Jurisdiction Shopping and Foreign Location Choice: The Role of Market and Nonmarket Experience in the European Solar Energy Industry

We investigate how industry-support policies targeted at specific sectors influence foreign investment location choices. We argue that firms engage in jurisdiction shopping, selecting to invest in countries with more generous policy support, but that this effect varies markedly across firms. Our theory suggests that firms’ nonmarket experience positively moderates the effect of policy support on location choice, while the moderating effect of market experience should be less impactful. We further argue that, depending on the nature of their nonmarket experience, firms may become less attracted to places with more generous policy support. We find support for our predictions by examining foreign investments of firms entering the solar energy industry in the EU during 2004-2013. We discuss implications of our study for international business, research on public policy and strategy, and the energy transition.

Although this body of research has generated impressive insights into the behavior and performance of international businesses, it has given less attention to two important features of the relationship between institutions and international business. First, the literature on institutions within international business research has often viewed formal institutions as having a homogenous impact on firms' international strategies, either because of researchers' explicit emphasis on isomorphism (Kostova, Roth, & Dacin, 2008), or because of their focus on more aggregate levels of analysis such as the country (Aguilera & Grøgaard, 2019). Second, as policies typically place constraints on firms' potential to extract rents from their operations, much of this research has focused on examining whether firms tend to avoid institutional constraints by investing and locating their operations in countries with lax regulatory standards (Cole, 2004;Kellenberg, 2009;Rezza, 2015;Rugman & Verbeke, 1998). However warranted, these two tendencies limit our potential to fully explain how policies supporting specific economic sectors can engender heterogeneous international strategies depending on firm-specific features.
Our study aims to fill this gap by leveraging the growing importance of institutional support for industries in the form of supportive policies targeted at specific industries, a trend that exposes firms to opportunities and challenges related to international expansion. Countries across the globe have recently provided substantial policy support to sectors such as green buildings, clean energy, and electric vehicles, in order to facilitate industry transitions (Georgallis, Dowell & Durand, 2019;Holtsmark & Skonhoft, 2014;Jones, York, Vedula, Conger, & Lenox, 2019). However, while evidence of how industry support policies stimulate the growth of their target sectors continues to accumulate (Cimoli, Dosi, & Stiglitz, 2009), we lack a good understanding of whether they affect firms' foreign investment decisions, and why some firms are more or less attracted to policy-supported industries in foreign markets.
To shed light on these questions, we examine how industry-support policies affect firms' foreign investment location choices. Conceptualizing policy support as a country-specific advantage (CSA) (Clarke, Tamaschke, & Liesch, 2013;Rugman, Verbeke, & Nguyen, 2011), we argue that firms will engage in 'jurisdiction shopping', whereby they search for and select to invest in countries with the most supportive policy frameworks. However, heterogeneity in firms' market and nonmarket experience will lead to markedly different responses to industry-support policies. We argue in particular that, because learning in unfamiliar environments tends to be domain-specific (Kim, Delios, & Xu, 2010;Oh & Oetzel, 2017), firms' nonmarket experience, conceptualized here as prior experience with similar policies, will positively moderate their tendency to engage in jurisdiction shopping while their market experience will have a less pronounced effect. Further, though nonmarket experience should generally enhance firms' attraction to foreign countries' policies, we expect that the nature of that experience-in particular, whether nonmarket experience is adverse-may lead some firms to prefer locations with a less generous policy support. We test and find support for our predictions by analyzing foreign location choices in a setting where policy support was paramount: the solar energy industry in European Union countries over the period 2004-2013. We contribute to international business research by examining why firms differ in the degree to which they engage in jurisdiction shopping when selecting among potential locations for foreign investments. Specifically, we add to research on institutions in international business by showing how supportive formal institutions heterogeneously affect firms' international strategies. We also contribute to research on the FSA/CSA framework by highlighting that country-specific advantages are not only viewed differently by firms depending on their firm-specific experience (Chen, Li, & Shapiro, 2015;Clarke, Tamaschke, & Liesch, 2013), but also that some firms may view the same host country's supportive policies as a disadvantage rather than an advantage. Moreover, we add to the strategy and international business literatures by disaggregating experience into component parts, addressing recent calls for a more nuanced understanding of how context interacts with firm attributes, including firm experience, in shaping international strategy (Oh & Oetzel 2017;Liu & Vrontis, 2017;Fremeth & Shaver, 2014). Finally, our findings regarding jurisdiction shopping and firm experience provide insights relevant for practitioners in the renewable energy sector, and inform current debates on the role of industry-support policies in fostering the energy transition. We elaborate on these contributions and on the implications for the energy transition in the discussion section.

THEORETICAL BACKGROUND AND HYPOTHESES' DEVELOPMENT
A firm's decision about where to locate international investments is one of the most central strategic choices in international business, as such a choice affects most other cross-border economic transactions (Banalieva & Dhanaraj, 2013;Belderbos, Olffen, & Zou, 2011;Nachum, Zaheer, & Gross, 2008). Prior research has identified the institutional context as a critical determinant of international business activity in general and location choices in particular (Albino-Pimentel, Dussauge, & Shaver, 2018;Banalieva & Dhanaraj, 2013;Blonigen, 2005;Henisz, 2000;Maggioni et al., 2019). For example, prior literature has shown that the imprinting effect of firms' home-country institutional environment can affect firms' ability to adjust their operating knowledge strategies across borders (Kriauciunas & Kale, 2006). Furthermore, in the case of international expansion, prior work has revealed that firms' sensitivity to host country regulations depends on their home country's institutional setting (Albino-Pimentel et al., 2018;Cuervo-Cazurra & Genc, 2008;Holburn & Zelner, 2010). More generally, the internationalization literature has discussed how home and host country institutional environments allow firms to develop knowledge and experience, or more generally, resources and capabilities, that they can later translate into firm-specific advantages (FSAs) when making subsequent foreign market entry decisions (Delios & Henisz, 2003a, 2003bDunning & Lundan, 2008;Rugman et al., 2011). All in all, this research has established that the institutional environment, both in local and international markets, has a major impact on firms' international location choices.
Institutional contexts have been analyzed using a variety of perspectives, including institutional economics (North, 1990), sociological institutionalism (Scott, 2013), and comparative institutionalism (Hall & Soskice, 2001), among others (see Aguilera & Grøgaard, 2019 for a review). In referring to institutional context, this study relies on North's perspective on institutions for two reasons. First, as possibly the "most widely adopted strand of institutional theory in IB" (Aguilera & Grøgaard, 2019), this perspective allows us to position our study within-and contribute to-the broader literature on internationalization and location choice strategies, which has examined the role of institutions as determinants of foreign investment. Second, the strong emphasis that the new institutional economics has placed on formal institutions (Aguilera & Grøgaard, 2019;Albino-Pimentel et al., 2018;Coeurderoy & Murray, 2008;Hotho & Pedersen, 2012), and on the policy environment in particular (Dorobantu, Kaul, & Zelner, 2017;García-Canal & Guillén, 2008;Henisz, 2000) is consistent with our primary interest in how firms react to public policy. Following North's conceptualization, we view institutions as the "rules of the game" that organizations need to abide by and that provide order and the incentive structure for an economy. Though institutions can be informal, derived from customs, norms, or cultural values, the regulatory dimension of institutions is of utmost importance because policies clearly define choice sets that determine "the profitability and feasibility of engaging in economic activity" (North, 1991: 97). As a result, policy frameworks have been central not only to economic institutionalists interested in the economic performance of states (North, 1991), but also to international business scholars interested in cross-border business activity (Henisz, 2000;Delios & Henisz, 2003a;2003b;Albino-Pimentel et al., 2018).
However, two features of this research have so far inhibited its potential to fully explain why firms react differently to the same formal institutions, and its applicability to the context of industrysupport policies, such as policies enacted to facilitate industry transitions. First, most research in international business which uses institutional theory relies on the assumption that firms are homogeneously affected by institutions (Aguilera & Grøgaard, 2019;Hung, 2005). Indeed, despite the many useful insights on how institutions affect international business, research in this area has largely focused on the country or industry as the primary levels of analysis (Aguilera & Grøgaard, 2019;Hotho & Pedersen, 2012), and has thus been insufficient to address how the same institutional environment elicits heterogeneous responses by firms making international business decisions. This is unfortunate, as recent evidence suggests that firms do react differently to the same formal institutions (Bu & Wagner, 2016), and to industry support policies specifically (Georgallis & Durand, 2017).
Second, much of the research that studies the regulatory aspect of institutions views government policy as a constraint on firms' behavior. As a result, these studies view firms as making their location choices in a way that allows them to avoid such constraints. Prior research has found, for example, that polluting firms prefer to avoid countries with stringent environmental regulations, and that multinationals often avoid countries with strong employment protection legislation (Dowell et al., 2000;Olney, 2013).
Building on North's neo-institutional economics approach, Ahuja and Yayavaram (2011) suggest that this is a result of firms engaging in jurisdiction shopping, a process whereby each firm assesses the institutional conditions at different geographic locations and selects to avoid those with the most stringent regulation. Accordingly, the idea of jurisdiction shopping has been recently utilized in the international business literature to explain patterns of location choice across countries (Brunetta, Giustiniano, & Boccardelli, 2017;Findlay, 2014;Tallman, Luo, & Buckley, 2018). However, the emphasis of research on jurisdiction shopping on regulation as a constraint and jurisdiction shopping as an avoidance strategy may be too narrow. Even though policies and formal institutions are generally restrictive, policy-makers can also manipulate the institutional context to stimulate firms' investment decisions (Lazzarini, 2015), by using policy incentives to induce entrepreneurship, attract technologically advanced businesses, or enable industry transitions (Georgallis, Dowell & Durand, 2018;Mesquita & Lazzarini, 2008;Jones et al., 2019).
Below we take a step towards addressing these oversights by examining how formal institutions, notably industry-support policies for specific industries, differentially affect firms' foreign location choices. Our theory relies on two fundamental assumptions. First, jurisdiction shopping is not merely an avoidance strategy as implied or explicitly stated in some prior research (Ahuja & Yayavaram, 2011;Brunetta et al., 2017;Findlay, 2014;Tallman et al., 2018), but a more general process by which firms select between alternative foreign locations. Second, not all firms engage in jurisdiction shopping to the same extent. Rather, firms are heterogeneous in their ability to scan and understand the policy environment of foreign countries, and some firms may even view the same policies differently depending on their prior experience.
To develop our theory, we draw from the international business literature on country-specific advantages and firm-specific advantages, namely the CSA/FSA framework (Hillemann & Gestrin, 2016;Rugman & Verbeke, 1998;Rugman et al., 2011;Verbeke & Kano, 2012) and from studies on the influence of experience on international strategy (Chen et al., 2015;Clarke et al., 2013). According to the CSA/FSA framework, internationalization decisions are largely based on firm-specific advantages (FSAs) and country-specific advantages (CSAs). CSAs refer to advantages presented by the foreign countries where the firm plans to operate, such as better access to resources, industrial clusters, or a growing market.
FSAs refer to firm-specific own assets or competences such as innovation capacity, brand name, knowledge and experience. We argue that institutional support can be conceptualized as a CSA that is, however, viewed differently by firms due to their own FSAs stemming from their firm-specific experience. To more clearly ground our conceptualization we briefly introduce our empirical context, the European solar energy industry, before developing specific hypotheses.

Policy Support and the Solar Energy Industry
The development of the solar energy industry is a prominent example of how institutional support in the form of favorable policies can stimulate investment in new economic sectors. Over the last two decades, governments have intensified their efforts to address climate change, in part by promoting alternative energy. For instance, as early as in the mid-1990s the German and Japanese governments embarked on demonstration projects which disseminated information on solar energy technology, aiming to establish the German and Japanese industry, respectively, as global market leaders (Hoffmann, Pietruszko, & Viaud, 2004). The centerpiece of these policy frameworks was the introduction of feed-intariff policies, economic instruments through which governments guarantee the purchase of clean power that is fed into the electricity grid at a price (tariff) that is above the actual price of electricity. Soon after, Italy and Spain also introduced feed-in-tariffs for solar energy in order to contribute to "environmental protection, industrial policy, and employment creation" (del Rio & Mir-Artigues, 2014). Feed-in-tariffs quickly spread across the world in the first decade of this century, making it the most popular policy measure for spearheading the energy transition, and leading to an unprecedented policy-induced growth of the solar power industry, especially in Europe (Georgallis, Dowell, & Durand, 2018;Hoppmann, Peters, Schneider, & Hoffmann, 2013). It was not long before firms started to enter foreign countries in order to reap the benefits of feed-in-tariff policies, making the solar power industry an ideal setting to understand how industry-support policies shape foreign investment, and what types of firms they tend to attract.

Industrial Policies and International Location Choice
Extensive research examines how industrial policies stimulate the development of new industries, create powerful companies, or promote innovation and local employment (Lazzarini, 2015;Spencer, Murtha, & Lenway, 2005). However, unless a policy is strictly defined to benefit local firms, these benefits can spill over outside the country's borders. Policies that favor a specific sector of economic activity are likely to attract foreign firms that operate or consider operating in that sector. In other words, such policies become CSAs for firms considering different locations for international investments. For example, feed-in-tariff policies have led to increased interest in and remarkable growth of several EU countries' markets Hoppmann et al., 2013), but the question of who would be serving those markets was 'up for grabs'. Foreign firms could step in and take advantage of the favorable regulatory environment. Indeed, some critics of European countries' renewable energy policies have argued that public money is being spent inefficiently, as deployment policies sometimes benefit foreign firms instead of the local economy (Peters, Schneider, Griesshaber, & Hoffmann, 2012).
For a firm considering international investment, policy support adds a novel problem to a familiar question: what is the best possible location to invest in? International market selection requires a screening and comparative assessment of different locations and the opportunities that each of them have to offer (Maitland & Sammartino, 2015a;Papadopoulos & Martín, 2011). Firms may be attracted to a certain location for a variety of reasons. In particular, a rich body of international business literature has identified a set of country attributes that can be conceived of by firms as CSAs, such as market size, availability of key resources, investments by competitors, or country similarities that facilitate business interactions (Basuil & Datta, 2019;Blonigen, 2005;Hillemann & Gestrin, 2016;Verbeke & Kano, 2012;Yaprak, Yosun, & Cetindamar, 2018).
In addition to these factors, policy tends to be a prominent feature of the foreign environment that firms assess in order to select the best possible location. For example, prior research has found that heavily polluting firms attempt to save up on costs of cleanup or fines by locating their operations in countries with lax environmental regulation (Dowell et al., 2000), and that multinationals often evade public policy controls by selecting to invest in countries with weak safety or employment protection legislation (Ahuja & Yayavaram, 2011;Olney, 2013). Some have argued that this can lead to a 'race to the bottom' as countries competitively undercut their ethical or environmental standards to attract foreign investment (Olney, 2013). According to Ahuja and Yayavaram (2011), this is the result of firms engaging in jurisdiction shopping, whereby managers of each firm scan the institutional conditions at different geographic locations, assess the constraints that they pose on the firm, and then make their international location choices in such a way as to avoid countries with the most stringent regulation.
Extending their argument, we suggest that jurisdiction shopping is not merely an avoidance strategy, but a more general process by which firms select between foreign locations. As such, regulation does not always induce a race to the bottom; it can also induce a 'race to the top' (Bu & Wagner, 2016) when stimulating policy instruments offer rent-generating opportunities for foreign firms. In line with this idea, we argue that supportive policies constitute CSAs, as they render investment in a specific country a more attractive proposition for foreign entrants. Ceteris paribus, the more generous the policy incentive, the more attractive it will be to do business in that country. Thus, if firms do engage in jurisdiction shopping, we should expect a firm to survey the policy conditions in different countries and select to invest in the country with the most generous policy support. For example, electronics companies are lured to Southeast-Asian countries such as China and India which offer preferential loans and grants to attract investment (Bloomberg, 2015;Capital Trade Incorporated, 2009). Also, countries compete for investment in e-mobility by offering subsidies for electric vehicles (Zgut, Zbytniewska, Hosnedlová, & Szalai, 2019), and, in the solar energy setting, firms are incentivized to invest in countries with more generous feed-intariff policies (Georgallis & Durand, 2017). More generally, conditional on the choice to invest in a foreign market, firms should exhibit a preference to locate in countries with more generous industrysupport policies. Formally, we expect that: H1: The more generous the industry-support policy in a country, the greater the likelihood that a firm will choose to invest in that country.

The Moderating Role of Firm Experience
As discussed above, international business research using institutional theory has often focused on cross-country institutional differences (Aguilera & Grøgaard, 2019;Hotho & Pedersen, 2012) at the expense of examining how individual firm characteristics condition their strategies regarding countrylevel formal institutions (Bu & Wagner, 2016). Similarly, most research on the effectiveness of industrial policy has investigated aggregate levels of analysis rather than firm-specific characteristics (Georgallis & Durand, 2017;Hoppmann et al., 2013). However, firms are heterogeneous in their ability to conceive or capture the benefits associated with a specific institutional environment (Ghemawat, 2001;Nachum et al., 2008). In particular, different types of firm-specific advantages (FSAs) may not only make firms more or less able to deal with certain institutional environments, but may even shape how a country characteristic is conceived of, as either a CSA or a country disadvantage that may render international operations difficult or too risky (Basuil & Datta, 2019). One such FSA that is likely to constitute a source of variation in firms' strategies and moderate their responses to industrial policy is the level and the nature of their prior experience (Clarke et al., 2013).
Firm experience, defined as the practical knowledge that firms derive from participation in a particular activity, market, or country (cf. Chang and Rosenzweig, 2001;Oh and Oetzel, 2017), has frequently been shown to affect international investment strategies (Belderbos et al., 2011;Delios & Henisz, 2003a, 2003bMaitland & Sammartino, 2015b;Oh & Oetzel, 2017). Experience is an important FSA because it implies learning, or the acquisition of knowledge, which can later be leveraged as the firm encounters the choice of whether to invest in other locations, or the choice of where to invest (Buckley & Casson, 1985;Johanson & Vahlne, 2009;Oh & Oetzel, 2017). Furthermore, acquired experience allows firms to apply their knowledge in new contexts and thus affects the expected benefits of entering a particular location.
The significance of experience stems from managers' bounded rationality regarding strategic choices, including international investment location choices. Under a rational actor model, decisionmakers have full information availability and unrestricted capacity to assess the costs and benefits of international expansion. However, under the assumption of bounded rationality, which has been identified as a more realistic basis for interpreting international business decisions (Kano & Verbeke, 2019), there are cognitive constraints on managers' ability to scan the environment and identify all information relevant to the decision, as well as on their capacity to make sense of that information in order to assess whether entry in a given location is warranted (Maitland & Sammartino, 2015a;Papadopoulos & Martin, 2011). This can lead to significant inter-firm differences because not all firms have faced similar situations in the past. Firms with substantial experience are better able to identify "what information to look for, how to analyze it, and its implications for the firm's expansion" (Maitland & Sammartino, 2015a: 738), whereas inexperienced firms do not hold this type of FSA when expanding abroad and have limited ability to assess target countries.
Indeed, prior research has shown that firm experience is an important determinant of international location choices (Delios & Henisz, 2003a;Hernandez, 2014;Oh & Oetzel, 2017). For instance, experience acquired in the firms' home country has been shown to influence how firms assess political uncertainty in candidate foreign locations (Cuervo-Cazurra & Genc, 2008;García-Canal & Guillén, 2008;Holburn & Zelner, 2010), and experience acquired from investments made in other host countries has influenced firms' subsequent location choices (Delios & Henisz, 2003a, 2003b. Overall, international experience is critical for firms' international investment location choices. However, whereas most prior research has used experience as a general property of the firm that applies broadly to international expansion, recent findings suggest that the learning associated with experience tends to be domain-specific rather than general in nature. For instance, Kim, Delios, & Xu (2010) suggest that learning is most effective when prior experience is specific to the product market or industry, but not in other domains. Similarly, a study by Oh and Oetzel (2017) shows that a firm's potential to leverage its experience with political risk across borders is specific to the type of risk involved. Thus, it is important to consider in which domain experience has been accumulated because that may affect the specific benefits that firms gain from that experience. Below, we argue that firms' market and nonmarket experience can be conceived of as FSAs, which differentially affect their propensity to select countries with more generous policies, i.e. countries with a greater level of the CSA that is the focus of this study. In other words, we expect firms' market and nonmarket experience to moderate the effect of industry-support policies on location choice.

Market and nonmarket experience:
To assess whether conditions are favorable for expansion, firms need to survey both the market and nonmarket environment of potential host countries. The market environment of business activity refers to interactions between firms and market stakeholders such as suppliers, buyers, and competitors (Baron, 2003). Market experience, i.e. experience gained through interactions with market stakeholders, increases with both the length of a firm's stay in a particular country and with the breadth of its operations (i.e. the number of countries) (Maitland & Sammartino, 2015a). The underlying knowledge acquired through market experience can include, for example, a better understanding of consumer preferences, greater knowledge of the market dynamics, or the ability to draft more favorable contracts with suppliers. As prior research shows, market experience is relevant for foreign expansion decisions when it is specific to the market or industry for which the firm is considering foreign expansion because of the similarity of activities that a firm needs to perform across different countries (Kim et al., 2010). Thus, market experience within the solar industry should moderate the influence of market attraction features on a firm's likelihood to invest in a particular country in the same industry. Specifically, having experience investing in the solar industry should influence how sensitive a firm is to industry-support policies when deciding whether to invest in a foreign location, because firms with more market experience are better able to appreciate the relevance of the policy environment for their performance compared to competitors in the same industry.
Nevertheless, we do not expect market experience to have the strongest moderating impact on the relationship between policy support in a potential host country and the firm's location choice. Market experience does not necessarily imply the specific ability to scan and understand elements of the nonmarket environment (such as public policy), as these capabilities may rest on nonmarket experience.
The nonmarket environment includes interactions with nonmarket stakeholders such as governments, regulators, or civil society (Baron, 2003); accordingly, nonmarket experience reflects the firm's experience with nonmarket stakeholders. 1 Just as firms can develop market experience from exposure to business interactions in the home country or in other host countries, so they can develop nonmarket experience from exposure to policies across their domestic and international operations. For example, Conergy used its experience in the German and French markets as a stepping stone to invest subsequently in markets such as Italy, Czech Republic, and Greece -countries that had, at the time, some of the most attractive feed-in-tariffs in Europe. Their nonmarket experience was likely critical in triggering these location choices, as it offers domain-specific knowledge that allows the company to navigate the policy environment (e.g. routines to swiftly secure installation approvals, grid connection licenses, or environmental permits), and the necessary know-how to assess the potential benefits of the policy. Along these lines, prior work has shown that firms in regulated industries enter countries characterized by similar types of firm-regulator interactions, because the experience with governments that their home-business entails and the specific knowledge that they acquire as a result allow them to better do business in similar foreign countries (Cuervo-Cazurra & Genc, 2008;García-Canal & Guillén, 2008). Prior studies also suggest that firms can gain nonmarket experience not only from their home country's but also from their foreign operations (Albino-Pimentel et al., 2018;Delios & Henisz, 2003a).
In sum, the influence of experience on firm decisions tends to be domain-specific in that it depends on the type of the experience gained (Maitland & Sammartino, 2015a, 2015bOh & Oetzel, 2017). This is important for our purposes because market and nonmarket experience are associated with fundamentally different types of knowledge. A firm's ability to assess supportive policies in potential host countries is likely to stem not from general experience with the market, but rather from domain-specific experience with the nonmarket environment. Specifically, firms that have gained experience with similar policies (in their home country or in other host countries where they operate) develop knowledge that equips them with greater ability to evaluate the implications of such policies for their business and, in turn, render them more responsive to similar policies in foreign countries. Thus, nonmarket experience should positively moderate the impact of industry-support policies on location choice. Having gained a better understanding of supportive policy from prior experience, a firm is likely to incorporate the policy environment more centrally in its evaluations of different locations. Moreover, firms with relevant nonmarket experience may also have developed better routines to apply for and obtain the benefits from industrial policies in a timely manner, which increases the perceived rent-generating capacity that the firm can expect from investing in a location with such a CSA. Thus, firms' propensity to engage in jurisdiction shopping may rely more on nonmarket, rather than market experience. Formally, we put forth that: H2: Nonmarket experience positively moderates the impact of industry-support policy on the likelihood that a firm will choose to invest in a particular country.
H3: The moderating impact of nonmarket experience on the relationship between industry-support policy and location choice is stronger than the moderating impact of market experience.
Adverse Experience: While the previous section established the importance of domain-specific experience, it is also important to consider the nature of that experience. In addition to the existence of a supportive policy environment in the host country, changes in the policy conditions may also trigger a search for opportunities to invest abroad and influence location choice. Despite the opportunities afforded by supportive policies, dependence on these policies also creates uncertainty for prospective entrants (Dutt & Joseph, 2019;Georgallis & Durand, 2017). The case of solar energy policies illustrates this point.
While policies such as feed-in-tariffs were highly successful in bringing solar technology to the market, in some cases they were 'too successful' in attracting investment: the number of investments escalated the costs of keeping the policies in place (del Rio & Mir-Artigues, 2014). This trend, along with the decreased costs of solar technology, led some European governments to lessen or even abolish their support for the industry. Such changes inevitably created an adverse experience for some firms, potentially affecting whether they engage in jurisdiction shopping for future investment locations. Adverse experience is particularly relevant for firms considering investments in countries with similar policy incentive schemes.
Prior work would suggest that firms facing a more adverse nonmarket environment are likely to seek opportunities in other countries. For example, firms are more likely to enter another geographic region when they face extreme rivalry in the places where they operate (Fuentelsaz, Gomez, & Polo, 2002) or when they need to 'escape' adverse institutional constraints (Witt & Lewin, 2007). Anecdotal evidence from the solar energy industry supports this idea; for instance, feed-in-tariff cuts in Spain were characterized by industry insiders as "chaotic" (García-Castrillón, 2016), and several companies that had experienced them-such as Spanish solar company FPV and the Italian OPDE group-started to turn their attention to neighboring countries (Price, 2011). Generally, firms that have recently experienced policy adversity, such as a repeal or substantial reduction in the level of policy support that is offered in their home or host countries, are likely to seek opportunities elsewhere.
Yet, we expect the pattern of location choice that these firms exhibit to be quite different from that of other firms. In general, firms prefer to invest in locations where the government is committed to the policies and rules it initiates (Henisz, 2000;Holburn & Zelner, 2010;Murtha & Lenway, 1994;Spencer et al., 2005). Being exposed to detrimental policy changes may affect firms' perspective on the stability and credibility of the policies, and, in turn, their future investments. Just as individuals avoid risky situations in the aftermath of negative experience (Sitkin & Weingart, 1995), so can firms develop a disbelief and even avoidance towards policies that they associate with adverse conditions, even if those policies are generally supportive. We posit that as firms experience adversity due to discontinuous changes in the nonmarket (policy) environment, they become less attracted to those same policies in other foreign markets. They may become especially skeptical of those countries that offer very generous policy support, considering those institutional conditions unsustainable and therefore risky. In other words, we argue that firms with greater adverse experience may conceive of industry-supportive policies not as a CSA but as a country-specific disadvantage that deters them from investing in countries with more generous policies.
We thus expect that, contrary to the general pattern predicted above, firms with greater adverse nonmarket experience will be less attracted to the countries with the most generous policy support when making their international location choices. Therefore, we propose the following moderating effect: H4: Adverse nonmarket experience negatively moderates the impact of industry-support policy on the likelihood that a firm will choose to invest in a particular country.

DATA AND METHOD
We test our hypotheses on a sample of foreign greenfield investments in the solar energy industry carried out by firms originating from one of the current members of the European Union (EU) and into other EU countries during the 2004-2013 period. We chose to focus on investments within the EU for two reasons. First, due to availability of high-quality data on policies and on firm market, nonmarket and adverse experience. Because most EU governments relied on the same type of policy to promote the solar energy industry, the feed-in-tariff (FiT), this setting offers a "natural laboratory" with "comparable crosscountry data" on FiT policies (Georgallis et al., 2018: 16), enabling us to measure and study the effects of firms' nonmarket experience on their international strategies. Furthermore, several EU countries have relied on FiT policies for a long period of time, which helps us ensure that firms accumulate different types of nonmarket experience, including adverse experience, associated with such policies. Finally, and importantly, even though most EU countries used FiT policies to support solar energy, FiT policies varied markedly in terms of their generosity across both time and national contexts. Thus, the EU solar energy industry setting offers variation on our main independent variables, which capture the generosity of potential host countries' policies and the experience firms have gained from their prior exposure to this aspect of the nonmarket environment.
Second, beyond the benefits for our research design, this setting allows us to test our theory in a context with substantial practical relevance. Our observation period, which marks the proliferation of FiT policies across the EU , was characterized by exponential market growth. The size of the EU solar energy market grew by more than 4,000 percent from 2004 to 2013, reflecting the welldocumented importance of FiT policies for this setting (Georgallis & Durand, 2017;Hoppmann et al., 2013;Peters et al., 2012). 2 Moreover, the European market was the leading region for solar photovoltaic panels during our observation period in terms of cumulative installed capacity, and accounted for more than half of the worlds' annual installations during our entire observation period except the last year, when other markets such as China started to emerge as key players (Lins, Williamson, Leitner, & Teske, 2014).
In sum, largely driven by FiT policies, the EU was historically the main market for solar energy. 3 Furthermore, this industry is key for mitigating climate change, one of the most urgent challenges of our times (Jones et al., 2019), and realizing the energy transition from brown to green energy has been one of the main motivations for the deployment of FiT policies (Gawel et al., 2014).
We use data on greenfield investments from the Financial Times (FT) fDiMarkets database, which includes information on greenfield projects by firms across all countries in the globe and has been used in a series of recent research papers in management and international business (e.g. Albino-Pimentel, Dussauge, & Shaver, 2018;Duanmu, 2014). Our sample includes 202 greenfield investments in the solar energy industry carried out by 89 firms and corresponding to 5,454 investment choices in EU countries over the 2004-2013 period. Unless otherwise stated, all predictor variables were measured one year prior to the focal investment year to avoid issues related to simultaneity or reverse causality.

Dependent Variable
Our dependent variable, location choice, is binary and equals 1 if a firm chooses a specific host country for an investment in the focal year, and 0 otherwise. We considered all 28 current members of the EU as potential targets for an investment. Accordingly, for each investment carried out, we created a set of 27 potential foreign locations (all EU member countries except the firm's home country). That is, for each investment in our data, we built a choice set of all other EU countries in which the focal firm did not make an investment, and assigned a value of 0 for the dependent variable in these cases. Following this approach, each observation reflects a possible location choice based on the 202 investments in our dataset.
After accounting for missing data, our final sample includes 5,451 observations.

Independent Variables
To test H1, we created the variable host-country FiT policy generosity, a continuous variable that measures the average price (in €/KWh) of the FiT offered by a potential host country. This variable captures variation in the generosity of the policy across potential host countries; the highest the FiT price, the more attractive the policy (Georgallis & Durand, 2017). As discussed above, we focused on feed-intariffs because prior research has documented that they were the key policy instrument used to generate a market in this setting (Georgallis & Durand, 2017;Hoppmann et al., 2013;Peters et al., 2012), and their prevalence offered the opportunity to use comparable and high quality data across countries . Based on our prediction in H1, we expect that higher levels of host country FiT policy generosity should positively affect a firm's choice of location for an investment. Data on FiT policies were collected and then cross-referenced from a variety of sources, including industry reports, solar industry trade press, the International Energy Agency's Policies and Measures database, and the RES LEGAL Europe website, which covers legislation on support schemes for renewable energy sources.

Moderating Variables
We created three variables to test the moderating hypotheses. First, to test H2 we created the variable firm nonmarket experience, a count variable that measures the number of years that a firm has been exposed to FiT policies across all its investments in the solar energy industry, including in its homecountry as well as in foreign countries. For each country where the firm had already invested in solar we counted the number of years when there was a FiT. Then, we added these up across all countries to capture a firm's overall experience with the same type of policy (a FiT policy). We rely on the interaction between host country FiT policy generosity and firm nonmarket experience to test the moderating role of nonmarket experience on the relationship between supportive policy and location choice (H2). We expect that the effect of host country FiT policy generosity will be higher for firms with greater firm nonmarket experience.
Second, to test H3 we created the variable firm market experience, a count variable that equals the number of years a firm has accumulated in solar energy across all its investments. We first counted the number of years a firm has operated in each country where it was active in the solar energy industry, including the firm's home country. Then, we summed these up across all countries. We assume that by investing for a longer time in a larger number of countries in a given industry, firms develop experience in dealing with market stakeholders, including competitors, suppliers and buyers, in that industry. Based on H3, we expect the interaction of firm market experience and host country FiT policy generosity to have a weaker effect on location choice compared to the interaction of firm nonmarket experience and host country FiT policy generosity.
Finally, to test H4 we created the variable firm adverse experience, a count variable that measures the number of times a firm has been exposed to a significant negative change in FiT, i.e., either a repeal or a substantial reduction of the FiT (Georgallis & Durand, 2017), across all countries in which the firm operated. We consider that such adverse experience should affect firms' sensitivity to policy support in a different way than simple nonmarket experience. Based on H4, we expect the effect of host country FiT policy generosity to be negatively moderated for firms with greater firm adverse experience.

Control Variables
We controlled for several factors that could influence the choice of a host country in our setting.
We included the variable peer investments, the cumulative number of investments in the solar industry by firms other than the focal firm in the potential host country before the focal investment. This variable captures potential bandwagon or isomorphic effects on a firm's location choice (Vedula & Matusik, 2017), as well as the effects of competition (Hawk, Pacheco-De-Almeida, & Yeung, 2013). We used data from FT fDiMarkets database to calculate this variable. The variable inward FDI (the ratio between inward FDI and the country's GDP in a given year) controls for the overall attractiveness of a potential host country. We used data from the World Bank World Development Indicators to compute this ratio.
Following prior research (Berry, Guillén, & Zhou, 2010;Nachum et al., 2008), we controlled for the variable geographic distance, the great circle distance in kilometers between the largest city in the home country and the largest city in the potential host country. Data for this variable were obtained from the Centre d'Etudes Prospectives et d'Informations Internationales (CEPII). To account for the institutional distance between a firm's home country and a potential host country, we first classified each member country of the EU on whether their economic system is highly liberal or highly coordinated, or neither (Jackson & Deeg, 2008;Vasudeva, 2009). We then created the variable institutional distance, which was assigned the value of 1 when a firm's home country's type of economic system differs from that of a potential host country, and 0 otherwise. We also included a set of dummy variables identifying countries with specific institutional features that might affect their attractiveness as location for investments in the solar energy industry. The dummy PIGS receives the value 1 when the potential host country is either Portugal, or Italy, or Greece, or Spain, countries where institutions are considerably weaker than other EU countries (Kickert, 2011). The dummy variable EE country is assigned the value of 1 when the potential host country is in Eastern Europe, a set of countries that is more likely to differ institutionally relative to the home-countries of the major investors from the EU (Bevan, Estrin, & Meyer, 2004;Dikova & Van Witteloostuijn, 2007). Finally, we included the dummy variables high corporatism and high liberalism, which are 1 when the potential host country is characterized by high corporatism or high liberalism, respectively, and 0 otherwise.
In addition, we controlled for several additional host country features that can affect a firm's location choice. The variable host country political uncertainty captures how easy it is for top political authorities in a host country to arbitrarily change policies. We used a continuous variable, the reverse of the political constraints measure drawn from the POLCONV dataset, to compute this variable (Henisz, 2000). The variable host country's ease of doing business measures the lack of barriers to opening a new business in a given host country and year. Data for this variable was drawn from the Heritage Foundation Economic Freedom index.
Because countries differ in available solar resources and in how developed the solar energy market is, we included three variables specific to this setting. First, as a firm's location choice might be driven by local market-seeking goals, we controlled for solar energy growth using the four-year compound annual growth rate in demand for solar cells in a given host country Henderson & Cool, 2003). Second, to control for the presence of a strong upstream production cluster that could affect a firm's location decision (Grøgaard, Gioia, & Benito, 2013), we included total solar cell production, a variable capturing the total combined production (in MW) of all solar photovoltaic cell producers in a potential host country in a given year. Third, to further capture potential resourceseeking drivers of a firm's location choice, we included the variable potential solar energy, the total potential solar energy that can be generated in a given host country per year (in MWh/year).
The variable economic growth captures the change in a given host country's gross domestic product (in billion USD) in a given year. Countries exhibiting higher economic growth are more attractive to foreign investors as they show better prospects for business expansion (Blonigen, 2005). We used data from the World Bank to measure this variable. Moreover, the variable environmental values captures the extent to which citizens in a potential host country care about the environment and thus potential latent preferences for the solar market. This variable is based on a survey question from the World Values Survey, which inquires citizens on whether they agree with giving part of their income for the environment. We used the percentage of respondents answering 'Agree' to obtain our measure (Jones, Malesios, & Botetzagias, 2009) and extrapolated missing values for the years when a survey was not issued (York & Lenox, 2014).
Finally, we controlled for other policies that might also provide support for investments in the solar industry in a potential host country and can act as partial substitutes for FiT policies. We measured other solar energy regulations as a combination of tax incentives, direct financial subsidies, quota obligations, and tender systems. Because there was little variation across each of these types of policies, we followed recent research  and grouped them together as a binary variable that takes a value of 1 when at least one of these policies is in place at a given host country and year, and 0 otherwise.

Empirical Approach
Our empirical analysis is designed to assess how FiT policies and firm nonmarket, market and adverse experience affect location choice. Our unit of analysis is thus the firm-investment-year, so that each investment made by a firm in a given year is compared to all EU country options that the firm could have chosen as a location instead. We followed prior literature on firm international location choice and used the conditional logit approach to model this choice (Alcacer & Chung, 2007;McFadden, 1974;Nachum et al., 2008;Shaver & Flyer, 2000).
Since location choice depends on both the attributes of the candidate countries a firm could choose to locate an investment, as well as on firm-specific characteristics (such as FSAs that can be used abroad), the conditional logit approach is advantageous as it examines firms choice among candidate locations conditional on the choice to make an investment abroad (Belderbos, Olffen, & Zou, 2011;Nachum et al., 2008). It has thus been the preferred model in location choice research because conditioning on the investment choice offers two benefits. First, it allows the researcher to focus precisely on how host country attributes, such as policy conditions in our case, affect the location decision. Second, the set-up of the model is conceptually equivalent to a model that includes both firm and firm-year (investment) fixed effects, so it does not require controlling for firm attributes because those do not vary across potential locations in the same year (Belderbos et al., 2011). The model allows us to indirectly control for unobserved firm attributes and focus on those firm-specific characteristics that we use in our theorizing through interactions between these firm characteristics and characteristics of the potential location. Thus, consistent with prior research (e.g. Nachum et al., 2008), we used interactions between different types of firm experience and the host country's FiT policy generosity to test the moderating hypotheses regarding these firm-level variables. Table 1 shows descriptive statistics and pairwise correlations between the variables included in our analysis. The table reveals that several variables in our models exhibit substantial variance across the observations in our analysis. For instance, the variable host country FiT policy generosity varies from 0 to 0.67 with an average of 0.21, suggesting that policy support for the solar energy industry is indeed a factor that differentiates countries in the EU as potential targets for investments during our period of analysis.

RESULTS
We also observe substantial firm heterogeneity concerning their nonmarket, market and adverse experience. The level of correlation between these variables is not very high, which suggests that firms also differ regarding the type of experience they develop over time. Finally, with nearly all pairwise correlations below 0.4, the correlations between the variables in our analysis do not cause concern (Kennedy, 2003). *******Insert Table 1 around here******* Firms appear to choose the location of their investments across a large range of EU countries. As figure 1 reveals, EU countries vary markedly in the number of foreign solar energy investments received during our period of analysis. In the map, countries that received a greater number of investments are represented in darker color. That is, countries colored in "light grey" received a small percentage of the investments in our setting, while those colored in "black" received the highest percentage of investments.
The map shows that countries with large solar resources, such as those located in Southern Europe and the Balkan peninsula, received a large proportion of the foreign investments; but so did France (9.6%) and the U.K. (11%), countries with average or low levels of solar radiation. Somewhat surprisingly, Germanythe largest solar market for most of our observation period-is not among the countries receiving the greatest proportion of international investments. 4 Overall, the pattern of investments suggests that the choice of country for a foreign investment in the solar energy industry is not straightforward; further investigation is needed to understand the underlying mechanisms driving firms' investment location decisions in this industry. *******Insert Figure 1 around here******* Anecdotal evidence from industry press offers preliminary support for the idea of jurisdiction shopping and the allure of feed-in-tariffs in particular. Several countries were reportedly considered more attractive for investors following the introduction of such policies, and in one interview the CTO of renewable energy company REC went so far as to say that "the solar market is getting so international that if any country pays a little more attractive feed-in tariff, it will get all the modules until demand is satisfied there" (Schmela, 2008: 53). The role of nonmarket experience also appears important. Following the announcement of a new feed-in-tariff in the Czech Republic, an executive from Sharp predicted a "wave of investors from Western Europe that are familiar with the subsidy regulations of the feed-in law" (Köpke, 2006: 27). And, alluding to the role of adverse experience, investors and lenders started "asking more questions about the risks" in the aftermath of feed-in-tariff cuts in major markets such as Germany and Spain (Hering, 2009: 53). As we discuss next, similar dynamics are borne out by the data. *******Insert Table 2 around here******* The table reveals overall support for our predictions. Model 2 shows evidence of jurisdiction shopping (i.e. firms seem to prefer locations with higher levels of feed-in-tariffs for solar energy when choosing the location for a foreign investment), as the coefficient of FiT policy generosity is positive and significant (β=4.478, p=0.000), providing strong support for H1. Model 3 shows that the interaction of FiT policy generosity and firm nonmarket experience is positive and significant (β=0.210, p=0.028). This provides support for H2, which predicted that nonmarket experience would make firms more sensitive to host country policies to support solar energy industry investments. Further, model 4 shows that the interaction of FiT policy generosity and firm market experience is not significant (β=-0.325, p=0.259).
This provides initial support for H3, which predicted that nonmarket experience should exacerbate the effect of FiT policy generosity more than market experience. To further assess this hypothesis, we re-ran models 2 and 3 using seemingly unrelated estimation approach, which allows the comparison of coefficients across equations. Results of this analysis confirm that the amplifying effect of nonmarket experience is indeed statistically greater than the effect of market experience (p-value=0.039). This result also holds when both coefficients are included simultaneously in the same equation (Model 6). We thus find support for H3.
Finally, model 5 shows that the interaction of FiT policy generosity and firm adverse experience is negative and significant (β=-1.252, p=0.030), providing support for H4. It appears that, as we predicted, firms with more adverse experience with FiTs in their home country or in the foreign countries where they operated become less attracted to countries where policy support is more generous. Model 6, which includes all interactions together, confirms the above conclusions.

Additional Tests
We carried out a series of robustness tests to assess the sensitivity of our findings to different empirical approaches that address specific features of our setting. *******Insert Table 3 around here******* First, as discussed earlier, it was relevant to focus on the EU as our empirical setting because the region pioneered and predominantly used feed-in-tariff policies and thus offered variation in terms of the generosity of these policies and the accumulation of different types of experience by firms. Although our data is restricted to EU countries, we believe that this is not a serious threat to the validity of our findings.
A problem would arise only if the results were driven by (some) firms having acquired experience outside the EU, which is not captured in our data. To eliminate this possibility we carried out a robustness test in which we excluded all investments made by firms that had prior experience outside the EU from our sample. Even though this reduced the sample to 119 investments by 70 firms, the results (see model 7 in Table 3) are consistent with our main findings. This suggests that considering only EU countries as investment options does not affect our conclusions.
Second, Germany was the largest solar market during our period of analysis, home to a great proportion of the investments in our dataset (69 out of 202 investments), and the first country to implement a FiT policy to stimulate the industry. It is thus possible that our results are driven by specificities of the German setting and investments by firms from this particular country. To assess this, we carried out our analyses after excluding observations associated with investments from German firms.
Our results remain robust with this alternative approach (see model 8 in Table 3).
Third, it is possible that a firm's experience in a particular country may offset the effects of other types of experience, such as nonmarket and market experience. In particular, once a firm has invested in a country, it may not be inclined to invest in it again and instead seek other locations for diversification.
Alternatively, prior investments in a country may make it more attractive. To address this, we added to our models a variable that took the value of 1 when the focal firm had invested in the potential host country prior to the focal investment year, and 0 otherwise. The results remain robust with the inclusion of this variable (see model 9 in Table 3).
Fourth, the results regarding H4 indicate that that firms with adverse nonmarket experience are deterred by generous FiT policies. One potential limitation here is that a host country's FiT policy generosity may be conflated with a history of sudden policy changes. If that were the case, our results might have been driven by firms being deterred not by the generous FiT policies but by the risk posed by the history of adverse changes in the FiT policy in said country. Thus, in two robustness tests we controlled for such regulatory uncertainty of the host country's FiT using either (i) the number of recent discontinuous negative changes in the FiT policy (those that occurred over the last year), or (ii) the number of discontinuous negative changes over a longer period (those that occurred over the past five years). As shown in models 10 and 11, adding these variables does not affect our findings. Firms with adverse experience are still more likely to avoid countries with a generous FiT policy even after we account for the riskiness of the country's FiT policy as reflected in past behavior.
Finally, the investing firms in our sample include both firms that were already present in renewable energy and those not originating from the renewable energy sector. We thus decided to explore how these two types of firms are affected by policies and negative change in policies when making their location decisions. We classified firms into renewable energy specialists and non-renewable energy firms based on their industry (NAICS) codes and run our models in samples split for each category. We find that both renewable energy and non-renewable energy firms are attracted by generous FiT policies; however, the effect of FiT policy is only moderated by nonmarket experience and adverse experience in the case of non-renewable energy firms (see models 12 and 13 in Table 3). Specifically, non-renewable energy firms are more attracted by generous FiTs when they have prior experience with FiTs, but also less attracted by such policies when they have adverse experience. We elaborate on this result in the discussion section.

DISCUSSION
By analyzing foreign location choices in the context of the European solar energy sector, we offer new insights on how firms behave when faced with cross-country differences in terms of supportive policy. Conceptualizing feed-in-tariff policies as a CSA, we find that policy support leads firms to engage in jurisdiction shopping (Ahuja & Yayavaram, 2011), whereby they search for and select to locate in countries with the most favorable policy frameworks. But not all firms engage in this behavior to the same extent. Firms' tendency to engage in jurisdiction shopping is exacerbated by firm-specific attributes, namely their nonmarket experience, but not by their market experience. Moreover, nonmarket experience has a complex impact on how firms respond to extra-jurisdictional policy, as the nature of this experience can reduce, rather than increase, some firms' tendency to engage in jurisdiction shopping. We discuss below the implications of our findings for research and practice.

Contributions to Research
The idea that institutions impose significant constraints on firm behavior has been a mainstay of the international business literature (Dorobantu et al., 2017;Henisz, 2000;North, 1990), with much of this work focusing on how policies impose constraints on firms, leading them to engage in jurisdiction shopping in order to avoid countries with stringent regulatory frameworks. However, research on jurisdiction shopping has often depicted it merely as an avoidance strategy (e.g. Ahuja & Yayavaram, 2011;Brunetta et al., 2017;Findlay, 2014;Tallman et al., 2018). By focusing on how supportive policies affect firms' location choices, our findings challenge this assumption and show that jurisdiction shopping characterizes a broader set of phenomena whereby firms assess institutions -not just unfavorable but also supportive ones -in foreign locations. More importantly, by considering which firms are more likely to engage in jurisdiction shopping and which are less so, we help address a key limitation of much of the institutional research in international business: the tendency to assume that firms are homogenously affected by institutions (see Aguilera & Grøgaard, 2019:22 for a discussion). We thus contribute to the international business literature by theorizing and showing that institutional support by the state differentially affects firms' location choices depending on firm-specific experience.
We also contribute to research on the CSA/FSA framework (Hillemann & Gestrin, 2016;Rugman & Verbeke, 1998;Rugman et al., 2011;Verbeke & Kano, 2012). Our findings are aligned with the idea that FSAs (firm experience in our context) affect location strategies in conjunction with CSAs of the host country (supportive FiT policies in our context). Prior research using this framework has suggested that experience is a firm-specific advantage that increases the likelihood of a firm selecting to locate in countries with strong CSAs (Clarke et al., 2013). Similarly, by relying on the CSA/FSA framework, we are able to demonstrate how firm experience matters in determining firms' response to institutions. We advance this research, however, as our results suggest that the impact of firm experience is less straightforward than previously considered. Firms' experience with the solar market does not appear to influence their propensity to engage in jurisdiction shopping, not unless the firm has specific nonmarket experience-and in particular experience with the same type of policy as that of the target countries. Thus, differentiating between market and nonmarket experience is a nuanced but critical insight that needs to be considered when studying firms' location choices, and potentially other important strategic decisions in the international context. Further, the effects of nonmarket experience do not apply across the board.
Firms that have adverse experience are less attracted to locations with very generous policies, potentially viewing these policies as a disadvantage, rather than an advantage, of the potential host country.
This last result is also consistent with recent research on how firms are affected by bounded reliability issues, which reflect "imperfect efforts to make good on open-ended promises" (Lumineau & Verbeke, 2016), in their international business operations. One of the bounds on reliability discussed by Verbeke & Greidanus (2009) is the benevolent preference reversal manifested as a phasing out of prior (over)commitments. While this notion has been applied primarily to the context of multinational enterprises (Eden, 2009;Kano & Verbeke, 2015;Verbeke & Greidanus, 2009), it is also evident in the government policy context: even if policy makers had good intentions when creating supportive policies, overcommitments such as generous but costly policy support may lead them to policy repeals or reductions. Such adverse changes may, in turn, be seen as unfulfilled commitments and lead to perceptions of the government, or of similar policies, as unreliable. Interestingly, our results indicate that negative perceptions about policies can spill over to other countries indirectly through firms' adverse experience. The implications are thus not constrained to the specific country where adverse policy changes have occurred. Rather, adverse changes may alter the perceived reliability of policies beyond a given national context and potentially affect industry transition efforts by countries that did not modify or withdraw their policies, as we elaborate below. Overall, our results point to future opportunities to expand ideas concerning bounded reliability to the supra-national level.
Moreover, by considering how the type and nature of firms' international experience condition their reaction to supportive policies, we respond to the calls for a more nuanced examination of how firms interact with their context in international business research (Liu & Vrontis, 2017;Oh & Oetzel, 2017).
Specifically, we contribute to the large body of research on how firm experience affects international business strategy. This work includes research using the CSA/FSA framework, but also the broader global strategy literature on international experience and learning (e.g. Albino-Pimentel et al., 2018;Belderbos et al., 2011;Delios & Henisz, 2003a;Oh & Oetzel, 2017). Despite the substantial insights offered, much of this work does not differentiate between market and nonmarket experience, or views experience as a general property of the firm at a given point in time. Our results indicate that both these tendencies should be reconsidered. First, some research has inferred nonmarket experience by looking at firms' presence in a particular market (e.g. Cuervo-Cazurra & Genc, 2008;Holburn & Zelner, 2010). But market experience does not necessarily offer capabilities to deal with the nonmarket environment; those may become available to the firm only insofar as it also deals with specific nonmarket requirements. We thus align with Oh and Oetzel (2017) in highlighting the need to unpack the effects of general market experience and domain-specific (e.g. policy-specific) nonmarket experience to better understand foreign entry strategies.
Second, experience is not a general property of the firm but depends on the confluence of the firms' presence in a market and the institutional environment that governs that market. For instance, two firms with similar nonmarket experience (e.g. both having operated in countries with a feed-in-tariff scheme) may behave in markedly different ways depending on the nature of that experience (e.g. whether they have experienced sudden drops in feed-in-tariff policies). Overall, by exploring the conditions under which firms engage in jurisdiction shopping, we call attention to the broader need to unpack different types of experience and the nature of that experience, in order to arrive at a richer understanding of firms' reactions to policy support outside their jurisdictions.
Finally, the focus on interfirm differences allows us to contribute to strategic management research. Although institutional support is ubiquitous in the global economic environment (Cimoli et al., 2009), strategy scholars have rarely addressed why firms respond differently to formal institutions that favor their sector (Georgallis & Durand, 2017), particularly to policies that are set outside the firms' current jurisdiction (Fremeth & Shaver, 2014). Observing the "surprisingly scant integration" between strategic management and industrial policy research, Lazzarini (2015) noted the need for scholars to examine how industry-specific supportive policies differentially affect firm strategies. Relatedly, Fremeth and Shaver (2014) recently called for research on how extra-jurisdictional policy affects market entry strategies. Our study of heterogeneous location choices in response to policies in the European solar power sector directly heeds these calls.

Practical Implications: industry-support policies and the energy transition
Industrial policy has always been an important engine of economic development. One of the most recent examples of this trend has been the renewable energy sector. Driven by a variety of motivationsaddressing climate change and pollution, reducing dependency on foreign resources, or creating jobs (Davies & Allen, 2013;del Rio & Mir-Artigues, 2014;Flamos, Georgallis, Doukas, & Psarras, 2009)many countries across the globe have enacted policies supporting renewable energy, fostering efforts toward an energy transition. These policies may not only be attractive to local players but may also attract foreign entrants. This paper explores the conditions under which this is the case.
Consistent with the idea of jurisdiction shopping, we find evidence that firms are indeed attracted by industry-support policies. The implications of this result depend on the goals of policy-makers. On the one hand, if the government simply wants to encourage the growth of the sector and facilitate the energy transition, then it can be seen as a positive externality. On the other hand, if its goal is to strictly to create national champions, these policies' consequences may be unintended (Davies & Allen, 2013;Peters et al., 2012). For instance, Peters et al. (2012) found that although demand-pull policies encourage local innovation, they also lead to innovation spillovers; policies established by a focal country stimulate innovation in neighboring countries as well. Similarly, our results indicate that industrial policies that may be set up to advance the local industry encourage, at the same time, the entry of foreign competitors.
A related and critical contingency that may advance this debate is the nature of foreign entrants that are attracted by supportive policies. Our findings indicate that jurisdiction shopping is moderated by firms' idiosyncratic experience. Specifically, supportive policies are not more likely to attract firms with greater market experience but rather tend to attract firms with more experience in dealing with similar policies. Moreover, in our robustness tests we find that both firms that specialize in renewable energy and those that have entered from other sectors are attracted to generous FiT policies, but nonmarket experience appears to be more important for the latter group of firms. One possible explanation for the latter result is that non-renewable energy specialists have to rely more on their own experience to understand how to assess public policies, while firms more ingrained into the sector can gain information about policies through other means (e.g. as they are more embedded in industry networks and more exposed to trade groups or industry press), making experience with policies less important for them or even redundant. Of course, this argument is speculative and future research is needed to identify the precise mechanism driving these differences. Regardless of the mechanism, however, the direct effect of FiT policies is positive and significant for both renewable energy specialists and firms coming from other industries. The finding that non-renewable specialists, in particular, are attracted to countries with generous FiT policies suggests that these policies incentivize firms to transition to renewable energy, notably solar energy.
While we are far from a full-scale systemic transition, viewing the energy transition as "a process that unfolds over time" requires individual firms to initiate such a transition (Delmas, Lyon, & Maxwell, 2019); our results confirm the intuition that supportive government policies can incentivize foreign firms to do so.
For managers, this result suggests that supportive policies are not only incentivizing renewable energy firms to invest in a particular country, but that they also attract competition from non-specialists. It also appears that generalist firms rely more on their own experiences to choose a location, an insight that may help managers anticipate future competitor moves. For policy-makers, our findings suggest that the energy transition can be facilitated by industry supportive policies. At the same time, the results regarding the role of adverse experience indicate that, for policy-makers interested in attracting foreign investment, it is not enough to avoid sudden changes to policies domestically. Since firms' response to supportive policies is affected by their adverse experiences in other countries as well, governments willing to use such policies to attract investments should invest resources in convincing other governments using similar policies to not arbitrarily change the rules of the game. This could, for instance, include relying on multilateral organizations as guarantors of stability, or lobbying supranational institutions (such as the EU) to regulate sudden policy changes, as these may render policies unreliable and deter foreign firms from transitioning into renewable energy. More broadly, this research signifies how the nature of firm experience speaks to the debate about foreign investment in response to industry support policies, and the role of states in industry transitions.

Limitations and Future Research
This study's findings are constrained by the usual limitations of single-industry studies. Our data only permit us to derive conclusions about a specific economic sector, albeit a very important one. EU countries exhibited variation in terms of their propensity to enact supportive policies for solar energy, but there was substantial consistency in terms of which policies they chose, with feed-in-tariffs being the policy predominantly used across the EU. This was an advantage for our research design, as we were able to observe jurisdiction shopping more clearly by comparing how firms are attracted to the same type of policy. However, other countries emerged as key markets in the solar industry since the end of our period of analysis, so it is important to consider whether our results are generalizable beyond this timeline and context.
We believe that these results are generalizable to a number of countries beyond the EU that used FiT policies to support the industry. In fact, the proliferation of FiT policies over the past few years may have provided more opportunities for firms to acquire nonmarket experience in a variety of contexts, making our findings even more relevant. On the other hand, our conclusions should be translated with care to settings where the type of supportive policy differs across countries. For instance, US states mostly used Renewable Portfolio Standards (RPS) to incentivize solar energy (Dutt & Joseph, 2019); since these were less common in other countries, it is unlikely that firms could gain sufficient experience that would translate into an FSA when entering the USA. Similar insights might apply to China and other South East Asian countries, which focused mainly on production-side subsidies. On the other hand, our theory is general in that it should apply beyond the context of feed-in-tariffs and solar energy, as a variety of industries benefit from policy support such as subsidies, grants, or favorable loans (Cimoli et al., 2009).
Indeed, many countries have offered policy support to industries such as automobiles, semiconductors, or chemicals (Brahm, 1995;Tyson, 1993) and more recently we have witnessed a growing trend of policy support for sustainable industries such as wind energy, green buildings, or electric cars (Jones et al., 2019;Lins et al., 2014;Mazur, Contestabile, Offer, & Brandon, 2015). We expect our theory to generalize to industries where similar policies are enacted across countries, allowing for the acquisition of domainspecific experience that firms can benefit from when making subsequent foreign investment choices.
One of the main findings of this study is that the nature of firms' nonmarket experience has substantial implications for how they react to foreign country policies. International business research can increase its explanatory power by incorporating adverse experience in models of international decisionmaking. But we have only scratched the surface of the richness of this topic. Future research is needed to identify different forms of experience and confirm that merely having experience does not sufficiently predict how firms cope with supportive or constraining institutional environments.
Overall, our research has made the case for a more detailed understanding of why firms respond differently to foreign countries' supportive policies. We hope future research will build on this study to explicate the full implications of supportive policies for foreign entry, and offer additional insights to policy-makers interested in stimulating industry transitions.