Abstract
Despite many studies investigating how lobbying expenditures enhance performance outcomes, limited attention has been given to understanding the underlying mechanisms driving firm lobbying behaviors—and, particularly, where firms target their lobbying expenditures. We argue lobbying breadth serves as a risk management strategy to both hedge against possible government intrusion and minimize disapproval of firm actions from government officials. As such, we posit firms’ strategic risk will positively relate to breadth of lobbying firms’ target. Further, if lobbying breadth serves as risk management strategy, then other aspects that affect decisions about the amount of risk protection a firm may need should also affect this relationship. We argue that CEOs’ ownership and firms’ political uncertainty may exacerbate exposure from firm risk taking to government actions, thus strengthening the insurance relationship. Similarly, we argue available slack, as a form of internal insurance against government actions, will weaken the relationship between strategic risk taking and lobbying breadth. We find support for most of our arguments in a sample of U. S. manufacturing firms. This research extends understanding of lobbying beyond the predominant focus on expenditures and securing beneficial outcomes. Exploring how lobbying breadth serves an important risk management role also offers insights into firm nonmarket strategy.
Corporate political activity (CPA), or “attempts to shape government policy in ways favorable to the firm,” has long been seen as a risk mitigation tool (Hillman, Keim, & Schuler, 2004: 837), both in providing insights that help firms minimize losses from undesired government actions and in influencing the perceptions of government officials so they act as firms desire. Indeed, use of CPA by firms to limit risk exposure is a long-held topic of scholarly interest (e.g., Buchanan & Tullock, 1965; Olson, 1965) that continues to draw the attention of researchers (e.g., Hillman et al., 2004; Schuler, Rehbein, & Green, 2019). As Hadani and Schuler (2013: 165) note, despite widespread belief that CPA can mitigate firms’ risk, “empirical studies have not consistently borne this out” (see also Mellahi, Frynas, Sun, & Seigel, 2016). For example, two studies using meta-analysis, “the preferred method for accumulating evidence” on a topic (Combs, Ketchen, Crook, & Roth, 2011: 178), offer conflicting findings as to how CPA helps mitigate risks and curry favor (Hadani, Bonardi, & Dahan, 2017; Lux, Crook, & Woehr, 2011).
Motivated at least in part by these conflicting findings on the outcomes of engaging in CPA (e.g., Cooper, Gulen, & Ovtchinnikov, 2010; Ridge, Ingram, & Hill, 2017), a smaller but growing body of research shifts focus away from expenditures on CPA and toward firms’ allocations of those expenditures across a wider breadth of government (defined as the extent of government activities or entities the firm is attempting to engage via CPA activities). The assumption underlying this work is that since “it is difficult for firms to predict what public policymakers will decide” (Hadani et al., 2017: 355), dispersing CPA across a wider breadth of the political sphere helps firms—in effect, by “spreading their bets” (Haspeslagh, 1982)—firms are argued to benefit via two mechanisms: (a) better informing them about risks in the political environment, thus facilitating protective adjustments, and (b) increasing their chances of influencing government officials to act as desired (e.g., Keim & Baysinger, 1988; Meznar & Nigh, 1995). 1
While the growing body of research on how firms allocate CPA suggests generally there may be benefits to widening breadth (see Minefee, McDonnell, & Werner, 2020, for a discussion of some exceptions), extant work focuses almost entirely on the outcomes of such actions (e.g., government contracts and firm performance in Ridge et al., 2017; stock returns in Cooper et al., 2010; survival and growth in Zheng, Singh, & Mitchell, 2015). Little focus has been given to understanding what drives firms’ CPA breadth, however, raising important questions for our knowledge of the topic. Like the assumption that CPA expenditures help firms, there is a need to further develop and substantiate our understanding of this important issue (Hadani & Schuler, 2013), particularly given a growing number of studies that allude to the risk mitigation role of targeting an increased breadth of government but that neglect drivers of this behavior beyond environmental factors (e.g., Hiatt, Carlos, & Sine, 2018). At the same time, firms do not have unlimited resources (Penrose, 1959) and thus must decide what to allocate for CPA (Hadani et al., 2017).
Better understanding what drives CPA breadth can offer insights for avoiding wasteful allocations that may be akin to blindly “throwing money at the issue,” thus helping avoid ill-advised reductions in CPA that may both limit information needed and hinder chances of influencing government officials to act as desired. We thus join others in pointing to a need for research that develops and substantiates our understanding of what drives CPA breadth (e.g., Hiatt et al., 2018; Ridge et al., 2017).
To address the paucity of research on drivers of CPA breadth and enhance our understanding of this growing aspect of firm nonmarket strategy, we build on extant theory suggesting firms broaden their lobbying—the most common and largest form of CPA (Milyo, Primo, & Groseclose, 2000) 2 —as risks in their political environment increase (e.g., Keim & Baysinger, 1988; Meznar & Nigh, 1995). Extending the logic that risk may drive firms to spread their bets, we argue that as firm strategic risk—defined as resource allocations with more outcome uncertainty (Hoskisson, Chirico, Zyung, & Gambeta, 2017; Palmer & Wiseman, 1999)—increases, so too will lobbying breadth. We then expand the foundational theorizing to address other aspects that may influence a firm’s risk profile and, thus, affect the degree to which the firm tries to broaden lobbying to protect against strategic risk.
We thus argue CEO vulnerability to financial losses in the form of firm stock ownership will strengthen the relationship between strategic risk and lobbying breadth. Relatedly, we theorize that if lobbying breadth is indeed a risk mitigation tool, firm political uncertainty will affect the degree to which firms respond to strategic risk by spreading lobbying across a wider breadth. Finally, protection offered by lobbying breadth is likely not as essential if a firm has alternative risk mitigation available. Since available slack “insulates the technical core of the organization from environmental turbulence” (Tan & Peng, 2003: 1250), we argue that as firms’ available slack increases, the need for firms to engage in lobbying breadth in response to strategic risk is lessened. In a sample of publicly traded manufacturing firms in the United States, we find broad support for our theorizing. Collectively, our study contributes to CPA theorizing and poses implications for both future research and practice, each of which we highlight in the Discussion.
Conceptual Background
Considerable empirical work addresses how firms can spread their bets or “hedge their risks” to guard against uncertain outcomes emanating from their strategies (Bromiley, McShane, Nair, & Rustambekov, 2015). In particular, scholars have found firms utilize various risk management strategies to mitigate their own strategic risks. For example, Liebenberg and Hoyt (2003) found an association between firms’ debt and the appointment of chief risk officers (CROs). In a similar vein, Pagach and Warr (2011) show firm debt, earnings volatility, and stock performance were associated with the likelihood of employing a CRO (see also Almeida, Hankins, & Williams, 2017; Rampini, Sufi, & Viswanathan, 2014). Previous studies also find that managerial motives of risk reduction figure prominently in company corporate risk strategies (e.g., Denis, Denis, & Sarin, 1997). For example, Tufano (1996) found that CEOs of gold-mining firms were likely to manage gold price risk when they have a significant share of firm stock. Recently, Bradley, Pantzalis, and Yuan (2016) provided evidence that lobbying and corporate political action committee expenditures serve as a risk mitigation mechanism, with politically active firms enjoying lower cost of debt when policy uncertainty is high relative to firms that are less politically active.
At the same time, organizational scholars have long recognized that if stakeholders do not approve of firm actions, they engage in boycotts, badmouthing, withdrawing access to critical resources, or even revoking the right to do business (Fombrun, Gardberg, & Barnett, 2000). As such, substantial research has emerged that considers how firms use their resources to mitigate risks from stakeholder disapproval to ensure that when negative actions arise, these strategies can minimize the losses from the event. As a case in point, organization scholars posit that corporate social responsibility (CSR) can function as an “insurance” strategy to help firms avoid questionable practices and build reserves of goodwill or reputational capital (Jia, Gao, & Julian, 2020), forming a sort of safety net that protects them when they encounter crises (Fombrun et al., 2000). Godfrey, Merrill, and Hansen (2009) found that market declines for firms that face lawsuits were lower if they actively engaged in CSR. Similarly, Minor and Morgan (2011) find that CSR acts as insurance by reducing stock market drops for firms that increase their CSR activities following product recalls. Koh, Qian, and Wang (2014) also find that firm CSR inversely relates to litigation risks. Similar findings have emerged in the impression management literature that considers how firms strategically use private information in order to manage market perception of negative events. Specifically, Graffin, Haleblian, and Kiley (2016) found that firms use impression offsetting by releasing positive information ahead of negative events to strategically minimize negative market stakeholder responses.
Research directed at understanding firm risk mitigation strategies, however, pays minimal attention to government stakeholders. However, such stakeholders, through regulation and the provision of resources, substantially affect firms (Pearce, Dibble, & Klein, 2009) from both a strategic (e.g., Abdurakhmonov, Ridge, & Hill, 2021; Ridge, Hill, & Ingram, 2018) and a performance (e.g., Hillman, Zardkoohi, & Bierman, 1999) standpoint. Indeed, such logic is alluded to but not empirically investigated in the CPA literature (Hadani et al., 2017). Particularly, CPA scholars suggest that firms engage in CPA to protect against any “unwanted political interferences” and curry “preferential treatments from government officials (Hillman et al., 2004)” (Mellahi et al., 2016: 155). Despite a long history of CPA in practice, and decades of scholarly attention to the topic, whether such actions offer firm protection or influence that is ultimately helpful is far from a settled question (e.g., Hadani et al., 2017; Lux et al., 2011).
As Mellahi et al. (2016) highlight in reviewing CPA scholarship, the bulk of research attention on the topic focuses on firms’ engagement in CPA. Hadani et al. (2017), referring specifically to the works of Lord (2000, 2003) and Cooper et al. (2010), highlight a smaller number of studies investigating how firms can benefit not by simply engaging in CPA but by spreading CPA across the political landscape. This line of research builds on the long history of resource allocation scholarship (e.g., Haspeslagh, 1982; Williamson, 1975), the premise of which is that “since the returns associated with any single resource allocation [like CPA] are uncertain, firms benefit from increasing resource allocation breadth because ‘spreading bets’ across a range of possibilities may increase the probability of at least some success” (Ridge et al., 2017: 1142). Thus, firms gain access to information that helps them engage in protection from unwanted political interferences and/or increasing the chances of influencing officials to act as desired. In their investigation of firm lobbying activity, Ridge et al. (2017) point to substantial within-political-environment variance that suggests other variables and contextual nuance should also be considered (see also Hadani et al., 2017). We thus direct our attention to addressing such factors and, in doing so, answer the call of Ridge et al. (2017) and Hiatt et al. (2018) for theoretical and empirical work that investigates this important question.
We suggest that one particularly germane element associated with firms’ resource allocations in the form of lobbying breadth (i.e., the extent of government activities or entities targeted with lobbying activity) is strategic risk, or “corporate strategic moves that cause returns to vary . . . [and] for which the outcomes and probabilities may be only partially known” (Baird & Thomas, 1985: 231). Sanders and Hambrick (2007) argue that strategic risk comprises three elements: the amount at stake, the range of possible outcomes, and the probability of loss that may even threaten the firm’s existence. Strategic risk increases as resource commitments rise and outcomes either become increasingly unpredictable or pose a large potential for sizeable harm (e.g., insolvency, windfall loss). For example, low-risk strategies generally require minimal levels of resource commitments and are unlikely to pose a threat to the firm if unsuccessful, whereas high-risk strategies are the opposite in each case. Thus, strategic risk often entails managers engaging in strategic choices associated with increasing probabilities of loss (Hoskisson et al., 2017) and is determined by the proactive strategies employed by firm managers in allocating resources (Palmer & Wiseman, 1999).
We suggest that viewing lobbying breadth as a form of insurance or risk management tool against strategic risk that buffers a firm from environmental uncertainty stemming from the government may advance our understanding of differences in lobbying breadth undertaken by firms beyond what engaging in lobbying and associated expenditures can tell us. In particular, we argue it is important to view lobbying breadth as an insurance or risk management tool by focusing on the preventive or buffering goals of lobbying, which are often overlooked (Hillman et al., 2004), as well as its role in helping to generate political capital (Kim, 2019). Thus, even though some stakeholders may approach firm lobbying activity with suspicion based upon the perceived legitimacy or “acceptability” of it (Jia, 2018; Minefee et al., 2020), when considered legitimate, a broad risk management view of lobbying by firms may suggest important positive benefits from this nuanced aspect of firm political behavior. Particularly, this view of lobbying moves beyond the primary focus of the CPA literature on the benefits firms accrue from such actions (Mellahi et al., 2016) and emphasizes how the breadth of CPA may provide additional insight above and beyond the amount that firms expend or the simple fact they engage in CPA at all.
Theory and Hypotheses
Central to strategic management is managing risk exposure, or the extent to which external actions affect a firm (Miller, 1998). Strategic risk increases a firm’s vulnerability to external changes or potential negative actions that may adversely impact the firm as well as the perception of governmental stakeholders regarding the firm’s actions, threatening future performance. This is because increased strategic risk leads to uncertainties due to both government officials’ disapproval of investments associated with new products or processes and their discretion to initiate potential regulatory actions that may hamper these investment decisions. For instance, in the subprime mortgage crisis, many banks had taken on large amounts of risk through mortgage-backed securities and collateralized debt obligations; yet many of these banks continued to function normally and even profit until the external shock of falling real estate prices (Reinhart & Rogoff, 2008). In other words, a negative governmental view of a firm may create a large amount of uncertainty as to not only if and how a governmental action will occur but also where and when it may originate. This in turn may limit the firm’s ability to manage potential risks emanating from the government.
We expect firms that increase strategic risk will accordingly expand their levels of lobbying breadth in an attempt to insure the firm against future potential negative actions across the multitude of possible agencies or legislative acts from which they may originate. In this way, lobbying breadth acts akin to insurance, protecting a firm from the political environment by expanding both influence and information exchange through ties of the firm’s political activities across a larger portion of the political sphere that may affect it. This proactive approach to managing the environment is directed at “reduc[ing] a government’s effects on an organization’s activities, either by insulating the organization from government effects or changing those effects” (Blumentritt, 2003: 206).
Lobbying breadth benefits firms through two primary mechanisms: (a) gathering and providing a variety of information that can help firms wielding influence over decisions (e.g., Cao, Fernando, Tripathy, & Upadhyay, 2018; Lux et al., 2011) and (b) establishing political ties within wider branches of government so to prevent or minimize government disapproval of firm actions (Fisman, 2001; Hillman, 2005). Nownes (2006) argues that by engaging in lobbying activity, firms actively gather information about government activities in an attempt to identify potential actions that may affect their firm. Accordingly, firms both actively and reactively work to influence governmental actors as to potential courses of actions. Firms gather information to identify potential government actions and proactively attempt to influence action as it is difficult to eliminate government policy after it has been enacted.
Indeed, several interviews by Nownes (2006) of lobbyists suggest it is much easier to stop a policy than it is to start a new one or remove one already enacted. Thus, most firm lobbying activity is enacted on the belief that “it is better to ‘get out in front’ of proposed policies and attempt to change them or kill them if necessary than it is to react to policies once they are adopted” or even to try to propose their own policies rather than wait for others to do so (Nownes, 2006: 46). By spreading lobbying activity across a larger number of governmental targets in the form of agencies and pieces of legislation (i.e., increasing lobbying breadth), firms are able to gather information about what possible policies may impact the firm from a diverse set of agencies. Thus, firms can plan accordingly so as to minimize possible damage and try to stop or modify the policy before it is enacted across expanded levels of government.
Moreover, by expanding lobbying breadth, a firm is also able to build relationships across a diverse group of government officials so that when issues arise, the contacts and resources are in place to address those issues (Hillman & Hitt, 1999). Importantly, this mechanism suggests that lobbying breadth not only insulates a firm from potential downside risk associated with adverse governmental actions but also assists in effectively protecting the firm from unfavorable perceptions by government entities that may arise as the firm engages in high levels of strategic risk. Thus, firms decide not only whether to try to gather information or curry influence with any one government official but also whether to expand the breadth of officials with potentially valuable information that should be influenced or considered so as to broaden the coverage of the firm’s protection.
This breadth of connections within distinct government agencies can also help firms to ease potential backlash from their actions and build political capital that represents a shield when firms commit misconduct or encounter failure. Indeed, because government officials are boundedly rational (March & Simon, 1958), all else equal, they may be favorable to a firm even in the face of some actions that violate their expectations if these firms have ties to government agencies. Thus, the political capital that provides the firm ties to government officials may also influence those officials’ heuristics built on prior perceptions of the firm (e.g., Antia, Kim, & Pantzalis, 2013; Kim, 2019). Lobbying breadth can essentially function as insurance during unexpected negative firm-specific events to avoid, or reduce, any loss arising from firm actions. This is so because political capital generated should reduce the overall severity of resentment by encouraging government officials to give the firm “the benefit of the doubt” (Godfrey et al., 2009). As such, we expect that as firms increase strategic risk, they will expand their lobbying breadth in an attempt to manage this risk by altering or at least staying aware of potential negative actions emanating from the government that may bring about negative outcomes associated with that risk.
Hypothesis 1: Strategic risk will be positively related to lobbying breadth.
The basis of our theorizing is that firms increase their lobbying breadth to protect against potential negative actions emanating from the government when the firm’s exposure increases due to proactively engaged strategic risk. Thus, if our theorizing is correct, as noted earlier, other aspects that also play a role in decisions on the amount of protection an entity may require should affect this decision (i.e., potential loss of key decision makers, political uncertainty that may exacerbate the likelihood of loss, alternative risk mitigation via slack).
Executive Personal Loss
As the potential of financial losses for key decision makers increases with strategic risk, the need to protect against loss likely increases as well. This possibility of loss is central to the executive-compensation and strategic-risk literatures. Indeed, as the key decision maker in a firm, CEOs have principal influence on the types of risk a firm takes and are heavily influenced by the manner in which they are paid (Finkelstein, Hambrick, & Cannella, 2009). For instance, executive-compensation research provides substantial evidence of how stock ownership may impact the amount of risk executives are willing to tolerate (Hoskisson et al., 2017). While some compensation components, such as stock options, have been recognized as a means to reduce risk aversion (e.g., Grant, Markarian, & Parbonetti, 2009; Hall & Murphy, 2002), stock ownership has generally been shown to decrease strategic risk (e.g., Devers, McNamara, Wiseman, & Arrfelt, 2008), because as risk increases, the possibility of downside stock movement also increases, threatening the value of owned stock. Simply put, as an executive’s ownership stake in the firm increases, thus reducing his or her diversification level, the executive may be inclined to protect that value by avoiding risk.
Prior studies find when CEOs are overinvested in their firm, they are likely to be prudent in risk taking. Denis et al. (1997), for example, finds that a CEO who has significant exposure to stocks of his or her firm was positively associated with diversification, which was attributed to a risk reduction strategy. Similarly, as noted earlier, Tufano (1996) finds that CEOs of gold-mining firms were likely to manage gold price risk when they were overinvested. As CEOs may perceive that their stock ownership increases the value they can lose, as stock ownership increases, CEOs will be inclined to expand the levels of lobbying breadth in order to ensure that they (a) are aware of potential governmental actions that may occur that can harm the firm and (b) can be active in managing governmental perceptions of the firm by building relationships with a wider array of government officials. In this way, we expect when firm strategic risk and CEO stock ownership both increase, the firm will be increasingly likely to “insure” the stock ownership by attempting to mitigate potential negative consequences of governmentally induced shocks or negative perceptions through CPA breadth.
That is, CEO stock ownership induces the CEO to be concerned with potential negative actions that may emanate from the government because any potential downfall from strategic risk hurts not only firm performance but also the CEO’s personal wealth. Hence, the need to protect stock value is likely particularly strong as strategic risk increases to higher levels, and as a result, executives would be likely to expand lobbying breadth to both preclude future potential negative actions across a broad breadth of the government and have a large array of government officials to call upon should a need arise when the CEO’s stock ownership increases. Such a move is akin to CEOs using risk management when highly exposed to company stock. Thus, we posit the following:
Hypothesis 2: The relationship between strategic risk and lobbying breadth will be stronger as CEO ownership increases.
Political Uncertainty
Beyond personal executive loss, political uncertainty, defined as uncertainty associated with changes in regulation or public policy and future government actions (Pástor & Veronesi, 2012), is also important when considering the coverage firms seek to protect against loss and potential governmental adverse reaction in the face of firm actions. We have argued that strategic risk increases a firm’s risk profile, and thus firms expand lobbying breadth in order to swing the potential odds of loss in their favor. Political uncertainty, however, increases the difficulty with which executives can accurately forecast future scenarios and/or odds of those scenarios (Alvarez & Barney, 2005). Thus, if lobbying breadth is a form of reducing these odds, then the amount of lobbying breadth needed should additionally be impacted by political uncertainty that influences the difficulty with which executives can make these predictions about future scenarios.
Specifically, political uncertainty is vital to managers, as political instability creates uncertainty as to the viability and effectiveness of risks taken, which may ultimately have a significant influence on the value of a firm (Butler & Joaquin, 1998). Sometimes labeled “policy risk” (e.g., Bradley et al., 2016), political uncertainty has been linked to weaker economic conditions and high volatility in stock prices (Pástor & Veronesi, 2013). Firms have also modified investment expenditures (Julio & Yook, 2012) and engaged in expense mitigation strategies, such as corporate downsizing (Kim, Pantzalis, & Park, 2012), in response to political uncertainty. While political uncertainty often peaks in presidential election years, it is also a key concern to firms surrounding ongoing public policy debates (Bradley et al., 2016).
Indeed, as previously argued, the scope of regulatory uncertainty is salient to firm outcomes in many industries (Kingsley, Vanden Bergh, & Bonardi, 2012), and firms are often exposed to idiosyncratic risks arising from changes in governmental regulations and policies (Jens, 2017). At the same time, relationships are particularly important to sway decisions in favor of the firm when needs arise (Ridge et al., 2017). In the context of a firm’s strategic risk, instability in the government and associated changes that can affect the likelihood and degree to which a strategic action is viable and effective are important as a firm increases risk. While all firm actions come with some level of risk (cf. Sanders & Hambrick, 2007), when trying to protect against declines from potential negative actions as strategic risk increases, firms will likely consider not only the strategic risks they take but also the political uncertainty and their need for potential contacts to call upon. Further, conditions of political uncertainty may enhance the firm’s perceived need to increase lobbying breadth since the firm may be less sure from where potential negative actions may arise. Indeed, studies show firms modify their political strategies, like campaign contributions and lobbying, in response to shifts in political uncertainty (Bradley et al., 2016). Thus, we argue that firms with high levels of political uncertainty, and consequently at increased mercy of shocks emanating from the government, will likely seek to expand “coverage” through greater lobbying breadth as strategic risk increases.
Hypothesis 3: The relationship between strategic risk and lobbying breadth will be stronger as firm political uncertainty increases.
Internal Risk Financing
A factor past research identifies as closely associated with components of firm risk is available slack (i.e., cash, or financial slack), which is important to firms in guarding against risk (Bourgeois, 1981). Firms with such slack are able to respond to environmental jolts effectively (Cheng & Kesner, 1997) and the absence of it reduces flexibility to adapt similarly (Miles, 1982). Specifically, Tan and Peng (2003: 1250) note such slack “may be employed as a buffer, which insulates the technical core of the organization from environmental turbulence.” Similarly, Jeffrey, Onay, and Larrick (2010) suggest having available slack creates a cushion that protects the firm against failure and thereby increases risk taking. In fact, studies by Hambrick and D’Aveni (1988) and Levinthal (1991) have shown that firm survival was significantly related to a firm’s available slack. Similarly, Acharya, Almeida, and Campello (2007) note a firm’s cash, or available slack, is tied closely to firm risk (see also Itzkowitz, 2013), while Bolton, Chen, and Wang (2011) consider a firm’s cash as part of a risk mitigation strategy. Relatedly, Kim and Bettis (2014) argue and find that cash, as a form of available slack, creates an adaptive benefit for firms that protects them from environmental uncertainties. In our study’s context, cash may serve as a form of internal risk financing because it represents a flexible form of slack, giving decision makers freedom to deploy it where needed (George, 2005).
Considering the buffering financial slack provides, we expect when a firm takes strategic risks but has an internal risk mitigation mechanism to help adapt to or manage new environmental changes (i.e., available slack), the firm has less need to guard against potential negative actions. This is because access to those slack resources are available to buffer against unfavorable governmental perceptions of the firm and harmful regulatory actions. Thus, the firm will be better prepared to weather new environmental shocks that stem from the government and may be less inclined to actively combat those shocks by broadening lobbying breadth. Indeed, some studies have even suggested that too much firm slack can lead to firms becoming complacent and suffering from inertia, thereby not proactively guarding against increased strategic risk (Agrawal, Catalini, Goldfarb, & Luo, 2018; Kim, Kim, & Lee, 2008). With high levels of available slack, a focal firm may experience less need to have wide coverage against government actions because its slack provides it flexibility in adjusting its strategic risks in the face of potential negative actions. Thus, we expect that available slack will weaken the relationship between strategic risk and lobbying breadth.
Hypothesis 4: The relationship between strategic risk and lobbying breadth will be weaker as available slack increases.
Method
Our sample is all publicly traded U.S. manufacturing firms in the Compustat and Execucomp databases from 2009 to 2016. We use manufacturing firms to be consistent with prior studies on firm strategic risk (Benischke, Martin, & Glaser, 2019; Schumacher, Keck, & Tang, 2020) and to aid comparison across firms given measures are often nonuniform across industries (Devers et al., 2008). Lobbying and political activity data were collected from the Center for Responsive Politics (CRP), an organization “that examines money and lobbying in elections, government actions, and public policy” (Ridge et al., 2017: 1147). CRP provides data for all lobbying directed toward the federal government. We obtain data on firm risk from RavenPack, which uses a patented algorithm to classify and aggregate press releases to enable scholars to capture firms’ strategic risk taking (e.g., Guo, Sengul, & Yu, 2020; Ye, O’Brien, Carnes, & Hasan, 2021). All other variables were collected from Compustat. To establish temporal spacing, independent and control variables were collected in time t, while dependent variables were collected at time t + 1. The final sample includes 485 firms and 3,119 firm-year observations.
Measurement
Our dependent variable, lobbying breadth, was measured following Ridge et al. (2017), using the number of government agencies and the number of legislative acts listed as being lobbied by a focal firm in a focal year. As the targeting of agencies and pieces of legislation are distinct due to their different outcomes (i.e., agencies interpret and enforce laws; legislation creates law) and because among lobbying firms the means and standard deviations between each target are statistically significantly different (agency mean = 6.49, SD = 5.92; legislation mean = 11.69, SD = 21.18; p = .000), we standardized each target of lobbying and summed to create the aggregate measure of lobbying breadth, rather than simply summing them. This procedure follows the prior literature that has used the standardized measure (Ridge et al., 2017) and is theoretically appropriate given the differences in the outcomes of the differential targets and statistical difference between the two. Further, in each year, there are thousands of potential legislative acts that could be targeted, while the number of agencies remains fixed (Opensecrets.org provides a count of 209 different agencies in which lobbying occurs). Thus, this measure provides an indication of the amount of the government the firm targeted with lobbying expenditures.
In measuring strategic risk, we utilized data sourced through RavenPack (e.g., Connelly, Tihanyi, Ketchen, Carnes, & Ferrier, 2017; Guo et al., 2020). “RavenPack uses a patented algorithm to classify and aggregate press releases” and “improves measurement precision by avoiding limitations of manual extraction (e.g., double-counting; missing information; fatigue and human error in coding)” as well as problems with other measures related to firm strategic moves, such as risk (Hill, Recendes, & Ridge, 2019: 811). Our measure is calculated as the number of four major types of strategic actions in a given year as reported by RavenPack (market expansion–related actions, legal actions, acquisitions, and strategic alliances). Actions such as these involve a significant commitment of financial or fixed assets and are difficult to implement, will not yield returns for a long time, and are hard to reverse (Connelly, Tihanyi, Certo, & Hitt, 2010). Thus, these actions provide both significant outlays and time delays to returns, making them risky investments with uncertain long-term implications, which directly relates to our theoretical exposition of strategic risk (as we detail later in the Robustness Tests section, our findings are robust to an alternate measure of strategic risk that is widely used as well).
CEO ownership is measured as the total value of outstanding firm stock owned by the CEO. To measure political uncertainty, we use a measure developed by Hassan, Hollander, Van Lent, and Tahoun (2019) in which they conduct a content analysis of quarterly earnings calls to calculate the amount of attention a firm places on political uncertainty. 3 The authors created risk measures using pattern-based sequence classification or bigrams in quarterly earnings calls to determine the proportion of earnings calls the firm devotes to political uncertainty. To do so, they identify and validate words related to discussion of regulatory uncertainty in quarterly earnings calls and divide the number of such words by total words used. Because our data are yearly, and this variable is created quarterly, we calculate political uncertainty as the average quarterly political uncertainty over the focal year’s quarterly measurements. Finally, available slack is measured as the value of cash on hand in the firm because this is a direct measure of uncommitted liquid resources and has been suggested to be a type of internal cushion for firm actions (Kim & Bettis, 2014).
We control for several factors that may influence the level of lobbying breadth in which a firm may engage. First, we control for firm size, measured as the firm’s sales (Devers et al., 2008), because firm size likely drives a portion of lobbying activity. We additionally include a measure of firm performance, measured as the firm’s return on assets, and performance volatility, calculated as the standard deviation of the quarterly return on sales, as this is an indicator of the fluctuation of performance over the prior four quarters. 4 The expenditures that a firm directs toward political activity also likely impact the breadth of lobbying engaged, and thus we include lobbying expenditures as the natural log of total federal lobbying expenditures (Ridge et al., 2017). Additionally, because we test the effects of available slack on the relationship between strategic risk and lobbying breadth, and because slack has been theorized to take several forms, we also control for recoverable slack, measured as the amount of firm inventory divided by firm sales (Miller & Leiblein, 1996).
We also account for government involvement with a firm in two different ways. First, as a firm is highly taxed, the necessity of lobbying activity may present itself; thus, we include a measure of taxes paid as the total taxes paid scaled by total sales. Second, the dependence the firm has on the government for revenue also likely impacts the level of lobbying the firm will engage. Thus, we include a measure of government dependence as the value of federal contract obligations secured by the firm divided by total sales (Abdurakhmonov et al., 2021). Data pertaining to firm government contract obligations were gathered from www.usaspending.org. We also include a variable for breadth of CEO political donations by counting the number of distinct political campaigns the focal CEO individually contributed using Federal Election Commission individual contributions data (Gupta & Wowak, 2017). We include CEO tenure as the number of years the CEO has been employed in the position in the firm to account for the decision making of the CEO in the context of firm risks (Hoskisson et al., 2017).
Analyses and Results
Descriptive statistics and correlations appear in Table 1, while results of our analyses appear in Tables 2 and 3. Not all firms lobby in every year, so our dependent variable is left censored in such cases. We thus used a tobit regression model to account for the censored dependent variable (Greene, 2000). To account for industry and temporal effects, we included industry and year dummies in all analyses. Given firms do not randomly choose their levels of strategic risk, estimated relationships between strategic risk and outcomes may be subject to endogeneity (Shaver, 1998). There may also be interdependencies as focal variables may simultaneously reinforce each other over time. As Semadeni, Withers, and Certo (2014: 1070) note, addressing endogeneity, which “occurs when an independent variable is correlated with the error term,” requires “a balancing act” as efforts to correct for endogeneity when it is not present and/or faulty approaches may beget “results that are inferior to those reported” without such corrections. Noting this quandary, Hill, Johnson, Greco, O’Boyle, and Walter (2021: 134) suggest (a) offering a “clear diagnosis” of the source and presence of endogeneity in a given study so as to avoid applying the “wrong treatments” (i.e., analytical tool) that “either will not address the actual cause [of endogeneity] or may even exacerbate the problem” and (b) presenting results from analyses attempting to address endogeneity alongside “naive results” without any correction (see also Semadeni et al., 2014).
Descriptive Statistics and Correlations
Note: Correlations > |0.0389| are significant at the .05 level.
Scaled by 100 for presentation.
Two-Stage Residual Inclusion (2SRI) Model
Note: N = 3,119 firm-year observations; 485 firms. Year and industry dummies included in all analyses.
P values in parentheses (two-tailed t tests).
p < .10
p < .05
p < .01
p < .001
Naive Model
Note: N = 3,119 firm-year observations; 485 firms. Year and industry dummies included in all analyses.
P values in parentheses (two-tailed t tests).
p < .10
p < .05
p < .01
p < .001
Following Hill et al. (2021), we first diagnose three possible sources of endogeneity in our study: (a) we do not observe the decision-making processes underlying both levels of risk and levels of lobbying, so an omitted variable may be present (Hambrick, 2007); (b) the relationship may be subject to simultaneity or reverse causality as lobbying breadth influences strategic risk and vice versa (Greene, 2000); and (c) the level of strategic risk is selected by the firm, giving rise to “selection of treatment” concerns (Shaver, 1998).
To address potential endogeneity from omitted variables, we assessed the impact threshold of a confounding variable (ITCV; Frank, 2000). Results suggest that to alter the inference, an omitted variable would have to exhibit a pattern of positive correlation with the dependent variable and negative correlation with the independent variable as well as that the strength of that pattern would have to be stronger than any predictor currently observed. As Hubbard, Christensen, and Graffin (2017: 2262) state, “Assuming that we have a reasonable set of control variables, this suggests the results are not likely driven by a correlated omitted variable.”
To address simultaneity and selection of treatment, we employ instrumental variables using two-stage residual inclusion (2SRI) that is amenable for tobit analyses (Terza, Basu, & Rathouz, 2008). In Stage 1, we regress all control variables, including an instrumental variable of CEO compensation, measured as the logarithm of CEO pay, on our independent variable and then save the residuals (which we label 2SRI) to include in our second-stage models (this first-stage model is presented in Appendix A). As per Hill et al. (2021: 124), sound instruments must meet two conditions: “First is relevance, which implies that an instrumental variable is related to the endogenous predictor [and] can be tested directly (Stock, Wright, & Yogo, 2002). Second is exogeneity, which implies that [the instrumental variable] is uncorrelated with the residual.” To assess relevance and exogeneity, we first conducted a Montiel-Pflueger robust weak-instrument test. The effective F statistic of this test is 60.497, which exceeds the critical value of 37.418 at a confidence level of 5%, suggesting that our instrumentation is strong and valid (i.e., relevant).
Next, we conducted a Davidson-MacKinnon test of exogeneity on a fixed-effects regression estimated via our instrumental variable. The null hypothesis of this test is that any endogeneity among the variables would not have deleterious effects, and thus if the null is not rejected, the test indicates endogeneity is not biasing our results. Our results indicate a statistic of 0.76 with a p value of .3833, thus not rejecting the null and indicating our model is asymptotically consistent and endogeneity does not negatively bias our results. We report the results of analyses that address possible endogeneity using 2SRI in Table 2 and the “naive results without any adjustment in Table 3. The results are relatively similar across both sets of analyses, enhancing the robustness of our conclusions to efforts to attenuate endogeneity (Hill et al., 2021; Semadeni et al., 2014).
Beyond concerns of endogeneity, as fixed effects are not specified for tobit analyses, we approximate a fixed-effects estimator using the hybrid approach. Specifically, following Bliese, Schepker, Essman, and Ployhart’s (2020: 81) recommendation for instances where a researcher can recover the “gold standard,” we employ the hybrid approach (Raudenbush, 2009) and include the firm mean level of strategic risk as a control variable and then subtract the firm-level mean from our strategic risk variable before including it in the model. Notably, our results are robust with or without this adjustment, offering support for our findings.
Parameter estimates for lobbying breadth appear in Tables 2 and 3. Model 1 includes control variables only. Interestingly, when considering Model 1, available slack is positively related to lobbying breadth (p < .001), supporting prior arguments about slack enabling firms to allocate resources toward a CPA strategy (Schuler, 1996; Schuler, Schnietz, & Baggett, 2002). Moreover, as expected, government dependence positively relates to a firm’s level of subsequent lobbying breadth (p < .01).
Hypothesis 1 argues strategic risk will be positively related to lobbying breadth. To test this hypothesis, we add strategic risk to Model 2 in both Tables 2 and 3. The results support Hypothesis 1, as the coefficient for strategic risk is positive and significant in both sets of analyses (p < .05) with a 95% confidence interval of [0.02, 0.23] from Model 2 in Table 2, suggesting that as firms increase strategic risk, they additionally increase lobbying breadth, supporting our theorizing that lobbying breadth is increased as an insurance mechanism for increased strategic risk.
From a practical perspective, using results in Model 2 of Tables 2 and 3, the findings of Hypothesis 1 indicate that for every one-standard-deviation unit increase in strategic risk, lobbying breadth is increased by approximately 10%. Given the mean targeted number of agencies and pieces of legislation, this translates to a little less than one agency (0.649) and over one (1.169) piece of legislation would be targeted as strategic risk increases by one standard deviation. With the cost of lobbying any single agency or piece of legislation at times being quite high in many instances, as firms can spend untold millions on single issues (Nownes, 2006), as well as the left-censored nature of data, an average of a 10% increase can be quite substantial, in many cases exceeding the amount expended on even the highest-paid individual in the firm. Moreover, as it may be untenable to interpret a main effect in the presence of moderators, we follow Busenbark, Graffin, Campbell, and Lee (2021) and assess the effect at thresholds from very low to very high for each moderator following Model 6, in which all moderators are included. We find that the relationship in Hypothesis 1 is statistically significant across the entire range of values for each moderator.
In Hypothesis 2, we argued the vulnerability a firm’s key decision maker (i.e., the CEO) has to financial loss in the form of CEO ownership will strengthen the relationship between strategic risk and lobbying breadth. This hypothesis is offered some support in Model 3 of Tables 2 and 3. Specifically, the interaction of CEO ownership and strategic risk is positive and significant in Model 3, albeit at traditionally lower thresholds in Table 2 (p = .10) than in Table 3 (p = .04), with the 95% confidence interval in Table 2 being [−0.005, 0.058]. Nonetheless, the result is practically meaningful as the naive model suggests that the increase of lobbying breadth is 3% higher when strategic risk and CEO ownership are high (one standard deviation above the mean), while the more conservative model in Table 2 suggests the increase is approximately 1%.
In Hypothesis 3, we argue political uncertainty will strengthen the relationship between strategic risk and lobbying breadth. Results offer mixed support, as the interaction term is in the specified direction yet does not reach significance (p = .62) in Model 4 of Table 2 but does reach statistical significance (p = .01) in Model 4 of Table 3. The 95% confidence interval for the interaction in Table 2 is [−0.07, 0.11] and [0.02, 0.12] in Table 3. This finding suggests that our results are contingent on the degree to which endogeneity is present in our analyses (Hill et al., 2021). As there is a reason to suspect endogeneity is present, we urge caution in viewing the results for Table 3, as even low levels of endogeneity can result in inaccurate estimates (Semadeni et al., 2014).
Hypothesis 4 states that available slack reduces the relationship between strategic risk and lobbying breadth, arguing an internal mechanism through which a firm may insure itself against shocks minimizes the need for external insurance in the form of lobbying breadth. To test this hypothesis, we add the interaction of available slack and strategic risk in Model 5. The results provide support for Hypothesis 4, with Model 5 suggesting a significant negative interaction (p = .04 in Table 2; p = .00 in Table 3) and the 95% confidence interval in Table 2 being [−0.4, −0.00]. Similarly, these results are practically significant as the results from Table 2 suggest that the increase in lobbying breadth at high levels of strategic risk is approximately 5% higher when available slack is one standard deviation above the mean relative to at the mean. Further, Model 6 includes all interactions simultaneously and provides similar results to the individually specified interaction models, albeit with weaker results for available slack in Table 2 and nonsignificant results for CEO ownership and political uncertainty in Table 3.
Robustness Tests
Beyond presenting both our findings in models that seek to attenuate endogeneity (Table 2) and the naive results without such attenuation (Table 3), we conducted additional robustness tests as well. First, we tested an alternate strategic risk measure widely used in prior research (e.g., Devers et al., 2008; Mannor, Wowak, Bartkus, & Gomez-Mejia, 2016). This measure is created through a factor analysis of three components of the firm’s financial statement as a proxy for strategic risk: research-and-development spending, capital expenditures, and long-term debt. Each component is widely considered to be an important strategic factor that contributes to firm risk (Kish-Gephart & Campbell 2015; Martin, Gomez-Mejia, & Wiseman, 2013). The results of our theoretical model remained substantively the same with this alternate measure of strategic risk (see Appendix B).
Second, although our measure of lobbying breadth follows prior research (Ridge et al., 2017) and is theoretically relevant, as an alternate measure of our dependent variable, we use the simple summed measure of targets (i.e., agencies and bills). The results of this alternate model tested with a Poisson model as shown in Appendix C, given the dependent variable is a count measure, are relatively similar to our main model for strategic risk and available slack. However, the moderation of CEO ownership and political uncertainty do not reach significance (p > .10).
Finally, our logic is that CEOs’ stock ownership acts as a moderator, but it is also plausible that stock ownership affects strategic risk—either positively to earn gains (Jensen & Murphy, 1990) or negatively as to avoid loss (Wiseman & Gomez-Mejia, 1998)—and therein lobbying breadth so that our model is mediated. While there is reason to expect that a direct effect of stock ownership on strategic risk in isolation is too simplistic by not accounting for contingent factors—such as other aspects of compensation, firm performance, and governance (among others)—that would effect whether and how stock ownership affects strategic risk (for a review, see Devers et al., 2008), we also test a mediated relationship with stock ownership and strategic risk measured in time t and lobbying breadth in time t + 1. Consistent with evidence about the contingent effect of stock ownership on strategic risk, we did not find a direct relationship, and we also did not find support for a mediated effect. We nonetheless note that the relationships between stock ownership and firm risk are complex and additional contingencies may be at play, and though the results of ITCV suggest our principle findings are likely robust to omitted potential interactions, future research can shed further insight on this issue.
Discussion
In this article, we sought to examine the risk-mitigating role of CPA breadth by departing from prior research that overwhelmingly focuses on the outcomes of CPA breadth (Ridge et al., 2017) and shift scholarly attention to critical internal firm actions, such as strategic risk. We argued that because strategic risk increases a firm’s exposure to potential negative actions in the external environment, and specifically from one of the most influential external entities (i.e., government; Abdurakhmonov et al., 2021), firms employ lobbying breadth—the largest form of CPA—as a risk management strategy that creates a type of political insurance coverage by being able to influence and/or garner information from an increased breadth of the government as well as build political capital to minimize government stakeholder disapproval of firm actions.
Further, we tested the underlying assumptions of an insurance or protective framing by also considering the level of CEO financial vulnerability to risk in the form of stock ownership as well as external hazards, such as firm political uncertainty, and internal risk financing, such as available slack. Specifically, we argued that if lobbying breadth functioned similarly to insurance coverage, CEOs vulnerable to risk through stock ownership or firms with increased levels of political uncertainty would find it important to insure against potential negative actions as strategic risk increases. However, if lobbying breadth is an external insurance mechanism, then firms with an increased level of internal risk management (i.e., high levels of available slack; Kim & Bettis, 2014; Tan & Peng, 2003) will be less likely to seek external risk management, weakening the relationship between strategic risk and lobbying breadth. Overall, we find support for these relationships in our sample of manufacturing firms. Cumulatively, these results substantiate our theory of CPA breadth as a type of insurance mechanism emanating from firm strategic risk and offer a number of contributions to theory and practice.
First, we contribute to the CPA literature by empirically testing theory about lobbying breadth acting as a risk management tool from an organization’s perspective. While prior research alludes to the possibility that targeting an expanded amount of government is a risk mitigation mechanism for firms (Cooper et al., 2010; Hadani et al., 2017), we believe this is the first attempt to empirically test a broad theoretical model of lobbying breadth as a form of insurance in a wide sample of public firms. In contrast to past studies that predominantly utilized surveys to ask respondents about the buffering nature of their strategy (Meznar & Nigh, 1995) or alluded to possible hedging or risk mitigation of firm political activity breadth (Hadani et al., 2017), this article directly ties and tests theory about the need to insure the firm against potential negative actions through lobbying breadth, directly measuring government involvement. Specifically, we theorize and empirically demonstrate the relationship between strategic risk and the extent of government agencies and pieces of legislation targeted by a firm. In so doing, we advance understanding of why firms expand the amount of government they lobby despite possibly reaching a point of diminishing returns (Ridge et al., 2017). These findings provide direct empirical evidence that CPA should be considered not only as a rent-seeking tool but also as a tool to mitigate external risks.
Second, by considering lobbying breadth as a risk management strategy, we offer avenues for a new stream of research. While prior research identifies corporate strategies as outcomes of firm risk exposure, to our knowledge, we are the first in the management field to directly and empirically tie firm lobbying breadth to strategic risk as an insurance mechanism or a risk management strategy. This not only extends the limited research on firm risk management in the management literature but also advances our understanding of the scope of risk management strategies available for risk-averse executives (Godfrey et al., 2009). In so doing, we contribute to the study of an expanding portfolio of risk management strategies beyond traditional financial tools, such as forward or future contracts (Bromiley et al., 2015), CSR (Jia et al., 2020), and impression management (Graffin et al., 2016), directed at alleviating risks and stakeholder disapproval from firms’ actions (Godfrey et al., 2009). It is likewise important to consider the role of “moral hazard” in insurance and risk management portfolio of firms. While prior research has noted that insurance encourages risk taking, we find that firms purposefully expand their government targeting through lobbying as they expand their strategic risk. Thus, we advance the idea that lobbying breadth being a hedge may represent the “cost” or “price” for a firm’s expanded risk taking.
Finally, this article extends research on the interplay between the amount that firms devote to CPA directed at government and how those funds are spread across government. We provide a critical contribution to the lobbying literature by suggesting that the targeting of lobbying expenditures is at least as important, if not more so, as just how much is expended. By showing that firms with increasing levels of strategic risk also consider how this greater exposure to potential negative actions from the government may necessitate expanded nonmarket coverage through lobbying breadth, we provide evidence of an integrated consideration of not just that firms engage in lobbying expenditures but that they do so with consideration of how to expend their resources. Such findings are also of value to practice, as they offer insight into the use of lobbying breadth as a risk management strategy along with other, traditional risk mitigation tools. Further, because firms face resource allocation constraints (Penrose, 1959), gaining a better understanding of the firm-specific antecedents of lobbying can help practicing managers make better decisions related to firm political activities and avoid wasteful expenditures by “throwing money at the issue.”
In light of our findings, this study raises important questions about how lobbying breadth might be considered in the overarching portfolio of risk management strategies that firms employ. For example, future research may be well served to examine how firms that increase CPA breadth may approach other risk management strategies. One avenue for potential investigation is whether CPA breadth and other risk management techniques serve as substitutes or complements in a firm’s risk management portfolio. It could be possible that firms that expand CPA breadth may feel less inclined to expend on CSR, given their relatedness as nonmarket strategies (Dorobantu, Kaul, & Zelner, 2017; Mellahi et al., 2016) and well connectedness to the branches of government. It is also feasible that firms pursue these risk management techniques in concert, given that firms may reach broad stakeholder groups utilizing both strategies. This interchangeable use of these two nonmarket techniques may vary based on CEO ideological leanings. For example, liberal CEOs, given their preference for CSR, may be inclined to use CSR as an insurance strategy, while conservative CEOs may prefer CPA breadth (cf. Chin, Hambrick, & Treviño, 2013). Research examining these risk management strategies in tandem will not only serve “to integrate and synthesize these two strands of the nonmarket strategy literature” that reflect similar risk mitigation techniques (Mellahi et al., 2016: 144) but also contribute to political-ideology literature concerned with the use of nonmarket strategies, such as CSR and CPA (Gupta, Briscoe, & Hambrick, 2017; Unsal, Hassan, & Zirek, 2016).
However, lobbying breadth may also create impression management challenges for companies. Given the rise in social activism as an influence on firms (McDonnell & Werner, 2016), it is possible that lobbying breadth, while alleviating governmental uncertainty, runs a risk of unfavorable public perceptions if stakeholders are predisposed to be suspicious of such activity. Future studies that seek to bridge CPA research with other firm strategies related to managing stakeholder perceptions, such as impression management, may investigate how lobbying breadth, potentially viewed as less principled by stakeholders, may influence firm efforts to build reputation. To this end, stock ownership may also shed further insight. Our results suggest stock ownership does not directly influence strategic risk and therein lobbying breadth, but future research could investigate whether the contingent effects of stock ownership (Devers et al., 2008) are used a priori in concert with other governance or compensation mechanisms.
Another interesting avenue that future research could investigate is how lobbying breadth is modified as the environment changes. While firms may display variance in CPA breadth within the same environment (Ridge et al., 2017), increasing corporate activism may make lobbying breadth salient for some firms even in the face of minimal strategic risk. For example, liberal CEOs may have an expanded need for CPA breadth given their tendencies in taking a public stand on socially relevant issues, while this may be less necessary for conservative-leaning firms (Chin et al., 2013; Gupta et al., 2017). Further, growing partisanship in politics, especially in the United States, may make breadth of contacts within the government a salient consideration for firms. This could be true because of changing dynamics in government policies from one administration to another that may make the need for contact across a wide breadth of government more necessary compared with other risk management techniques. Another potential intriguing question is whether the benefits of lobbying breadth materialize when companies lobby across political party lines. While it is feasible that lobbying across party lines may help to alleviate political uncertainty due to power shifts in government, the benefits of lobbying breadth may decrease or even reverse as a firm “spreads its bets” by expanding its lobbying activity across an increasing number of party officials if a country’s political system is characterized by growing partisanship, such as in the United States. Future longitudinal studies that relate firm lobbying breadth to polarization in politics may be especially helpful in answering this critical question and bridge organizational and political science.
In this study, we focus on lobbying as a representation of CPA, as it is the primary and largest form (Hadani & Schuler, 2013; Lux et al., 2011) and the variables are publicly available and consistent across firms, allowing for comparability. However, while we provide robust evidence for our theorizing of lobbying breadth as political insurance, other forms of CPA may act as an insurance strategy as well. As such, future research may benefit from doing a qualitative study that focuses on corporate political strategies that are hard to capture through quantitative data. For example, top management teams of firms with increased levels of strategic risk may devote more attention toward understanding the regulatory agenda and/or potential political changes than top team members of firms with significantly lower strategic risk. Thus, strategic risk may increase not only the amount of government agencies and legislative acts a firm targets but also the amount of attention executives devote to political events. Future research may also be well served to uncover how CPA breadth influences government actions that impact a firm. While research examining benefits from CPA is inconclusive partly due to difficulty of measuring proximate benefits and access to data (Ridge et al., 2017), understanding how CPA breadth actually impacts stopping and/or promoting useful legislation through qualitative study and interviews of government officials is an area we see as an interesting avenue for future research.
While this article focuses on the interaction of strategic risk, CEO stock ownership, political uncertainty, and available slack as the predictors of the amount of government lobbied, there may be other interactive effects that can add nuance to our understanding of lobbying breadth as a risk management strategy. Thus, future research might be well served by investigating other critical factors that may influence this relationship. One such potential theoretically intriguing interactive effect that seems worthy of investigation is the effect of the institutional environment (DiMaggio & Powell, 1983). For example, aspects of the institutional environment that are idiosyncratic to the United States may affect whether lobbying is considered as a risk mitigation instrument. While we expect that the mechanisms relating strategic risk and political activities will operate in a similar manner across distinct institutional environments, in developing countries where institutional environments are either “incomplete” or “captured” (Dorobantu et al., 2017: 115), the effect of political strategies as risk mitigation may be stronger than in the United States or other advanced economies, where the institutional environment is more predictable.
Footnotes
Appendix
Poisson Model of Summed Target Measure
| Variable | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| Firm size | 0.00† (0.10) | 0.00† (0.08) | 0.00† (0.10) | 0.00* (0.04) |
| Recoverable slack | 0.30 (0.19) | 0.30 (0.20) | 0.30 (0.21) | 0.32 (0.17) |
| Taxes paid | 0.06 (0.89) | 0.04 (0.92) | 0.05 (0.90) | −0.01 (0.98) |
| Performance volatility | −0.09 (0.15) | −0.09 (0.15) | −0.09 (0.16) | −0.08 (0.17) |
| Firm performance | −0.42* (0.01) | −0.42* (0.01) | −0.42* (0.01) | −0.44** (0.01) |
| CEO political donation | −0.00 (0.44) | −0.00 (0.43) | −0.00 (0.46) | −0.00 (0.48) |
| Lobbying expenditures | 0.13*** (0.00) | 0.13*** (0.00) | 0.13*** (0.00) | 0.13*** (0.00) |
| Government dependence | 0.34† (0.05) | 0.34† (0.05) | 0.35* (0.05) | 0.39* (0.03) |
| CEO tenure | 0.00 (0.28) | 0.00 (0.28) | 0.00 (0.29) | 0.00 (0.30) |
| Available slack | 0.09* (0.03) | 0.09* (0.03) | 0.09* (0.03) | 0.17** (0.00) |
| CEO ownership | 0.00 (0.86) | −0.00 (0.43) | 0.00 (0.88) | 0.00 (0.88) |
| Political uncertainty | −0.10 (0.21) | −0.10 (0.20) | −0.12 (0.23) | −0.11 (0.20) |
| Strategic risk | 0.00* (0.04) | 0.00† (0.07) | 0.00 † (0.09) | 0.00** (0.00) |
| Strategic Risk × CEO Ownership | 0.00 (0.21) | |||
| Strategic Risk × Political Uncertainty | 0.00 (0.78) | |||
| Strategic Risk × Available Slack | −0.00* (0.02) | |||
| Constant | −0.00 (0.99) | −0.00 (1.00) | −0.00 (0.99) | −0.04 (0.91) |
| χ2 | 886.68 | 887.39 | 886.62 | 890.18 |
| Change in χ2 from control model | 4.10* | 5.65† | 4.18 | 9.90** |
| Change in χ2 from Model 1 | 1.54 | 0.08 | 5.66* |
Note: N = 3,119 firm-year observations; 485 firms. Year and industry dummies included in all analyses. P values in parentheses.
p < .10
p < .05
p < .01
p < .001
Acknowledgements
The authors would like to thank editor Brian Connelly and two anonymous reviewers for their helpful comments and guidance throughout the review process.
