to direct competitors within the same ecosystem

(e.g. Samsung vs. HTC within the Android ecosystem), participants need to be cognizant of potential

threats from different ecosystems (e.g. Nokia in the Windows Phone ecosystem vs. Samsung in the

Android ecosystem).

This study naturally has several limitations. First, the range of participants in the smartphone

industry is not limited to mobile device manufacturers and platform providers - other business ecosystem

participants (such as telecommunication operators and application developers) may also need to respond

strategically to patent wars. As patent wars are waged among device makers and platform providers, we

believe that our results show only the first-order effects. Limitations on data availability mean we cannot

directly show how other players respond strategically as patent wars intensify. Future research may gather

additional data to explore their impact on other types of industry participants. Second, the theoretical

argument introduced here may hinge on whether the product in question involves a patent thicket and is

subject to a patent war, so the generalizability of our theory should be taken with caution. Future studies

can explore other industry settings to extend our findings.

The worldwide smartphone market is still growing, and its dynamics are yet to play out in full.

Despite these limitations, we believe this study enhances our understanding of patents, patent

enforcement strategies, and the dynamics of patent wars in platform competition, and significantly

contributes to the emerging patent enforcement strategy literature.

26

References

Adner R. 2006. Match your innovation strategy to your innovation ecosystem. Harvard Business Review 84(4): 98.

Adner R. 2012. The wide lens: A new strategy for innovation. Portfolio/Penguin.

Adner R, Kapoor R. 2010. Value creation in innovation ecosystems: How the structure of technological interdependence affects firm performance in new technology generations. Strategic Management Journal 31(3): 306-333.

Agarwal R, Ganco M, Ziedonis RH. 2009. Reputations for toughness in patent enforcement: Implications for knowledge spillovers via inventor mobility. Strategic Management Journal 30(13): 1349-1374.

Ahuja G. 2000. The duality of collaboration: Inducements and opportunities in the formation of interfirm linkages. Strategic Management Journal 21(3): 317-343.

Alcacer J, Dezso CL, Zhao M. 2013. Firm rivalry, knowledge accumulation, and mne location choices. Journal of International Business Studies.

Alcacer J, Oxley J. 2012. Learning by supplying. Harvard Business School Strategy Unit Working Paper(12-093).

Barkema HG, Bell JHJ, Pennings JM. 1996. Foreign entry, cultural barriers, and learning. Strategic Management Journal 17(2): 151-166.

Basberg BL. 1987. Patents and the measurement of technological change: A survey of the literature. Research Policy 16(2–4): 131-141.

Bessen J, Meurer MJ. 2008a. Do patents perform like property. Academy of Management Perspectives 22(3): 8-20.

Bessen J, Meurer MJ. 2008b. Patent failure: How judges, bureaucrats, and lawyers put innovators at risk. Princeton University Press.

Bhagat S, Brickley JA, Coles JL. 1994. The costs of inefficient bargaining and financial distress: Evidence from corporate lawsuits. Journal of Financial Economics 35(2): 221-247.

Bittlingmayer G. 1988. Property rights, progress, and the aircraft patent agreement. Journal of Law and Economics: 227-248.

Bright AA. 1972. The electric-lamp industry: Technological change and economic development from 1800 to 1947. Arno Press.

Carrier M. 2012. A roadmap to the smartphone patent wars and frand licensing. CPI Antitrust Chronicle 2.

Clarkson G, Toh PK. 2010. ‘Keep out’ signs: The role of deterrence in the competition for resources. Strategic Management Journal 31(11): 1202-1225.

27

Cohan P. 2012. 5 reasons to scrap our patent system. http://www.forbes.com/sites/petercohan/2012/10/08/5-reasons-to-scrap-our-patent-system-1-apples-siri/ (last accessed Feburary 2013).

Cohen WM, Nelson RR, Walsh JP. 2000. Protecting their intellectual assets: Appropriability conditions and why u.S. Manufacturing firms patent (or not). National Bureau of Economic Research Working Paper Series No. 7552.

Delgado M, Kyle M, McGahan AM. 2013. Intellectual property protection and the geography of trade. Working Paper.

Duhigg C, Lohr S. 2012. The patent, used as a sword. http://www.nytimes.com/2012/10/08/technology/patent-wars-among-tech-giants-can-stifle-competition.html?pagewanted=all (last accessed Feburary 2013).

Feldman R. 2012. Rethinking patent law. Harvard University Press.

Gilbert R, Shapiro C. 1990. Optimal patent length and breadth. The RAND Journal of Economics 21(1): 106-112.

Ginarte JC, Park WG. 1997. Determinants of patent rights: A cross-national study. Research Policy 26(3): 283-301.

Grabowski HG, Vernon JM. 1982. The pharmaceutical industry. In Government and technical progress, a cross-industry analysis. Nelson RR (ed.).

Graham S, Vishnubhakat S. 2013. Of smart phone wars and software patents. Journal of Economic Perspectives 27(1): 67-86.

Griliches Z. 1990. Patent statistics as economic indicators: A survey. Journal of Economic Literature 28(4): 1661-1707.

Grindley PC, Teece DJ. 1997. Managing intellectual capital: Licensing and cross-licensing in semiconductors and electronics. California Management Review 39(2): 8-41.

Hall BH, Ziedonis RH. 2001. The patent paradox revisited: An empirical study of patenting in the us semiconductor industry, 1979-1995. RAND Journal of Economics: 101-128.

Haselton T. 2012. Apple ceo tim cook: Patent wars are a “pain in the ass”. http://www.technobuffalo.com/2012/05/29/apple-ceo-tim-cook-patent-wars-are-a-pain-in-the-ass/ (last accessed February 2013).

Hausman JA, Leonard GK. 1997. Superstars in the national basketball association: Economic value and policy. Journal of Labor Economics 15(4): 586-624.

Iansiti M, Levien R. 2004a. The keystone advantage: What the new dynamics of business ecosystems mean for strategy, innovation, and sustainability. Harvard Business Press.

Iansiti M, Levien R. 2004b. Strategy as ecology. Harvard Business Review 82(3): 68-81.

Isaacson W. 2011. Steve jobs. JC Lattès.

28

Jaffe AB. 2000. The u.S. Patent system in transition: Policy innovation and the innovation process. Research Policy 29(4–5): 531-557.

Joshi AM, Nerkar A. 2011. When do strategic alliances inhibit innovation by firms? Evidence from patent pools in the global optical disc industry. Strategic Management Journal 32(11): 1139-1160.

Klemperer P. 1990. How broad should the scope of patent protection be? The RAND Journal of Economics 21(1): 113-130.

Kyle MK, McGahan AM. 2012. Investments in pharmaceuticals before and after trips. Review of Economics and Statistics 94(4): 1157-1172.

Lampe R, Moser P. 2010. Do patent pools encourage innovation? Evidence from the nineteenth-century sewing machine industry. Journal of Economic History 70(04): 898-920.

Landes WM, Posner RA. 2003. The economic structure of intellectual property law. Harvard University Press.

Lanjouw JO, Lerner J. 1996. Preliminary injunctive relief: Theory and evidence from patent litigation. National Bureau of Economic Research Working Paper Series No. 5689.

Lanjouw JO, Lerner J. 1997. The enforcement of intellectual property rights: A survey of the empirical literature. National Bureau of Economic Research Working Paper Series No. 6296.

Lanjouw JO, Lerner J. 2001. Tilting the table? The use of preliminary injunctions. Journal of Law and Economics 44(2): 573-603.

Lanjouw JO, Schankerman M. 2001. Characteristics of patent litigation: A window on competition. RAND Journal of Economics 32(1): 129-151.

Lanjouw Jean O, Schankerman M. 2004. Protecting intellectual property rights: Are small firms handicapped? Journal of Law and Economics 47(1): 45-74.

Lechner C, Floyd SW. 2012. Group influence activities and the performance of strategic initiatives. Strategic Management Journal 33(5): 478-495.

Lee J-Y, Mansfield E. 1996. Intellectual property protection and u.S. Foreign direct investment. The Review of Economics and Statistics 78(2): 181-186.

Lemley MA. 1997. Economics of improvement in intellectual property law. Texas Law Review 75: 989.

Lemley MA, Shapiro C. 2005. Probabilistic patents. Journal of Economic Perspectives 19(2): 75-98.

Lerner J. 1995. Patenting in the shadow of competitors. Journal of Law and Economics 38(2): 463-495.

Lerner J. 2002. 150 years of patent protection. American Economic Review 92(2): 221-225.

Levin RC, Klevorick AK, Nelson RR, Winter SG. 1987. Appropriating the returns from industrial research and development. Brookings Papers on Economic Activity 1987(3): 783-831.

Liu J-L, Liu J-T, Hammitt JK, Chou S-Y. 1999. The price elasticity of opium in taiwan, 1914–1942. Journal of Health Economics 18(6): 795-810.

29

Loudermilk MS. 2007. Estimation of fractional dependent variables in dynamic panel data models with an application to firm dividend policy. Journal of Business & Economic Statistics 25(4): 462-472.

Mansfield E, Schwartz M, Wagner S. 1981. Imitation costs and patents: An empirical study. The Economic Journal: 907-918.

Mazzoleni R, Nelson RR. 1998. The benefits and costs of strong patent protection: A contribution to the current debate. Research Policy 27(3): 273-284.

Merges RP. 1996. Contracting into liability rules: Intellectual property rights and collective rights organizations. California Law Review 84(5): 1293-1393.

Merges RP, Menell PS, Lemley MA, Jorde TM. 1997. Intellectual property in the new technological age. Aspen Law & Business.

Merges RP, Nelson RR. 1990. On the complex economics of patent scope. Columbia Law Review 90(4): 839-916.

Meurer MJ, Bessen J. 2005. The patent litigation explosion. Paper presented at the American Law & Economics Association Annual Meetings.

Moore JF. 1996. The death of competition: Leadership and strategy in the age of business ecosystems. HarperBusiness New York.

Moser P. 2005. How do patent laws influence innovation? Evidence from nineteenth-century world's fairs. The American Economic Review 95(4): 1214-1236.

Mossoff A. 2011. The rise and fall of the first american patent thicket: The sewing machine war of the 1850s. Arizona Law Review 53(1): 165.

Narin F, Noma E, Perry R. 1987. Patents as indicators of corporate technological strength. Research Policy 16(2–4): 143-155.

Nordhaus WD. 1969. An economic theory of technological change. The American Economic Review 59(2): 18-28.

Oxley JE. 1999. Institutional environment and the mechanisms of governance: The impact of intellectual property protection on the structure of inter-firm alliances. Journal of Economic Behavior & Organization 38(3): 283-309.

Papke LE, Wooldridge JM. 1996. Econometric methods for fractional response variables with an application to 401 (k) plan participation rates. Journal of Applied Econometrics 11(6): 619-632.

Papke LE, Wooldridge JM. 2008. Panel data methods for fractional response variables with an application to test pass rates. Journal of Econometrics 145(1–2): 121-133.

Park WG. 2008. International patent protection: 1960–2005. Research Policy 37(4): 761-766.

Polidoro F, Toh PK. 2011. Letting rivals come close or warding them off? The effects of substitution threat on imitation deterrence. Academy of Management Journal 54(2): 369-392.

Priest GL, Klein B. 1984. The selection of disputes for litigation. Journal of Legal Studies 13(1): 1-56.

30

Radovsky D. 2012. Can patent trolling be legislated away? http://www.fool.com/investing/general/2012/08/02/can-patent-trolling-be-legislated-away.aspx (last accessed May 2013).

Sakakibara M, Branstetter L. 2001. Do stronger patents induce more innovation? Evidence from the 1988 japanese patent law reforms. The RAND Journal of Economics 32(1): 77-100.

Schroth C, Janner T. 2007. Web 2.0 and soa: Converging concepts enabling the internet of services. IT Professional 9(3): 36-41.

Scotchmer S. 1991. Standing on the shoulders of giants: Cumulative research and the patent law. The Journal of Economic Perspectives 5(1): 29-41.

Shane S, Somaya D. 2007. The effects of patent litigation on university licensing efforts. Journal of Economic Behavior & Organization 63(4): 739-755.

Shapiro C, Varian HR. 1999. The art of standards wars. California Management Review 41(1).

Shaver JM, Flyer F. 2000. Agglomeration economies, firm heterogeneity, and foreign direct investment in the united states. Strategic Management Journal 21(12): 1175-1193.

Shaver L. 2012. Illuminating innovation: From patent racing to patent war. Washington and Lee Law Review 69(3).

Somaya D. 2000. Obtaining and protecting patents in the united states, europe and japan. In Regulatory encounters: Multinational corporations and american adversarial legalism. Axelrad L, Kagan RA (eds.), University of California Press.

Somaya D. 2003. Strategic determinants of decisions not to settle patent litigation. Strategic Management Journal 24(1): 17-38.

Somaya D. 2012. Patent strategy and management: An integrative review and research agenda. Journal of Management 38(4): 1084-1114.

Somaya D, McDaniel CA. 2012. Tribunal specialization and institutional targeting in patent enforcement. Organization Science 23(3): 869-887.

Song J, Shin J. 2008. The paradox of technological capabilities: A study of knowledge sourcing from host countries of overseas r&d operations. Journal of International Business Studies 39(2): 291-303.

Stinchcombe AL. 1965. Social structure and organizations. In Handbook of organizations. March JG (ed.), Rand McNally & Company: Chicago.

Stuart TE, Podolny JM. 1996. Local search and the evolution of technological capabilities. Strategic Management Journal 17(S1): 21-38.

Tanriverd 砦 Hn, ee C-H. 2008. Within-industry diversification and firm performance in the presence of network externalities: Evidence from the software industry. Academy of Management Journal 51(2): 381-397.

Teece DJ. 1986. Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy 15(6): 285-305.

31

Toh PK, Kim T. 2012. Why put all your eggs in one basket? A competition-based view of how technological uncertainty affects a firm's technological specialization. Organization Science.

Von Hippel E. 1988. The sources of innovation. Oxford University Press.

Wagner J. 2001. A note on the firm size-export relationship. Small Business Economics 17(4): 229-237.

Wagner J. 2003. Unobserved firm heterogeneity and the size-exports nexus: Evidence from german panel data. Review of World Economics / Weltwirtschaftliches Archiv 139(1): 161-172.

Xia J, Tan J, Tan D. 2008. Mimetic entry and bandwagon effect: The rise and decline of international equity joint venture in china. Strategic Management Journal 29(2): 195-217.

Zaheer S. 1995. Overcoming the liability of foreignness. Academy of Management Journal 38(2): 341-363.

Zhang Y, Li H, Li Y, Zhou L-A. 2010. Fdi spillovers in an emerging market: The role of foreign firms' country origin diversity and domestic firms' absorptive capacity. Strategic Management Journal 31(9): 969-989.

Zhao M. 2006. Conducting r&d in countries with weak intellectual property rights protection. Management Science 52(8): 1185-1199.

Zhu F, Iansiti M. 2012. Entry into platform-based markets. Strategic Management Journal 33(1): 88-106.

Ziedonis RH. 2004. Don't fence me in: Fragmented markets for technology and the patent acquisition strategies of firms. Management Science 50(6): 804-820.

32

Table 1. Major patent lawsuits among smartphone vendors listed in chronological order Year Month Plaintiff Defendant Country suit was filed 2009 Oct Nokia Apple U.S.A. 2010 Mar Apple HTC U.S.A. 2010 Sep Apple Nokia U.K. 2010 Sep Apple Nokia Germany 2010 Oct Microsoft Motorola U.S.A. 2010 Oct Motorola Apple U.S.A. 2010 Dec Nokia Apple Netherlands 2010 Dec Sony Ericsson LG Electronics U.S.A. 2011 Apr Apple Samsung U.S.A. 2011 Apr Samsung Apple South Korea 2011 Apr Samsung Apple Japan 2011 Apr Samsung Apple Germany 2011 Jun Samsung Apple U.K. 2011 Jun Samsung Apple Italy 2011 Jul Microsoft Samsung U.S.A. 2011 Jul HTC Apple U.K. 2011 Aug Apple Samsung Australia 2011 Aug Apple Samsung Netherlands 2011 Sep Samsung Apple France 2012 May Microsoft Motorola Germany 2012 Nov Sony Ericsson Samsung U.S.A. Notes: Duplicate pairs in the same country are not listed, such as countersuits or additional related suits. Non-practicing entities (e.g., patent trolls) are not included. Table 2. Unit sales of smartphones to end users in 4Q2012 Country Sales to End Users (# of units) Market Share China 56,641,200 27.3% United States 39,690,700 19.1% United Kingdom 9,484,500 4.6% Japan 8,733,800 4.2% Germany 6,331,500 3.0% Brazil 6,277,000 3.0% Korea 5,307,700 2.6% France 5,178,700 2.5% India 4,606,100 2.2% Russia 4,562,500 2.2% Mexico 3,705,400 1.8% Italy 3,202,400 1.5% Spain 3,145,400 1.5% Canada 3,027,900 1.5% South Africa 1,191,100 0.6% Rest of Asia/Pacific 17,783,800 8.6% Rest of MEA 9,973,000 4.8% Rest of Western Europe 7,380,800 3.6% Rest of Latin America 6,777,500 3.3% Rest of Eastern Europe 4,661,300 2.2% Total 207,662,300 100% Notes: MEA = Middle East & Africa

33

Table 3. Unit sales of smartphones to end users in 4Q2012 for Top 10 vendors Vendor Sales to End Users (# of units) Market Share Origin Samsung 64,496,300 31.1% South Korea Apple 43,457,400 20.9% USA Huawei Technologies 8,666,400 4.2% China ZTE 8,310,200 4.0% China LG Electronics 8,038,800 3.9% South Korea Lenovo 7,904,200 3.8% China Research In Motion 7,333,000 3.5% Canada Nokia 7,094,300 3.4% Finland Sony Mobile Communications 7,069,400 3.4% Japan HTC 6,688,600 3.2% Taiwan Table 4. Descriptive statistics for variables in the regression analyses Mean S.D. min Max Country Share 0.05 0.18 0 1 PatentWarIntensity 50.37 68.90 0 251 IP_index 6.45 1.49 4 8.5 TechCapability 1314.35 2727.13 0 11707 WeakOrigin 0.55 0.50 0 1 LagCountrySalesShare 0.05 0.06 0.00 0.27 LagGDPshare (%) 5.07 4.39 0.70 20.94 LagHHI 0.32 0.17 0.10 1 LagNumFirms 13.53 5.46 1 38

34

Table 5. Shift of vendors’ sales towards countries with weak IP protection as the patent war intensifies

DV= country share of vendor's sale (1)

All vendors

(2) Not involved in litigation

(3) Involved

in litigation

(4) All vendors

(H1)

PatentWarIntensity * IP_index -0.0008** -0.0010** -0.0001 (0.0004) (0.0004) (0.0004) IP_index 0.2387 0.1978 0.2458 -0.4223 (0.4454) (0.5672) (0.2623) (0.5576) 3Q2008 * IP_index -0.0574* 4Q2008 * IP_index -0.0128 1Q2009 * IP_index 0.0438 2Q2009 * IP_index 0.0095 3Q2009 * IP_index 0.0021 4Q2009 * IP_index 0.0970 1Q2010 * IP_index 0.0545 2Q2010 * IP_index 0.1606 3Q2010 * IP_index -0.0093 4Q2010 * IP_index 0.0285 1Q2011 * IP_index -0.1660 2Q2011 * IP_index -0.1243 3Q2011 * IP_index -0.2466 4Q2011 * IP_index -0.3344** 1Q2012 * IP_index -0.3881** 2Q2012 * IP_index -0.3292* 3Q2012 * IP_index -0.3041* 4Q2012 * IP_index -0.3026

Control Variables LagCountrySalesShare 1.3264 -0.8691 9.5056*** 1.5563 (2.1319) (2.4899) (2.8358) (2.2220) LagGDPshare 0.2688 0.2572 0.1525 0.2471

(0.2964) (0.3775) (0.3517) (0.2835) LagHHI -0.4092 -1.0089 0.1061 0.3830 (0.6078) (0.8182) (0.4816) (0.7231) LagNumFirms 0.0475*** 0.0579*** -0.0320** 0.0496*** (0.0151) (0.0181) (0.0154) (0.0156)

Country dummies Yes Yes Yes Yes Quarter dummies Yes Yes Yes Yes Vendor Fixed Effects Yes Yes Yes Yes Observations 14,793 11,873 2,920 14,793 Notes: The dependent variable is each vendor’s sale in a given country-quarter divided by the vendor’s total sales in that quarter. Model (1) uses the entire sample, Model (2) and (3) split the sample by whether a vendor was involved in a patent litigation or not, and Model (4) uses quarter-IP index interaction terms to dissect the time trend and test which periods are affecting vendors’ firm strategy. Fractional logit models are used in all models. Heteroskedasticity-adjusted standard errors clustered at the vendor level in parentheses. Standard errors are omitted for the interactions between quarters and PatentWarIntensity in Model (4) for brevity. significant at 10%; ** significant at 5%; *** significant at 1%.

35

Table 6. The Moderating Effect of Vendors’ Technological Capabilities

DV=country share of vendor's sale

(1) Strong Tech

Capability

(2) Weak Tech

Capability

(3) All

vendors

(4) Not in

Litigation

(5) Involved

in litigation

(H2) PatentWarIntensity * IP_index 0.0003 -0.0011* -0.0034*** -0.0038*** 0.0014 (0.0004) (0.0006) (0.0008) (0.0010) (0.0018) IP_index -0.0039 0.5727 0.0850 0.0010 0.3153 (0.3326) (0.8318) (0.4285) (0.5525) (0.2785) PatentWarIntensity * TechCapability * IP_index 0.0009*** 0.0013*** -0.0002 (0.0002) (0.0004) (0.0002) TechCapability * IP_index 0.0057*** 0.0098*** -0.0091 (0.0022) (0.0033) (0.0059) PatentWarIntensity * TechCapability -0.0058*** -0.0088*** 0.0015 (0.0014) (0.0027) (0.0015) Control Variables LagCountrySalesShare 4.7985* -1.7125 0.9729 -1.2796 9.5226*** (2.5326) (3.2665) (2.1437) (2.5241) (2.8489) LagGDPshare 0.2161 0.1708 0.3129 0.3355 0.1538 (0.3340) (0.4726) (0.2830) (0.3622) (0.3531) LagHHI 0.1125 -0.6737 -0.2072 -0.6814 0.1066 (0.7194) (0.9577) (0.5659) (0.7620) (0.4847) LagNumFirms -0.0047 0.0853*** 0.0512*** 0.0614*** -0.0320** (0.0145) (0.0265) (0.0151) (0.0184) (0.0155) Country Fixed Effects Yes Yes Yes Yes Yes Quarter Fixed Effects Yes Yes Yes Yes Yes Vendor Fixed Effects Yes Yes Yes Yes Yes Observations 7,367 7,141 14,508 11,588 2,920 Notes: The dependent variable is each vendor’s sale in a given country-quarter divided by the vendor’s total sales in that quarter. Models (1) and (2) use a split-sample analysis that parallels model (3). Model (3) uses a three-way interaction term to test how vendors with various degrees of technological capabilities are affected as the patent war intensifies. Model (4) and (5) split the sample by whether a vendor was involved in a patent litigation or not and replicates the analysis in Model (3). Fractional logit models are used in all models. Heteroskedasticity-adjusted standard errors clustered at the vendor level in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1%.

36

Table 7. The Moderating Effect of Vendors’ Home Countries’ IP Systems

DV= country share of vendor's sale (1)

Weak Origin=1

(2) Weak

Origin =0

(3) All

Vendors

(4) Not in

litigation

(5) Involved

in litigation (H3) PatentWarIntensity * IP_index -0.0015** 0.0006 0.0025** 0.0042** -0.0003 (0.0007) (0.0004) (0.0010) (0.0021) (0.0008) IP_index 0.1543 -0.0410 0.1343 -0.0848 0.3239 (0.6457) (0.6998) (0.4300) (0.5516) (0.2975) PatentWarIntensity * WeakOrigin * IP_index -0.0046*** -0.0064*** 0.0005 (0.0011) (0.0021) (0.0009) WeakOrigin * IP_index -0.7492*** -1.0222*** -0.1434 (0.1500) (0.2190) (0.1744) PatentWarIntensity * WeakOrigin 0.0292*** 0.0436*** -0.0033 (0.0082) (0.0164) (0.0057) Control Variables LagCountrySalesShare -0.6601 3.9037 1.9024 0.2168 9.5418*** (3.7738) (2.6805) (2.1030) (2.6015) (2.8313) LagGDPshare 0.1212 0.3901 0.1717 0.1010 0.1530 (0.3348) (0.4843) (0.2670) (0.3539) (0.3535) LagHHI -0.6008 1.2743 0.0975 -0.1726 0.1068 (0.8065) (0.9898) (0.5987) (0.8633) (0.4851) LagNumFirms 0.0570*** 0.0182 0.0553*** 0.0669*** -0.0322** (0.0197) (0.0243) (0.0148) (0.0182) (0.0155) Country Fixed Effects Yes Yes Yes Yes Yes Quarter Fixed Effects Yes Yes Yes Yes Yes Vendor Fixed Effects Yes Yes Yes Yes Yes Observations 8,189 6,604 14,793 11,873 2,920 Notes: The dependent variable is each vendor’s sale in a given country-quarter divided by the vendor’s total sales in that quarter. Models (1) and (2) use a split-sample analysis that parallels model (3). Model (3) uses a three-way interaction term to test how vendors whose home country has a weak IP system are affected as the patent war intensifies. Model (4) and (5) split the sample by whether a vendor was involved in a patent litigation or not and replicates the analysis in Model (3). Heteroskedasticity-adjusted standard errors clustered at the vendor level in parentheses. significant at 10%; ** significant at 5%; *** significant at 1%.

37

Table 8. The consequence of vendor strategy as the patent war intensifies and vendor’s product release strategy (negative binomial regression) as a robustness check.

Notes: In Model (1), the dependent variable is the market share of Android-powered smartphones in a given country. The fractional logit model is used. In Models (2)-(4), the dependent variable is the number of product models released in a given country. Negative binomial models are used. Heteroskedasticity-adjusted standard errors in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1%. Standard errors are clustered at the country level for Model (1) and at the vendor level for Models (2)-(4).

(1) Android

market share

(2) Number of

product models

(3) Not

in litigation

(4) Involved

in litigation PatentWarIntensity * IP_index -0.0015*** -0.0002** -0.0004* -0.0002** (0.0003) (0.0001) (0.0002) (0.0001) IP_index 0.0320 0.2163*** 0.5940*** 0.1168* (0.0385) (0.0758) (0.1769) (0.0639)

Control Variables LagOSshare 3.0868***

(0.1268) LagGDPshare 0.0041 0.2025*** 0.2569** 0.1976*** (0.0130) (0.0541) (0.1081) (0.0487) LagHHI 0.1504 -0.1004 0.2500*** (0.1157) (0.2814) (0.0937) LagNumFirms 0.0285*** 0.0767*** -0.0159*** (0.0040) (0.0087) (0.0038) LagCountrySalesShare -0.0063 -4.2150*** 0.1177 (0.4447) (0.9044) (0.4179)

Country Fixed Effects Yes Yes Yes Yes Quarter Fixed Effects Yes Yes Yes Yes Vendor Fixed Effects No Yes Yes Yes Observations 285 27,770 20,983 6,787

38

Figure 1. Patent war intensity as measured by the frequency of media coverage of smartphone patent litigation.

0

50

100

150

200

250

300

number of news articles related to smartphone patent war

39

Figure 2a. Share of smartphone sales in strong and weak IP countries.

Figure 2b. Smartphone sales volumes in strong and weak IP countries.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Strong IP (%) Weak IP (%)

0

20000

40000

60000

80000

100000

120000

140000

Strong IP Weak IP

40

Figure 3. Android smartphone sales volumes in strong and weak IP countries.

Figure 4. Market shares of Android, iOS, and others in strong and weak IP countries.

0

20,000

40,000

60,000

80,000

100,000

120,000

Strong IP Weak IP

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Strong IP - Other Weak IP - Other Strong IP - Android

Weak IP - Android Strong IP - iOS Weak IP - iOS