Bending the Rules or Changing Them? MNE Responses to Institutional Challenges in Transition Economies

New Institutional Economics and Institutional Strategy

We first adopt North’s (1990) notion of institutions: “Institutions are the rules of the game in a society or, more formally, are the humanly devised constraints that shape human interaction” (p. 3). New institutional economics emphasizes the role of institutions in determining how exchanges between actors occur (Coase, 1937; North, 1990; Williamson, 2000). In recent strategy literature, scholars have extended this understanding from new institutional economics by elaborating the arguments regarding how firms proactively respond to institutional constraints, beyond merely being influenced by institutions (Ahuja & Yayavaram, 2011; Doh et al., 2012; Dorobantu et al., 2017; Marquis & Raynard, 2015). The theoretical reasoning of how and why firms influence institutions starts by recognizing the incompleteness of institutions. North (1990) maintains, “Institutions are not necessarily or even usually created to be socially efficient; rather, they or at least the formal rules are created to serve the interests of those with the bargaining power to devise new rules” (p. 16). In this theory, actors are seen as boundedly rational; actors try to maximize their utility (e.g., pursuing their own self-interests) by making choices, but with their limited knowledge and cognitive capacity. Furthermore, institutions consisting of informal norms, as well as formal rules and enforcement mechanisms, shape the limited set of choices available to actors (North, 1990).

The incompleteness of institutions is more acute in countries where the overall institutional development for market exchanges is insufficient, and thus their governance systems are relatively weak, as in transition economy countries (Khanna, Palepu, & Sinha, 2005). By governance, we mean “the traditions and institutions by which authority in a country is exercised” (Kaufmann, Kraay, & Mastruzzi, 2008, p. 5). The public institutions and policies created by government as a framework for economic, legal, and social relations constitute the country’s governance system (Globerman & Shapiro, 2003). Therefore, we focus on the level of country governance as an indicator to study the level of institutional development of home and host countries in our research.

In countries that have a weaker governance system, idiosyncratic local norms and rules still interfere with newly enacted promarket policies, which amplifies transaction costs for market exchanges. In this regard, Dorobantu and colleagues (2017) differentiate the costs incurred by the attributes of the institutional environment from the costs incurred by the attributes of the transactions (Williamson, 1979, 1981). For instance, while market imperfection inherently raises transaction costs due to information asymmetry and the unequal distribution of bargaining power, transaction costs may still change, depending on how tightly a country governs market transparency and protection of property rights. Dorobantu and colleagues (2017) define these institutional costs as the costs representing “a given institutional environment’s fixed effect on the transaction costs associated with each governance form, holding transaction attributes constant” (p. 117).

Facing substantial institutional costs, firms strategically act to reduce such costs (Dorobantu et al., 2017) and gain “influence rents,” extra profits gained by shaping the rules of the game to favor them (Ahuja & Yayavaram, 2011). We refer to such actions as institutional strategies. While a variety of institutional strategies have been identified and classified (Ahuja & Yayavaram, 2011; Dorobantu et al., 2017; Marquis & Raynard, 2015), we focus on two particular activities that aim to directly influence public policies—lobbying and bribery. Ahuja and Yayavaram (2011) classify these activities as co-optation and capture, referring to them as “the influence that firms may establish over the decision-making of institutions” (p. 1643). While various other political activities are possible for the purpose of co-optation and capture, such as campaign contributions, government membership on company boards, and participation in political action committees (Lawton et al., 2013; Lux, Crook, & Woehr, 2011), most of these political activities, with the exception of corporate lobbying, have been largely unobserved in transition economies (Weymouth, 2012). Therefore, we focus on corporate lobbying as a focal influencing activity in this study, and on bribery, as the most prevalent corrupt behavior throughout the world.

Lobbying and Bribery as Co-optation and Capture

Lobbying and bribery have often been compared, given their similar goals: Both activities are aimed at reducing policy uncertainty and “obtain[ing] help from the public sector in exchange for some favor” (Campos & Giovannoni, 2007, p. 1). These activities reflect a firm’s purposeful action to co-opt or capture adverse institutional environments by “bring[ing] the interests of a challenging group into alignment with its own goals” (Trumpy, 2008, p. 480). The difference between the two is, however, still unclear (Rodriguez et al., 2006). The outright distinction discussed in the literature is that lobbying is regulated and legal in many countries, whereas bribery is not (Harstad & Svensson, 2011; Rodriguez et al., 2006). In addition, some studies differentiate between the two by focusing on the means of influencing public policies. Lobbying provides information to persuade policy makers (Hillman & Hitt, 1999), whereas bribery involves paying money to government officials in anticipation of preferential treatment.

These distinctions, however, are challenged in an institutional setting with weaker governance. As opposed to countries with strong institutional systems that govern lobbying activities with explicit laws and regulations (e.g., the Lobbying Disclosure Act of 1995 in the United States), lobbying tends to remain “a gray area” with various accompanying manipulations in institutionally less-developed countries (Anand, 2006). According to Chari, Hogan, and Murphy’s (2012) comparisons of lobbying regulations across countries, although a number of transition economy countries such as Lithuania, Poland, and Hungary have enacted lobbying laws since the 2000s, these regulations are limited in scope in most cases, and few lobbyists are registered. Consequently, these regulations are largely ignored (Chari et al., 2012). In such countries where the lobbying process is weakly governed, it is difficult to observe how lobbying is exercised.

Given our research question and objective, we distinguish between lobbying and bribery by focusing on their distinct purposes rather than the different means used (e.g., information vs. money) (Campos & Giovannoni, 2007; Harstad & Svensson, 2011; Rodriguez et al., 2006). Accordingly, lobbying activities are directed at policy-making institutions, whereas bribery is directed at the bureaucracy (Campos & Giovannoni, 2007). This distinction is not only suited to our study context, but it also aligns well with the conceptualizations of lobbying and bribery in previous studies. For instance, raising the ethical issues involved in the law-changing aspect of lobbying in the United States, Keffer and Hill (1997) define lobbying as “simply an attempt to persuade members of city councils, community commissions, state legislature[s], or the US Congress to support legislation favorable to one’s goals or desires” (p. 1372). In contrast, bribery has more to do with the execution of laws. In his research on the bribery practices of firms in China, Gao (2011) defines bribery as “offering, promising, or giving something in order to influence a public official in the execution of his/her official duties” (p. 176). Based on this discussion, we conceptualize lobbying as a firm’s attempt to change the content of laws or regulations in ways that favor the firm, and bribery as a firm’s attempt to break or bend the laws or regulations for an operational purpose.

Costs and Benefits Involved in Lobbying and Bribery

The different purposes of lobbying versus bribery (changing vs. bending the rules) imply that the costs and benefits involved in the two are different. Understanding the different costs and benefits involved in each of these activities is important in that the understanding affects managers’ cost–benefit calculations, which lead to their decisions regarding how they approach institutional challenges in a given setting. Lobbying, for instance, involves high up-front implementation costs with uncertain payoffs. As lobbying is aimed at changing a law or regulation, it incurs costs for learning complex legal and regulatory details and the policy-making process, researching potential allies and opponents, building reliable relationships with supportive policy makers, and training lobbyists about firm-specific factors (Kerr, Lincoln, & Mishra, 2014). Given the high costs and long lead time, it is natural for a firm to assess the expected benefits and odds of succeeding before deciding to lobby. The expected benefits can vary on a case-by-case basis, but the common aspects of the benefits from lobbying are that, despite the high costs, once a rule is changed as a result of lobbying, the effect lasts for a long time (Campos & Giovannoni, 2007; Harstad & Svensson, 2011).

However, bribery can seem less expensive to implement since it is not about changing a law or regulation; rather, it is about obtaining preferential treatment from corrupt government officials. Therefore, extensive learning is not required, and the bribes—money or gifts—are easily fungible across different contexts; thus, the exchange is much simpler. Bribery, however, incurs higher ex post risks, as there is a chance that a firm’s bribery attempt will be uncovered. Citing Noonan (1987) who discussed pervasive bribery across countries, Martine, Cullen, Johnson, and Parboteeah (2007) state that “there is not a country in the world which does not treat bribery as criminal on its lawbooks” (p. 1407). Due to its illegality, once firms get caught, they will receive legal sanctions with financial penalties as well as the loss of legitimacy in the market. Despite such ex post risks, firms are easily tempted to engage in bribery, given its temporary yet immediate benefits, which include receiving “a good (e.g., a contract, a permit, etc.) or preventing a bad (e.g., compliance with regulation, payment of taxes, etc.)” (Cuervo-Cazurra, 2016, p. 36). Although corruption tends to be pervasive in weak institutional settings, and thus firms are under increased pressure to bribe, firms do not automatically or uniformly follow the bribe demander’s pressures. Instead, it is the firm’s own decision as to how actively and purposefully it seeks such unfair advantages by using bribery (Martin et al., 2007).

Institutional Imprinting

Given that MNEs originate from their home country, which has a different institutional environment from that of the host country, institutional imprinting theory provides a perspective that complements prior knowledge by allowing us to consider home country institutions in addition to those of the host country. Institutional imprinting theorists emphasize the critical influence of institutional conditions during the “founding” period in establishing organizational activities (Kriauciunas & Kale, 2006; Marquis & Tilcsik, 2013; Stinchcombe, 1965). Marquis and Tilcsik (2013, p. 199) refer to imprinting as “[. . .] a process whereby, during a brief period of susceptibility, a focal entity develops characteristics that reflect prominent features of the environment, and these characteristics continue to persist despite significant environmental changes in subsequent periods.” Firms tend to establish their organizational practices and capabilites that deal with institutional contingencies during the founding period. This founding period is particularly important because, in that period, firms tend to lack resources; therefore, to survive, they must develop the strategies and practices that the external environment endorses (Kriauciunas & Kale, 2006).

Once organizations are imprinted by the institutional conditions during the founding period, their practices and capabilities persist, even when the contexts change. Kogut (1993) argues that the country of origin affects the capability of firms because “the firm develops [. . .] out of the socioeconomic conditions of its home environment. Even as the firm internationalizes, it remains imprinted by its early developmental history and domestic environment” (p. 106). Holburn and Zelner (2010) also state that a firm’s experience in its home country institutional environment “arguably represents a more fundamental influence” in increasing its capacity to deal with institutional hazards overseas (p. 1292). These studies suggest that in addition to the host country’s institutional environment, home country institutions also exert their influence when an MNE chooses its influencing activities in the host country.

Institutional Influence on the Costs and Benefits of Lobbying and Bribery

As discussed earlier, lobbying incurs high up-front costs with uncertain payoffs while bribery bears ex-post risks with more certain payoffs. Accordingly, whether or not firms attempt to lobby is largely a function of the probability that lobbying will succeed, given the high costs; whether or not a firm engages in bribery is largely a function of the probability that bribery will be uncovered and, due to their illegality, the extent to which economic, legal, and social sanctions will be imposed. Considering a country’s institutional condition particularly in terms of the governance quality is important because the country’s governance directly affects the costs and benefits involved in lobbying and bribery.

When a country has a well-established governance system, the policy-making process tends to be monitored effectively, and thus is accountable because of checks and balances. Under such a system, political competition tends to be reasonable and fair, thereby enabling “market-like political competition” (Bonardi, Hillman, & Keim, 2005). Despite the high costs, under such conditions, firms can reasonably calculate the costs and benefits involved in lobbying (Hillman & Hitt, 1999). Thus, when there is a political agenda strong enough to pursue, firms may invest in lobbying. In contrast, in a country with weaker governance, the policy-making process is usually interrupted by idiosyncratic factors such as personal ties and private interests. In such an environment, as the state does not have to be accountable for its actions, private interests tend to supersede public interests. Thus, firms are hesitant to invest in lobbying due to the relatively high costs in relation to the potentially unfair and opaque process of policy making, which creates difficulty in calculating investment payoffs (Weymouth, 2012).

However, under a weaker governance system, monitoring corrupt behaviors as well as the legal and economic sanctions imposed on them tends to be weak, which makes corruption more prevalent (Martin et al., 2007). In an environment where corruption prevails, there could also be pressure for a firm to follow norms regarding how things are done in that environment (Spencer & Gomez, 2011). In that setting, social sanctions such as losing legitimacy and reputation with bribery are unlikely to be substantial. In addition, in such environments, government officials expect to “sell” their authoritative power, and they allocate their power and efforts disproportionally to bribers versus nonbribers (Lee & Weng, 2013).

In an opposite condition, however, both the demand for and supply of bribery are discouraged (Martin et al., 2007) because the economic and legal sanctions (e.g., imposed fines and potential prison sentences) are significant and the social sanctions (e.g., the risks of losing one’s legitimacy and reputation) are severe. Given all of these risks, the expected costs involved in bribery outweigh the potential gains, on both the demand and supply side. Therefore, in an institutional environment with a stronger governance system, firms have few incentives to engage in bribery for preferential treatment from government officials, who also have few incentives to take risks by receiving bribes. For this reason, it is difficult to expect bribery to be an effective strategy in influencing public policies in countries with a stronger governance system.

This discussion suggests that the expected costs and benefits involved in lobbying and bribery are different and are significantly affected by the institutional conditions in which the influencing activities are undertaken. A real challenge confronted by MNEs is that they are embedded in both their home and host countries; thus, their choice of influencing activities is also affected by their home country’s institutional conditions, independent of those of the host country. Furthermore, we consider that, even under the same institutional condition in a host country, the amount of institutional costs incurred by an MNE differs, depending on its level of exposure to local market transactions (Dorobantu et al., 2017; Feinberg & Gupta, 2009).

The Impact of Home Country Institutions on Lobbying and Bribery

While a firm repeatedly engages in a particular domain of activity (e.g., lobbying or bribery) under the home country’s institutions, institutional imprinting occurs. While doing so, the firm learns what is (and is not) allowed by external institutional environments. Even if a firm is not politically active in the home country, the firm still learns socially accepted political activities through social interactions, given that the legitimacy of CPAs and corruption are profoundly engraved in the ideologies, beliefs, and social norms in a given society (Marquis & Tilcsik, 2013).

These imprinted organizational practices tend to persist, even if the external contexts change. MNEs institutionalize a particular influencing activity and establish their own knowledge structures, perceptions, and attitudes toward the particular activity corresponding to the home country’s institutional contingencies. Once these structures, perceptions, and attitudes are established, they provide the firm with a framework and schemes to interpret events (Baron, Hannan, & Burton, 1999). Moreover, according to Schreyögg and Sydow (2011), “These initial frames and schemes imprint decision-making processes in organizations, and eventually, amount to replicated patterns” (p. 326).

The institutionalization of imprinted practices in a corporate setting affects managers’ sociocognitive processes, especially when they face uncertain and ambiguous conditions (Benner & Tripsas, 2012). First, the imprinting effect manifests itself via managers’ cognitive processes. Facing uncertainty, managers tend to make sense of confusing and often conflicting signals using references to their past experiences (Weick, 2000). Benner and Tripsas (2012) show that prior industry experiences influence how firms behave in a nascent industry. Such a tendency comes from the decision makers’ cognitive bias as a result of imprinting; with these accumulated experiences, decision makers easily overlook information and evidence suggesting other possibilities. In our study context, when MNEs need to conduct an influencing activity in a host country, they may take a course of action imprinted by the home country’s institutions rather than thoroughly searching for information regarding the host country’s different institutional contingencies and possible alternatives.

Second, the imprinting effect becomes persistent and even reinforced through institutional forces within the MNE’s internal institutional field. Institutional scholars have argued that firms tend to reduce uncertainty by imitating other firms. In our study context, when MNEs conduct business in a host country setting, they confront high uncertainty stemming from their “foreign” status (Zaheer, 1995). However, local firms cannot always provide a proper reference point for MNEs to imitate, especially in less-developed institutional settings. In this case, the MNE’s host country subunit is likely to resort to home-based practices, because although foreign subsidiaries are subject to institutional forces from the host country, they are also continuously influenced by the transfer of home-based practices (White, Fainshmidt, & Rajwani, 2018). As foreign subsidiaries rely heavily on home country headquarters’ support for critical resources, “subunits are often more dependent on the parent company than their local external environment . . . [Thus, MNEs] are likely to consciously create and strengthen their intra-organizational field so as to reinforce and disseminate a shared business model” (Kostova, Roth, & Dacin, 2008, cited by White et al., 2018, p. 3).

Taken together, we argue that MNEs’ influencing activities—lobbying and bribery—are institutionalized with repeated direct and indirect experiences in their home country. Once they are set corresponding to the institutional contexts of the home country, they persist even in the different institutional contexts of a host country, owing to the influence of imprinting on managers’ cognitive processing and the MNE’s internal institutional forces. Our earlier discussion suggested that an institutionally developed setting with a stronger governance system favors lobbying over bribery; an opposite condition favors bribery over lobbying. Hence, we propose the following:

Institutional Costs in Using the Local Market

Along with home country institutions, MNEs are directly affected by the host country’s institutional environment in deciding how to respond to institutional challenges in that country. Country-level institutional conditions are the same for all firms conducting business in those environments; thus, institutional costs are country-level fixed costs, as discussed earlier (Dorobantu et al., 2017). However, such institutional costs are incurred when firms use local markets. Therefore, depending on their exposure to the local market, the level of institutional costs faced by MNEs may vary (Dorobantu et al., 2017; Feinberg & Gupta, 2009).

As previously discussed, new institutional economists draw attention to institutions, since these institutions devise incentives and constraints for human exchanges and affect transaction costs (North, 1990; Williamson, 2000). The market system is far from perfect, given its information and power asymmetries; the issue of an imperfect market becomes more (or less) serious depending on the institutional conditions surrounding the market (Ahuja & Yayavaram, 2011). More specifically, insufficiently developed institutions tend to result in procedural ineffectiveness, capacity problems, and corruption, thereby aggravating these information and power asymmetry issues (Ahuja & Yayavaram, 2011). Responding to such institutional constraints, firms develop various institutional strategies.

One immediate and considerable solution for MNEs (even before considering specific influencing activities for co-optation and capture through lobbying and bribery) is avoidance. Ahuja and Yayavaram (2011) consider avoidance as “a simple mechanism to control the institution” by “preventing or delaying the institution from coming into existence or beginning to function effectively” (p. 1641). In an MNE context, by leveraging the MNE’s subunits dispersed across multiple countries, the MNE can minimize the use of local markets if these markets substantially incur institutional costs. Feinberg and Gupta (2009) indeed demonstrate that MNEs increase their internal market transactions to a greater extent (e.g., within-firm sales) in countries where institutional risk is high. They argue that by increasing internal transactions, MNEs can minimize their exposure to and dependence on the external institutional environment within a host country (Feinberg & Gupta, 2009). Given that institutional costs are considered a toll to use the market (Dorobantu et al., 2017), minimizing the MNE’s exposure to the local market and increasing its internal transactions can be a viable solution to reduce the institutional costs incurred in a host country.

When an MNE cannot exercise the avoidance strategy and must use the local market to a greater extent, higher institutional costs are incurred; therefore, the MNE will attempt to reduce such costs by adopting various institutional strategies, including both lobbying and bribery. Regarding lobbying, when the local markets are inefficient due to institutional limitations, changing the rules in ways that favor the MNE’s subunit would provide enjoyable benefits. However, given the high up-front costs of lobbying, MNEs will decide to lobby only when the expected benefits outweigh the high costs. If they are not strongly committed to the local market, the firm has few incentives to deal with the complex process and monetary costs involved in changing the rules. Thus, higher reliance on local markets for sales provides significant incentives for the MNE to pursue lobbying as, if successful, the MNE can enjoy benefits over a long period despite the high costs.

As MNEs use the local market to a greater extent, they must develop not only a long-term solution such as lobbying, they also need to deal with day-to-day operational challenges that warrant immediate solutions. In host countries with a weaker governance system, local firms would be prone to engage in bribery. As discussed earlier, the perception that “other firms do the same” can create normative pressure for an MNE to follow this norm and justify their illegal behavior of bribery as being the way things are done in the local environment (Spencer & Gomez, 2011). In such a setting, the institutionalized ways of monitoring and punishing bribery are also likely to be lenient. Therefore, the actual probability of getting caught and receiving social and legal sanctions is indeed lower. Such normative pressure, as well as lenient institutional conditions, will allow managers to inflate the benefits and discount the costs in their calculation. Therefore, firms are more likely to seek ways to break or bend the rules by bribing government officials. Considered together, as MNEs rely more on the local market for sales, they will be more likely to lobby and will also more actively engage in bribery to reduce the institutional costs incurred by increased exposure to the local market. Thus, we propose the following:

Interaction Effects

While the need to reduce institutional costs in using the host country market would lead to MNEs’ increased engagement in both lobbying and bribery, the home country’s institutional imprinting will still affect MNEs’ decision on how they approach institutional constraints in the host country. When MNEs are from a country with a stronger governance system, the home country’s institutional imprinting will affect decision makers’ cognitive processing and will exert institutional pressures over local subunits to execute the influencing activity aligned with the home country’s institutional condition (White et al., 2018), which is lobbying in this case. However, these MNEs will perceive the ex post risks involved in bribery—such as economic, legal, and social sanctions—as being higher than those from a weaker governance condition, partly due to decision makers’ cognitive bias resulting from unfamiliarity with bribery (since the home country’s institutions do not allow it) and partly due to the stronger formal institutional mechanisms of the home country that increase the odds of getting caught and receiving more severe sanctions.

In comparison, for MNEs from a weaker governance condition, lobbying is largely an unfamiliar practice because the home country’s institutional environment does not provide favorable conditions for this political action. In contrast, these MNEs are more likely to directly or indirectly experience corrupt practices—such as bribery—in their home country. Subsequently, such corrupt practices are likely to be imprinted and persistent through a similar sociocognitive process that we discussed earlier (Baron et al., 1999; Cooper, Folta, & Woo, 1995; White et al., 2018): Decision makers repeat the routines they are familiar with, while the intra-MNE institutional process leads to similar behaviors across subunits. As a result, these MNEs are more likely to pursue bribery over lobbying when they need to engage in influencing activities, as their reliance on the local markets becomes higher. Therefore, we propose an interaction effect, as follows:

Data and Sample

We test our hypotheses using data from BEEPS, jointly conducted by the World Bank and the European Bank for Reconstruction and Development (EBRD) in 2002 and 2005. As our study examines institutional strategies particularly pertaining to countries in which the overall institutional development for market exchanges is insufficient and thus their governance systems are relatively weak, we focus on the surveys conducted in transition economy countries while excluding a few others such as Germany and Ireland. Transition economies refer to the countries that belonged to the former Soviet Union, Eastern Europe, and East Asia. Across these countries, the degree of market building has varied since the initial shock (the collapse of the Soviet Union), but there is a reasonable level of consensus that transition economy countries have a “less developed government and regulatory structure, suggesting that market regulation, corporate governance, transparency, accounting standards, and intellectual property protection may not be as reliable or mature as those in more advanced economies” (Marquis & Raynard, 2015, p. 300; see also Peng & Heath, 1996; Shinkle & Kriauciunas, 2012).

BEEPS is a comprehensive firm-level survey conducted in each of the selected countries. Through face-to-face interviews, top managers and business owners responded to survey questions. When the respondents were incapable of answering a question, they often called the managers in charge of the relevant functions (http://ebrd-beeps.com/methodology). In selecting the sample, this survey adopted stratified random sampling based on the characteristics of the firm populations operating in each country. Therefore, the sample was not skewed toward large firms, as is often the case with many other studies using listed firms.

As the survey questions contain sensitive components, such as business–government relations and bribery-related topics, the EBRD and the World Bank carefully administered the survey using rigorous interview protocols. For example, the EBRD and the World Bank hired private contractors as interviewers with no involvement of government agencies to ensure interviewees’ candid participation. Also, to facilitate more active participation and unbiased answers from the interviewees, confidentiality was assured. Given these efforts, BEEPS has been recognized as a credible source of firm-level data, comparable across different countries (Krammer, 2019; Lee & Weng, 2013; Martin et al., 2007; Spencer & Gomez, 2011).

Although BEEPS has been conducted every 3 to 4 years until recently, only the surveys taken in the years 2002 and 2005 contain information regarding the home country of a focal MNE. Therefore, we combined the surveys conducted in these 2 years. Among the sample participants, we selected firms whose foreign ownership shares are greater than 50%, as this study focuses on MNEs operating in host countries. After excluding observations with missing variables, the final usable data contain 911 MNEs, with 37 home and 26 host country pairs (see the appendix for more information regarding the home and host countries included in this study). Among the 911 observations, only 50 participated in the survey in both years. Thus, this data set is largely cross-sectional. We compared the answers with our key variables between the samples that excluded observations for missing variables and those that did not by conducting t tests. No systematic differences were detected between these groups.

Variables

Analysis

To test our hypotheses with respect to lobbying (H1a, H2a, H3a), in which the dependent variable is binary, we used a logistic regression model for the analysis. In the other hypotheses related to bribery (H1b, H2b, H3b), we used an ordinary least squares regression model.

We checked for multicollinearity with the variance inflation factors (VIFs) of all variables in each model. The highest VIF is 2.39, and the average VIF in our full model is 1.24. Given that the VIFs are well below the commonly accepted threshold of 10 (Hair, Anderson, Tatham, & Black, 1998), our study does not suffer from multicollinearity. We used robust standard errors in our analysis.