For the last 25 years, David Baron’s (1995) nonmarket integrated strategy model (NISM) on the benefits of combining market and nonmarket activities has stood tall in the management field. His NISM was among the first to note that the companies most adept at aligning and combining their market activities with nonmarket activities, specifically corporate political activities or CPA (Aggarwal, Meschke, & Wang, Reference Aggarwal, Meschke and Wang2012; Funk & Hirschman, Reference Funk and Hirschman2017; Hillman, Keim & Schuler, Reference Hillman, Keim and Schuler2004; Hillman, Zardkoohi, & Bierman, Reference Hillman, Zardkoohi and Bierman1999; Katic & Hillman, Reference Katic and Hillman2022; Keim & Baysinger, Reference Keim and Baysinger1988; Lux, Crook & Woehr, Reference Lux, Crook and Woehr2011; Shaffer, Quasney & Grimm, Reference Shaffer, Quasney and Grimm2000), could realize better performance than firms that were less deliberate about or neglected such combinations. The NISM argues that nonmarket activities like a company’s lobbying of government officials about legislation or regulation (Drutman, Reference Drutman2015; Hall & Deardorff, Reference Hall and Deardorff2006; Holburn & Vanden Bergh, Reference Holburn and Vanden Bergh2014) or using politically tied directors (Hadani & Schuler, Reference Hadani and Schuler2013; Hillman, Reference Hillman2005) complement market activities, such as research and development (R&D), marketing, and mergers and acquisitions (M&As), in ways that exploit firm-specific market exchange imperfections (Oberholzer-Gee & Yao, Reference Oberholzer-Gee and Yao2018) and otherwise generate competitive advantages (Baron, Reference Baron1995; Funk & Hirschman, Reference Funk and Hirschman2017; Jia & Mayer, Reference Jia and Mayer2016). The logic is simple and appealing, especially if we consider firms operate in an open system where actors, including nonmarket ones, bring pressures and control resources critical to firms’ task environment (Hillman & Hitt, Reference Hillman and Hitt1999; Pfeffer & Salancik, Reference Pfeffer and Salancik1978).
The empirical basis for the NISM, however, is at best incomplete. Baron (Reference Baron1995, Reference Baron1997) used several cases to illustrate the model, such as how Calgene, a biotech agriculture company, used CPA to influence federal and state standards that enabled its research trials to continue and eventually led to regulatory approvals of their products for consumer use (although they never had commercial success and were eventually acquired by Monsanto) and how Kodak undertook political activities to break into the Japanese retail market for their (then robust) film business. Baron’s Kodak example further illustrates the fact that some successful past examples of integration may be obsolete and less relevant in today’s strategic competitive environment. Other scholars have tested only parts of the NISM or in restricted situations such as for heavily regulated industries. For example, Holburn and Vanden Bergh (Reference Holburn and Vanden Bergh2014) examined whether utilities firms that used CPA were more likely to receive regulatory approvals for mergers and acquisitions. While they found that CPA led to favorable decisions, they did not study the direct financial benefits of those regulatory outcomes. Furthermore, most empirical studies focusing on the CPA to financial performance link, rarely explore how CPA combines with market activities (e.g., Aggarwal et al., Reference Aggarwal, Meschke and Wang2012; Ansolabehere, de Figueiredo, & Snyder, Reference Ansolabehere, de Figueiredo and Snyder2003; Coates, Reference Coates2012; Cooper, Gullen, & Ovtchnikov, Reference Cooper, Gullen and Ovtchnikov2010; Hadani, Reference Hadani2012; Hadani & Schuler, Reference Hadani and Schuler2013; Hillman & Hitt, Reference Hillman and Hitt1999; McDonnell & Werner, Reference McDonnell and Werner2016; Smith, Reference Smith2000; Stratmann, Reference Stratmann1998; Witko, Reference Witko2011). The closest study to what we undertake is by Shaffer, Quasney, and Grimm (Reference Shaffer, Quasney and Grimm2000), who examined how aspects of market and nonmarket (CPA) strategy—independently—affected the performance of commercial airline firms, rather than testing integration per se. Overall, the evidentiary base of the NISM is thin, leaving open the critical question of integration (Hillman et al., Reference Hillman, Keim and Schuler2004; Lux et al., Reference Lux, Crook and Woehr2011) and suggesting the need for the evaluation and validation of the integrated strategy model.
Indeed, the NISM shows certain theoretical weaknesses. First, scholars have noted that the theory connecting CPA to market strategy is underdeveloped (Funk & Hirschman, Reference Funk and Hirschman2017; Holburn & Vanden Bergh, Reference Holburn and Vanden Bergh2014). Holburn and Vanden Bergh (Reference Holburn and Vanden Bergh2014, p. 450) wrote: “While firm performance depends on the design and implementation of both market and nonmarket strategies, the majority of research in these fields has developed separately, implicitly treating market and nonmarket components independently.” Specifically, the NISM inadequately acknowledges how the significant differences between the market and the nonmarket institutional environments affect uncertainty (Bonardi, Reference Bonardi2011; Drutman, Reference Drutman2015; Hadani, Bonardi, & Dahan, Reference Hadani, Bonardi and Dahan2016), power dynamics (Hillman & Hitt, Reference Hillman and Hitt1999), performance (Bonardi, Reference Bonardi2004, Reference Bonardi2011; Hadani & Schuler, Reference Hadani and Schuler2013; Hillman & Hitt, Reference Hillman and Hitt1999; Lux et al., Reference Lux, Crook and Woehr2011; Schuler & Rehbein, 1999), and overall strategic logic and dynamics (Drutman, Reference Drutman2015; Hansen, Reference Hansen1991; Smith, Reference Smith2000). Certain features of the “political marketplace” (Baumgartner et al., Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009; Bonardi, Reference Bonardi2008, Reference Bonardi2011; Bonardi & Keim, Reference Bonardi and Keim2005; Drutman, Reference Drutman2015; Kersh, Reference Kersh2002; Lenway et al., Reference Lenway, Schuler, Marens, Werner and Green2022), including uncertainty, causal ambiguity, and the strong reliance on lobbyists as intermediaries (Drutman, Reference Drutman2015; Hadani & Schuler, Reference Hadani and Schuler2013; Hall & Deardorff, Reference Hall and Deardorff2006), necessitate astute political skills and capabilities—including processes and structures to share political information with market decision-makers—that some firms may not possess, significantly challenging the integration of CPA and market activities (Bonardi, Reference Bonardi2011; Hillman & Hitt, Reference Hillman and Hitt1999; Kersh, Reference Kersh2002).
Second, NISM research underplays the temporal dynamics dimensionFootnote 1 of integration and its potential to highlight if, how, and when integration can occur. While firms tilt toward market activities by default (Hadani & Schuler, Reference Hadani and Schuler2013; Porter, Reference Porter1985), the temporal sequencing of CPA and market strategies should matter. For example, Baron’s (Reference Baron1997) case studies involve firms that initially focused on market strategies and later used CPA to legitimize their products or enter new markets. However, prior political experiences may also affect integration. Drutman’s (Reference Drutman2015) work strongly indicates that escalation of commitment may occur once firms become politically active, divorcing CPA from market needs (see also Schuler, Reference Schuler1996). However, little research addresses the temporality of nonmarket versus market strategies.
Lastly, NISM research is unclear on how to capture integration, often using combinations of one or two forms of CPA with one type of market activity (Baron, Reference Baron1995, Reference Baron1997; Bonardi, Reference Bonardi2004; Holburn & Vanden Bergh, Reference Holburn and Vanden Bergh2014). While we agree these studies capture certain aspects of integration, we offer a more holistic and comprehensive approach showing how various combinations of CPA and market activities affect different measures of performance.
To sum up, because the NISM viewpoint lacks in-depth empirical exploration and because CPA research has highlighted the differences between the political and market environments (Drutman, Reference Drutman2015; Hadani et al., Reference Hadani, Bonardi and Dahan2017; Gehl & Porter, Reference Gehl and Porter2020), we have a strong hunch that the prevailing NISM viewpoint does not fully account for the complexity of integrating political and market activities. Thus, the purpose of this abductive study is to discover the impact of firms’ integrating CPA and market activities on their financial performance. We ask several critical and related questions. First, what are the implications of firms’ integrating different types of political and market activities on financial performance? Second, is prior experience with different types of CPA beneficial or detrimental to integration with firms’ ongoing focus on market strategies? Third, we explore the configuration of political and market activities relative to industry benchmarks. In analyzing these critical questions, we throw a wide net by including an array of CPA activities such as campaign contributions, lobbying expenditures, executive-level revolving door connections, and the presence of a Washington DC office, alongside a broad spectrum of market activities such as R&D expenditures, marketing expenditures, and M&A attempts, and common measures of firm-level performance. We also explore alternative temporal sequencing of CPA and market activities, as well as analyze both firm and industry deviation level measures of integration and other strategic measures, post hoc.
We seek answers to these questions through an abductive reasoning approach: we examined the political and market activities of over 2,200 publicly traded small-, mid-, and large-cap firms from the S&P 1500 from 1998 to 2018, resulting in a sample of about 18,000 to 19,000 firm-year observations (depending upon the model specification) of politically and nonpolitically active firms. Using several statistical techniques, we were unable to find sufficient evidence that supports the conclusion of the NISM. Most models yielded either a nonsignificant relationship between integration and performance or a significant negative relationship; only a handful of instances showed a significant and positive relationship with performance. Additional post hoc analyses produced comparable results. Overall, our in-depth analyses show no clear financial performance benefits from integration, with several important and instructive caveats that can inform executive decision-making.
Instead, we find a nuanced and complex picture. The impact of integrating CPA and market activities on firm-level performance measures varies by the type of CPA, type of market activity, the type of performance measure, deviations from industry norms, and the temporal sequence of CPA and market activities. Most configurations yielded either no effects or negative effects on market value. For a few configurations, however, especially involving revolving door connections as the political strategy and R&D as the market strategy, integration led to small improvements in market value. In our post hoc theorizing, we discuss how attention to strategic control limitations, policy opportunity windows, visibility of market activities, and challenges to integration mimicry may be important elements for future NISM theorizing.
The untested assumptions about integration
The core of the NISM is that firms perform the best when they combine or integrate their market and nonmarket strategies. Baron (Reference Baron1995, pp. 47–48) writes: “A market strategy is a concrete pattern of actions taken in the market environment to create value to improve economic performance. …A nonmarket strategy is a concerted pattern of actions taken in the nonmarket environment to create value to improve overall performance…” Oberholzer-Gee and Yao (Reference Oberholzer-Gee and Yao2018) lay out a market failures framework to demonstrate how a firm’s market actions affect market structures and the likelihood of subsequent government involvement in that market via regulations. Integration can mitigate some regulatory risks.
One manifestation of actions is via investments. Thus, we might expect firms that invest heavily in both sets of activities to be best able to integrate. In contrast, firms that invest considerable resources in the market side of their business but little on the political side are unlikely to integrate. Likewise, firms investing heavily in political activities, but with little devotion to market activities, are also unexpected candidates for integration.
The NISM differs, however, from other accounts that claim firms’ political investments destroy financial value. For example, while Callander et al. (Reference Callander, Foarta and Sugaya2022) recognize that political and market investments are linked, they note that firms use their political power to attempt to pass legislation or regulations that remove competition from the marketplace. These authors explain that such activities make firms less efficient than if they had directed more investment into market investments. Other researchers argue that political investments are “directly unproductive profit seeking activities” (Bhagwati, Reference Bhagwati1982; Stigler, Reference Stigler1971); inferior to market investments in creating value. Additionally, NISM downplays the possibility that managers may engage in CPA for their own personal benefit and not that of their firm, the agency problem, damaging performance (Hadani & Schuler, Reference Hadani and Schuler2013). These views differ from Baron’s where integration is central for managers and firms to create value for the firm because CPA and market activities are mutually reinforcing.
The integrated strategy model relies on several conditions for it to work satisfactorily. First, firms must have the willingness and capacity to scan the socio-political environment (Bazerman & Watkins, Reference Bazerman and Watkins2004) to correctly analyze the social and political forces that might affect business or business practices. Second, firms need to develop relationships with government officials and other external actors to exchange information, obtain access, and otherwise exercise influence on policies (Baumgartner et al., Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009; Mizruchi, Reference Mizruchi1992; Smith, Reference Smith2000). Third, firms must establish the structures and processes, sometimes held in “outward facing” units like public and government affairs departments (Schuler & Rehbein, 1999), to blend nonmarket information with market information and other firm competencies into a coherent integrated strategy. Fourth, firms need to identify the linkages among their market needs and related activities and their arsenal of CPA activities, exhibiting a deep understanding of how to use the latter in the service of the former; a challenging task at the best of times given the nature of the political arena (Baumgartner et al., Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009; Drutman, Reference Drutman2015). Lastly, firms must have the structures, processes, and culture to resolve internal conflicts that can hamper the effective integration of political activities across their organization (Shaffer & Hillman, Reference Shaffer and Hillman2000).
Assuming these organizational structures and processes are in place, the logic of the NISM model supports the first, omnibus hypothesis of this study:
Hypothesis 1: Firms that invest more resources in CPA and market activities will outperform firms that invest fewer resources in CPA and market activities.
Yet, all forms of CPA may not have the same potential to integrate successfully with market strategies. While the NISM scholarship does not explicitly distinguish how different forms contribute to effective integration, other CPA theorizing suggests such differentiation. Many scholars consider monetary investments in campaigns such as by political action committees (PACs) as providing access to other activities, such as lobbying (Clawson, Neustadtl, & Scott, Reference Clawson, Neustadtl and Scott1992, Kalla & Broockman, Reference Kalla and Broockman2016). Hillman and Hitt (Reference Hillman and Hitt1999) called such activities transactional, contrasting them with relational activities that provide greater companies with more opportunities to exchange information with government officials (Hillman, Reference Hillman2005). As relational activities allow firms to share strategic goals with and receive information about policy directions from government officials, these have the greatest integration potential for maximum performance.
Firms conduct relational activities through various channels. They sometimes contract with external lobbyists and consultants, something we capture in our financial CPA (FIN CPA) variable. While external lobbyists might convey a company’s information with relative fidelity, Drutman (Reference Drutman2015) shows that they also balance their own needs with those of their clients (see also Kersh, Reference Kersh2002). Another avenue is that firms hire in-house lobbyists, a variable we capture with the DC office. Finally, firms may have their executives depart for government service or receive former government officials into their firms—revolving door connections (RDC). RDC aim to create relationships between the firm and policymakers, exchange information, and monitor policies and gather intelligence (Drutman, Reference Drutman2015; Hillman & Hitt, Reference Hillman and Hitt1999; Hillman et al., Reference Hillman, Zardkoohi and Bierman1999).
In comparison to hiring external lobbyists, RDC might provide firms with more influence in affecting policies and gathering information. Because of the personal connections, RDC generally have the highest levels of personal trust between the firm representatives and the government official (Wang & Wu, Reference Wang and Wu2020; see also Hillman, Reference Hillman2005). The revolving door provides many benefits: access to information (Hillman, Reference Hillman2005), reduced political transaction costs (Kingsley et al., Reference Kingsley, Bergh, Richard and Bonardi2012; Luechinger & Moser, Reference Luechinger and Moser2014), buffer against unfavorable policies (Lester et al., Reference Lester, Hillman, Zardkoohi and Cannella2008), and influence toward favorable policies (Bonardi, Holburn, & Vanden Bergh Reference Bonardi, Holburn and Vanden Bergh2006; Faccio, Masulis, & McConnell, Reference Faccio, Masulis and McConnell2007; Hillman, Reference Hillman2005; Ridge, Hill, & Ingram Reference Ridge, Hill and Ingram2016). Empirical studies of revolving door ties reveal positive financial outcomes after firms hire former government officials or appoint them to their board of directors (Goldman, Rocholl, & So Reference Goldman, Rocholl and So2009, Hillman Reference Hillman2005; Houston & Ferris Reference Houston and Ferris2015). We therefore suggest that over the other forms of CPA, firms that combine RDC with their market activities are likely to reap the benefits of integration.
Hypothesis 2: Firms that invest more resources in revolving door connections and market activities will outperform firms that invest fewer resources in revolving door connections and market activities.
Methods
To explore the financial performance of firms integrating their political and their market activities, we focused on a broad cross-section of S&P 1500 publicly traded firms, which reflect about 90 percent of all market value in the United StatesFootnote 2 , alongside all politically active publicly traded firms, which overlap substantially with the S&P 1500; the dataset reflects both politically and nonpolitically active firms. We obtained data from the Center for Responsive Politics (now, OpenSecrets.org) which collects and summarizes Federal Election Commission (FEC) and other public data on firms’ campaign contributions, lobbying expenditures, and revolving door connections (Hadani & Schuler, Reference Hadani and Schuler2013; McDonnell & Werner, Reference McDonnell and Werner2016; Skaife & Werner, Reference Skaife and Werner2020). We then matched all firms with financial and governance data in Compustat from 1998 to 2018, leading to a final sample of between 18, 847, and 19, 422 firm-year observations, varying across models and analytical approaches.
Variables
Dependent variable: Firm performance . We selected two widely used measures of performance. The first measure is end-of-year market value, a common financial market measure in strategy (Castellaneta et al., Reference Castellaneta, Conti and Kacperczyk2017; Cho & Pucik, Reference Cho and Pucik2005) and CPA research (Hadani & Schuler, Reference Hadani and Schuler2013) to indicate anticipated future performance. The second is return on assets (ROA), another popular measure of firm performance, based on historical accounting data (Arend, Patel, & Park, Reference Arend, Patel and Park2014; Barnett & Salomon, Reference Barnett and Salomon2012; Hillman, Reference Hillman2003). These reflect both forward and backward-looking measures of firm performance.
Independent variable: CPA. Given the breadth of CPA research (Abdurakhmonov, Ridge, & Hill, Reference Abdurakhmonov, Ridge and Hill2021; Aggarwal et al., Reference Aggarwal, Meschke and Wang2012; Cao et al., 2018; Cooper et al., Reference Cooper, Gullen and Ovtchnikov2010; Drutman, Reference Drutman2015; Hersch et al., Reference Hersch, Netter and Pope2008; Hadani & Schuler, Reference Hadani and Schuler2013; Hillman et al, Reference Hillman, Keim and Schuler2004; Lux et al, Reference Lux, Crook and Woehr2011; McDonnell & Werner, Reference McDonnell and Werner2016; Skaife & Werner, Reference Skaife and Werner2020), we focused on three well-used measures: (1) financial CPA (FIN CPA), a firm’s combined expenditures on both lobbying and political action committee contributions or PACs; (2) executive-level revolving door connections (REV DOOR); and (3) the presence of a corporate Washington DC government relations office (DC OFFICE). These measures represent distinct and complementary aspects of CPA, reflecting direct monetary investments, such as PAC contributions and hiring professional lobbying firms, and others focused on information monitoring and exchanges, specifically the use of DC offices and revolving door connections. We obtained data from OpenSecrets.org and the FEC (supplemented by Google searches) on FIN CPA and REV DOOR. We used the publication Washington Representatives for the DC OFFICE measure.
Independent variable: Market strategies: A central weakness of the NISM literature is that it does not clearly define the range of “market” activities. Thus, we draw upon the strategy literature to identify three principal strategies firms use to deal with their competitive environment.
First, we focus on a firm’s investments in research and development, R&D, because of its central role in overall innovation and differentiation (Barney, 1991; Porter, 2008; Callander, et al., Reference Callander, Foarta and Sugaya2022; Porter, Reference Porter1985, Reference Porter1996; Rumelt, Reference Rumelt2012), measured as R&D expenditures (from Compustat). Second, we include the firm’s marketing investments, measured as marketing expenditures (Compustat), an avenue of positioning and differentiation (Porter, 2008). Finally, we consider a firm’s merger and acquisition activity (M&A), a key growth strategy (Haleblian et al., Reference Haleblian, Devers, McNamara, Carpenter and Davison2009; Tuch & O’Sullivan, Reference Tuch and O’Sullivan2007). To create this measure, we examined media reports about any instances where a firm acquired or attempted to acquire another firm or entity. This avoids creating selection bias for those firms completing M&As only to include those that pursued mergers but did not complete them. We supervised a team of three research assistants to create such a measure by searching Lexis-Nexis and Factiva, using firm names alongside relevant search words (e.g., bid, target, acquisition, merge/r, intent to, acquire, buy out, buy, consolidate, regulatory approval, etc.) to identify any M&A attempts. This search yielded 2,120 such events.
We treated missing data in two complementary ways. First, we added a dummy selection variable with 0 indicating missing data and 1 indicating data was present initially, in the originating dataset. Second, and closely following the extent research, we coded missing data as 0, increasing the number of firm-year observations and the comprehensiveness and representation of the sampleFootnote 3 .
Control variables . We controlled for the year and 4-digit SIC code dummies. We added the abovementioned dummy variables (missing values) to account for market activity selection bias. We also controlled for lagged firm performance (generated by the general method of moments (GMM) model by default and varies by model; either market value or ROA)Footnote 4 , firm size (log of number of employees), firm diversification (Herfindahl measure), and institutional equity ownership (percentage measure taken from 13-f filings).
Integration: sequence and timing
NISM research does not provide specific guidance about integration as well as the timing of each activity relative to the others. Shaffer et al. (Reference Shaffer, Quasney and Grimm2000) study shows how nonmarket strategies contributed positively to three measures of performance in the presence of market strategies (and controls) but does not consider the timing of the interactions. Jia and Mayer (Reference Jia and Mayer2016) and Bonardi (Reference Bonardi2004) also refer to integration as a combination of CPA alongside some form of market activity; our concurrent interactions reflect such configuration. Baron’s theoretical development of the concept of integration and his examples clearly refer to firms engaging in R&D activities alongside CPA activities, without explicit focus on the sequence. Holburn and Vanden Bergh (Reference Holburn and Vanden Bergh2014) also rely on interactions to reflect aspects of market and CPA integration; they demonstrate the temporal interactions between CPA and market activities.
Thus, what is consistent with the limited empirical research on integration is that firms combine their CPA alongside market activities to varying degrees. We thus begin with the concurrent period integration on two measures of performance: market value and ROA. For these dependent variables, we examine all potential configurations of CPA vis-à-vis market activities. In total, our primary analyses examine nine interaction terms per each regression model: FIN CPA*R&D, FIN CPA*MKT, FIN CPA*M&A, REV DOOR*R&D, REV DOOR*MKT, REV DOOR*M&A, DC OFFICE*R&D, DC OFFICE*MKT, and DC OFFICE*M&A.
Second, NISM research is mostly silent on the timing of the integration of CPA and market activities; this opens the door to consider the temporal dynamics of the model. Specifically, we examined both sides of the lagging, first looking at firms with lagged market activity (and hence more experience) relative to current CPA experience (the most likely scenario) and, second, looking at firms with lagged CPA (more CPA experience) relative to current market experience (a less likely scenario). We explored the aforementioned nine interaction terms (with one type of strategy lagged and the other current and vice versa) with regard to both market value and ROA, our two measures of performance.
The first set of temporal interaction terms are interactions involving 1-year lagged R&Dt-1, MKTt-1, and M&At-1, with current FIN CPA, REV DOOR, and DC OFFICE. We ran these nine interactions in models with the two measures of firm performance (market value and ROA). Next, we looked at firms with lagged CPA experience relative to market experience and generated a new set of interactions; these involve lagged FIN CPAt-1, REV DOORt-1, and DC OFFICEt-1 with current R&D, MKT, and M&A.
Integration: Industry Deviation . NISM research has not addressed firms’ integration of nonmarket and market activities relative to industry norms. We step into this void and empirically explore whether deviation from industry norms might act as goalposts to guide firms’ decision-making in NISM. In other words, firms may view integration as a mimetic configuration relative to their industry peers (DiMaggio & Powell, Reference DiMaggio and Powell1983; Oliver & Holzinger, Reference Oliver and Holzinger2008) that may improve performance, although industry peers may also be unsure about integration. Thus, we created new variables based on each firm’s CPA and market strategies relative to each 4-digit SIC by subtracting each firm’s CPA and overall market expenditures from the 4-digit SIC industry mean to create new deviation variables, FIN CPA deviation, REV DOOR deviation, DC OFFICE deviation, R&D deviation, MKT deviation, and M&A deviation. We ran our regression models with these new variables and their nine interaction terms with the same two dependent variables.
Analytical approaches
We have unbalanced panel data that include CPA and market strategies’ variables that may be endogenously related to firm market value (Aggarwal et al., Reference Aggarwal, Meschke and Wang2012; Hillman et al, Reference Hillman, Keim and Schuler2004). Furthermore, CPA and market strategies may co-determine each other when considering the same firm across different time periods. To analyze unbalanced panel data with possible omitted variable bias, correlated errors, and endogeneity among the variables, we used the GMM dynamic panel data analytical approach (Arellano & Bond, Reference Arellano and Bond1991; see also Woolridge, Reference Wooldridge2002). GMM uses the first difference of the regression equation to remove subject effects and uses lags of the independent variable as instruments for differenced lags of the dependent variable to generate dynamic panel data instruments (Baum, Schaffer & Stillman, Reference Baum, Schaffer and Stillman2003; Pollock et al., Reference Pollock, Lee, Jin and Lashley2015). It is based on moment functions that rely on observable variables and unknown population parameters and generates estimates like those of a traditional 2SLS instrumental regression approach. Karthikeyan, Jonsson, and Wezel (Reference Karthikeyan, Jonsson and Wezel2016) note that GMM produces estimates that are more statistically efficient than 2SLS, especially in the presence of autocorrelation. Moreover, GMM helps reduce reverse causality and omitted variable bias (Marino et al., Reference Marino, Aversa, Mesquita and Anand2015). In all the models, we defined CPA and the two market variables as endogenous relative to the dependent variables. Given the use of lagged differences, GMM yields a smaller sample size compared to other analytical approaches (see post hoc analyses below). System GMM removes multicollinear variables from regression by defaultFootnote 5 .
Results
Given the abductive nature of this paper, we created several sections for the analyses. We first report the concurrent analyses, followed by the temporal analyses, and then the industry deviation analyses. Table 1 provides summary statistics and correlations among all variables.
Table 1. Summary statistics and bivariate cross-sectional correlations of variables
ROA = return on assets; N = 18,847-19,422; correlations greater than .02 are significant at p > .05 or higher. R&D and marketing investments in thousands.
Before revealing these results, we note that none of the interaction terms of market and CPA measures were significant when using ROA as a measure of performance. As such, we report only the models using market value as the dependent variable. We will discuss the implications of the lack of support for ROA as a dependent variable in the limitations section.
Concurrent analyses results
We proposed two hypotheses, an omnibus hypothesis proposing that firms that invest more resources in different forms of CPA and market strategies will outperform those firms that do not, and a second hypothesis that posits that of three forms of CPA, the integration of RDC with market strategies has the best potential to bring performance benefits.
As seen in Table 2, which examines concurrent activities, the overall pattern of results for market value as the dependent variable is mixed. Of the nine interaction terms in model 3, only three are positive and significant; the rest are either negative and significant or nonsignificant. The positive and significant interactions include FIN CPA and R&D (β = .00, p < .0001), REV DOOR and R&D (β = 149, p < .0001), and REV DOOR and M&A (β = 352700, p < .0001). The three negative and significant interactions include FIN CPA and MKT (β = −.00, p < .01), REV DOOR and MKT (β = −103, p < .0001), and DC OFFICE and R&D (β = −4206, p < .01). The three nonsignificant interaction terms are FIN CPA and MKT, FIN CPA and M&A, DC OFFICE and MKT, DC OFFICE and M&A (p > .10, across the board).
Table 2. GMM regression analysis—concurrent CPA and market strategies—market value
GMM = general method of moments; CPA, corporate political activity; ROA = return on assets; AR = autoregressive. †p < .10, *p < .05, **p < .01, ***p < .001, ****P < .0001. Standard errors in parentheses. Sargan test; H0: overidentifying restrictions are valid (when p > .05); all AR structures set at 2.
Temporal sequence analyses: 1-year lagged market activities with current CPA
Examining Table 4, model 8, we see another mixed picture. Here, most of the interaction terms are significant: five are positive, three are negative, while one is nonsignificant. Specifically, the following interaction terms are positively associated with market value: FIN CPA and R&Dt-1 (β = .001, p < .0001), FIN CPA and M&At-1 (β = .63, p < .05), REV DOOR and R&Dt-1 (β = 139, p < .0001), REV DOOR and M&At-1 (β = 364558, p < .001), and DC OFFICE and MKTt-1 (β = 18361, p < .001). The following interaction terms are negatively associated with market value: FIN CPA and MKT-1 (β = −.001, p < .01), REV DOOR and MKTt-1 (β = −262, p < .0001), and DC OFFICE and R&Dt-1 (β = −9099, p<.0001). Lastly, DC OFFICE and M&At-1 are not significant (β = 1668245, p > .05).
Table 3. GMM regression analysis—concurrent CPA and market strategies industry deviation measures—market value
GMM = general method of moments; CPA, corporate political activity; ROA = return on assets; AR = autoregressive. †p < .10, *p < .05, **p < .01, ***p < .001, ****P< .0001. Standard errors in parentheses. Sargan test; H0: overidentifying restrictions are valid (when p > .05); all AR structures set at 2.
Table 4. GMM regression analysis: Lagged market and current CPA strategies—market value
GMM = general method of moments; CPA, corporate political activity; ROA = return on assets; AR = autoregressive. †p < .10, *p < .05, **p < .01, ***p < .001, ****P < .0001. Standard errors in parentheses.
Temporal sequence analyses: 1-year lagged CPA with current market activities
Table 5, model 10, presents another mixed picture. We find three terms positively associated with market value, two terms negatively associated with market value, and four nonsignificant interactions; most are not positive. Specifically, the following interaction terms are positively associated with market value; FIN CPAt-1 and MKT (β = .00, p < .001), REV DOORt-1 and R&D (β = 187, p < .0001), and REV DOORt-1 and M&A (β = 335951, p < .0001). The following interaction terms are negatively associated with market value; FIN CPAt-1 and R&D (β = −.00, p < .01), and REV DOORt-1 and MKT (β = −328, p<.001). Lastly, the following terms are not significantly associated with market value; FIN CPAt-1 and M&A, DC OFFICEt-1 and R&D, DC OFFICEt-1 and MKT, DC OFFICEt-1 and M&A (all p > .10).
Table 5. GMM regression analysis: lagged CPA and current market strategies—market value
GMM = general method of moments; CPA, corporate political activity; ROA = return on assets; AR = autoregressive. †p < .10, *p < .05, **p < .01, ***p < .001, ****P < .0001. Standard errors in parentheses. Sargan test; H0: overidentifying restrictions are valid (when p > .05); all AR structures set at 2.
Industry deviation results
Regarding the industry deviation concurrent analyses, reported in Table 3, we again have mixed results. Of the nine interaction terms reported in model 6, which looks at market value as the dependent variable, three are positive and significant, three are negatively significant, and three are not significant. Again, most are not positive. Specifically, the positive and significant interactions include the terms of FIN CPADEV and R&DDEV (β = .00, p < .001), REV DOORDEV and R&DDEV (β = 130, p < .0001), and REV DOORDEV and M&ADEV (β = 300569, p < .0001). The negative and significant interactions include FIN CPADEV and MKTDEV (β = −.00 p < .01), REV DOORDEV and MKTDEV (β = −217, p < .0001), and DC OFFICEDEV and MKTDEV (β = −19746, p < .01). The nonsignificant interactions terms are FIN CPADEV and M&ADEV, DC OFFICEDEV and R&DDEV, and DC OFFICEDEV and M&ADEV (all p > .10).
Diagnostic analyses
We also examined the autoregressive (AR) structure and the exogeneity of the instruments generated by the GMM models. First, we ran the Hansen test to explore the most appropriate AR structure (with one or two residual lags) using the Stata abond post-estimation command, which reports the Arellano–Bond test statistic for serial correlation in the first-differenced residuals. In all models, the first-order statistics, AR(=1), were significant, rejecting the null, but the AR(=2) were not (p > 0.3). As such, we used the AR(=2) option in all models (as well as in the post hoc analyses). Second, to verify the validity of the differenced instruments generated by the GMM procedure, we followed Baum and colleagues’ (2003) C statistic test, comparing the main effects model to control models and comparing full effects models to main effects models. The models indicate that the C statistic does not reject the null, consistently (p > 0.4 and above). We provide the results of the diagnostic analyses at the bottom of Tables 2 through 5.
Effect Size Analysis
We report on effect sizes after controlling for all other variables in our regression models. We use the Stata margins DY/DX post-estimation command to assess their effect size. The effect sizes reflect the effect that one standard deviation in each of the two independent variables comprising each interaction term has on one standard deviation of the dependent variable, market value, when the term is significantFootnote 6 . On the one hand, the largest positive effect size is .232 for the interaction of R&D and RDC (and RDCt-1), meaning that a one-unit increase increases the standard deviation of market value by .232, less than a fifthFootnote 7 . On the other hand, the biggest negative effect size is -.135 for the deviation analyses (MKT and RDC), about 10 percent. Overall, for the significant interactions, the effect sizes involving RDC are small.
Summary table
To simplify the interpretation of the complex pattern of findings, we provide an omnibus summary table reporting the effects across all analyses (see Table 6). As the table notes, only RDC and R&D stand out, respectively, as one aspect of CPA and one aspect of market activities that might lead to positive performance outcomes when integrated. This pattern reiterates the predictive power of hypothesis 2. One analysis further stands out—the lagged market strategies one. As can be seen from model 8 (Table 4), five out of the nine interaction terms involving lagged market strategies with the latter CPA are positive and significant, albeit with mostly very small coefficients. Such an aggregate pattern is not evident for the other analyses; lagged market strategies are more likely to facilitate integration benefits, while the analyses for lagged CPA do not yield such positive outcomes (see Table 3).
Table 6. Summary table across models
Text in italics represents the dominant relationship between that cell’s interactions and market value as the dependent variable. Text in bold italics represents the dominant relationship of that variable and the dependent variable across the interaction terms.
Post hoc analysis: additional market strategies
Our original analysis explored three common market strategies linked to achieving competitive advantage and increased performance: R&D, marketing, and M&As. However, firms may take other strategic actions firms take to try to gain potential competitive advantage. Specifically, we added two variables related to innovation: the number of patents submitted to the USPTO’s open data portal (PAT)Footnote 8 , per firm per year and the mass media sentiment of innovation (from the Thomson Reuters Refinitiv Market Psych dataset), per firm per year. We re-ran all the same models as we did for the original analyses including these new variables. We find that for patents nearly all interaction terms with CPA are not significant. We further find that for perceived innovation most interaction terms are negative. The results are available on request.
Post hoc analyses: additional performance measures
We also considered other measures of performance. One measure of a firm’s future performance expectations is Tobin’s Q, the ratio of the market value of a firm to the book value of assets (Singh et al., Reference Singh, Tabassum, Darwish and Batsakis2018; Wernerfelt & Montgomery, Reference Wernerfelt and Montgomery1988; Yermack, Reference Yermack1996). Using GMM, we re-ran all analyses with Tobin’s Q as the dependent variable. We did not find statistical significance between the integration of the various CPA and market strategies and Tobin’s Q.
Post hoc analysis: 2-year lags
To further explore the temporal dynamics of the original analyses, we added a 2-year lag for both the market and CPA variables. We include all three measures of market activities. We used the (first-differenced) GMM analysis, and the results were overall like those of the 1-year lag, with two caveats.
Specifically, when moving from 1 to 2-year lags of CPA with concurrent market activities, the interaction term of MKT and FIN CPA t-2 becomes negative, as does R&D and DC OFFICE t-2, and M&A and DC OFFICE t-2. At the same time, the interaction term of FIN CPA t-2 and R&D becomes positive, as does FIN CPA t-2 and M&A. We report these results in an online appendix.
Post hoc analyses: system GMM
To ensure the robustness of our analyses, we also ran an Arellano-Bover (Reference Arellano and Bover1995) system GMM analyses. The system GMM model goes beyond the Arellano-Bond (Reference Arellano and Bond1991) GMM estimators as it not only corrects for the endogeneity of the lagged dependent variable and other potentially endogenous explanatory variables, but it is also robust to potential measurement error and weak instruments (Kukenova & Monteiro, Reference Kukenova and Monteiro2009). The system GMM estimators also control for arbitrary heteroskedasticity and autocorrelation in the disturbance error (Kukenova & Monteiro, Reference Kukenova and Monteiro2009). The results of these analyses are like the original ones.
Post hoc analyses: integration mimicry
An intriguing dynamic that could explain the overall pattern of results is the concept of integration mimicryFootnote 9 . Some firms might observe and replicate the integration strategies of others that have successfully combined CPA and market strategies. This imitation could neutralize any competitive advantage initially gained from integration, thereby obscuring the relationship between integration and performance.
To investigate this possibility, we selected qualitative comparative analysis (QCA), a post-observation clustering method (Ragin, Reference Ragin1987) that uses Boolean logic to examine the relationships between various predictors and outcomes at the level of individual observations (in our case, firms over time). Unlike traditional regression analyses, QCA is effective in identifying significant patterns among a large set of variables and can handle multiple predictors simultaneously. This allows us to explore how different combinations of CPA and market strategies are associated with various measures of firm performance.
We focused on firms with high performance as measured by market value. The QCA results indicate that only 2 percent of firms across all years and industries show both high market value and high levels of integration in CPA and market strategies (i.e., high levels of all CPA and all market strategies, concurrently). Additionally, 4 percent of firms have high market value while engaging in some CPA activities and some market activities. In total, only about 6 percent of firms seem to benefit from integration. Over time, we observed little fluctuation in the percentage of firms able to achieve high performance through integration, supporting our overall statistical findings that integration benefits are rare.
To further explore mimicry, we compared the frequency of firms with similar configurations of CPA and market strategies from 1 year to the next. We found that the number of firms adopting these configurations without achieving higher market value did not change significantly. This analysis, which encompasses all industries, indicates that the frequency of firms mimicking successful integrators does not increase over time.
We also conducted QCA analyses at the industry level (4-digit SIC codes) to address whether firms can replicate industry peers’ success in integrating CPA and market activities. We examined firms that integrated successfully and experienced higher performance, as well as those with similar CPA and market activities but without higher performance, over subsequent years. By analyzing year-to-year fluctuations at the 4-digit SIC level, we found that the frequency of firms exhibiting elevated levels of CPA, market activities, and performance remains stable. Concurrent analyses corroborate this stability.
Discussion
In this empirical and abductive study, we tested the NISM view that suggests that firms combining political and market activities will outperform those that do not. We used the most popular and established measures of CPA aspects (Bonardi et al., Reference Bonardi, Holburn and Vanden Bergh2006; Hadani & Schuler, Reference Hadani and Schuler2013; Hillman & Hitt, Reference Hillman and Hitt1999; Hojnacki et al., Reference Hojnacki, Kimball, Baumgartner, Berry and Leech2012) as well as focused on the most popular and relevant market activities as noted by CPA and strategy research alike (Bonardi & Keim, Reference Bonardi and Keim2005; Hadani et al., Reference Hadani, Bonardi and Dahan2017; Hillman, Reference Hillman2005; Lux et al., Reference Lux, Crook and Woehr2011). Our extensive 21-year longitudinal analysis, encompassing publicly traded politically active and nonpolitically active firms across nearly 20,000 firm-year observations, while controlling for variables like year, industry, firm size, performance, and institutional ownership, and employing rigorous analytical approaches, did not clearly indicate that integration increased firms’ financial performance. Despite exploring various combinations of market and political variables, time lags, and deviation from industry norms, and employing several estimation techniques, we found insufficient evidence that integration leads to positive financial performance as measured by market value. Furthermore, we did not find that integration led to positive financial performance when considering other dependent variables such as ROA or Tobin’s Q. Overall, our results do not yield enough combinations of integration to support Hypothesis 1. However, we found that revolving door connections—one of the political activities we studied—offered firms limited benefits in terms of market value, lending minor support to Hypothesis 2.
While existing NISM research is optimistic about the benefits of integration, it often fails to differentiate between various CPA approaches or types of market activities. Our study adds significant value by adopting a broad perspective, incorporating three popular aspects of CPA and market activities (R&D, marketing, and M&A). Using these measures, we applied a configurational approach to measure integration, represented by over 100 different interaction terms, allowing for a more precise analytical and theoretical focus. Our findings consistently show that, in most cases, firms combining their market and nonmarket activities do not perform better than those that do not integrate. Across various models, it is difficult to argue for the empirical validity of the NISM approach. We simply do not find enough evidence that integration leads to improved market value, as proposed in Hypothesis 1. The exception lies with revolving door connections: here, integration with certain market strategies resulted in small increases in market value, though the effect sizes were minimal. Thus, we provide qualified support for Hypothesis 2, indicating that of the political strategies examined, revolving door connections have the most potential for integration benefits.
Our temporal analyses reveal additional nuances. Specifically, two lagged market activities—R&D and M&A—stand out (see Table 4, Model 8), where the interaction terms with CPA tended to be positive. However, this dynamic did not hold for lagged FIN CPA configured with ex-post market strategies; when examining lagged FIN CPA interactions with R&D and M&A (Table 5, Model 10), the interactions were negative. Experience with R&D and M&A activities also shows potential for integration.
In the following sections, we delve deeper into these complex findings. We theorize four dynamics that might explain why firms often do not realize the benefits of integration, and, just as importantly, when they might. We draw upon Karl Weick’s insights on theory building. In his 1989 article, Weick noted: “If theorizing is viewed as disciplined imagination, where the ‘discipline’ in theorizing comes from consistent application of selection criteria to trial-and-error thinking, and the ‘imagination’ in theorizing comes from deliberate diversity introduced into the problem statements, thought trials, and selection criteria that comprise that thinking” (p. 516). We use our extensive analysis of the market value results from different configurations of CPA and market activities as our “trial-and-error” thinking. We believe these results could lead to further theorizing around strategic control limitations, policy opportunity windows, visibility through market activities, and challenges to integration mimicry.
Strategic control limitations
The NISM model seems to overestimate firms’ abilities to collect, understand, and act on information and cues from the distinct political and market environments. Baron’s original writings did not identify the internal structures necessary for effective integration. NISM assumes that firms (and their executives) are adept at managing uncertainty, information asymmetry, and control limitations arising from each environment. However, our findings show that integration is not positively associated with performance, challenging these assumptions.
We suspect that most companies lack the structures, processes, and personnel needed to effectively understand and translate the political environment into core business strategies. One challenge in operating within the political arena is the high level of uncertainty surrounding policy outcomes, including how firms’ (and other actors’) political activities might influence these outcomes (Baumgartner et al., Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009; Drutman, Reference Drutman2015; Hadani et al., Reference Hadani, Bonardi and Dahan2017; Hart, Reference Hart2004; Smith, Reference Smith2000). As North (Reference North1990) noted: “The political market has been, and continues to be, one in which the actors have an imperfect understanding of the issues affecting them” (p. 357).
Additionally, firms often lack control over certain aspects of CPA. For example, once firms donate money to politicians’ campaigns, they cannot dictate what those politicians will do (Hadani et al., Reference Hadani, Bonardi and Dahan2017; Hall & Deardorff, Reference Hall and Deardorff2006). Drutman (Reference Drutman2015) emphasized this point by saying, “Politics is not a vending machine where outside forces can put in their dollar, press a few buttons, and reach below for a neatly shrink-wrapped policy outcome” (p. 5). Even hiring outside lobbyists does not guarantee policy success (Aggarwal et al., Reference Aggarwal, Meschke and Wang2012; Cao et al., 2018; Coates, Reference Coates2012; Hadani & Schuler, Reference Hadani and Schuler2013; Skaife & Werner, Reference Skaife and Werner2020). Indeed, Hadani et al. (Reference Hadani, Bonardi and Dahan2017) in their meta-analyses note that firms may think that a specific policy outcome will benefit them but that there often is a disconnect between perceived public policy outcomes and firm-level benefits; they note that due to political outcome uncertainty (their term) even if firms are somehow able to obtain a specific policy outcomes that do not necessarily translate into firm-level benefits. Hence, aligning market strategies with hard-to-control political actors is more complex than the NISM model assumes.
One area where we did find greater efficacy in bundling market and political activities for performance improvement was through revolving door connections—the personal relationships formed between firms and government bodies via the movement of personnel (Etzion & Davis, Reference Etzion and Davis2008). We proposed that revolving door connections, due to their role as an information exchange conduit and the firm’s (at least partial) control over them, would mitigate the integration challenges mentioned earlier. Indeed, our results provide limited support for this view. Future research should focus on the qualities of specific personnel (e.g., their backgrounds, competencies, connections, etc.), as well as their positions within the organization, as key factors in the effectiveness of revolving door connections to support an integrated strategy.
Policy opportunity windows
An important finding is how the timing of market activities versus CPA affects market value. Certain lagged market strategies, when paired with specific CPA, showed minor performance benefits. We believe this reflects the differing time horizons that firms face when engaging in productive activities across the market and political environments.
The political environment is unique in that firms’ political activities must align with policy opportunity windows—times when the public policy environment is open for firms and other interest groups to have input on issues that matter to them (de Figueiredo, Reference de Figueiredo2004; Kingdon, Reference Kingdon2010; Smith, Reference Smith2000). Many factors influence the openness of the public policy environment, including elections and noteworthy events (Baumgartner et al., Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009; Georgiou, Reference Georgiou2004; Hojnacki et al., Reference Hojnacki, Kimball, Baumgartner, Berry and Leech2012; Kim, Urpelainen, & Yang, Reference Kim, Urpelainen and Yang2021; Weschle, Reference Weschle2021). De Figueiredo (Reference de Figueiredo2004) notes that lobbying activity intensifies around specific policy opportunity windows and diminishes when those windows close. Public attention to issues also plays a role in opening these windows. When issues become more salient to policymakers, they receive more attention, meaning firms must act quickly as opportunities arise (Smith, Reference Smith2000). Bonardi and Keim (Reference Bonardi and Keim2005) emphasize the importance of public attention in influencing issue salience explaining that CPA is effective only at certain times. Conversely, markets have their own temporal dynamics, shaped by macroeconomics and industry, shareholder, and stakeholder pressures (Hoskisson et al., 2001; Nadkarni & Barr, Reference Nadkarni and Barr2008). Often, the timing and pace of these two environments do not align.
The episodic nature of policy opportunity windows means that firms must be prepared to act swiftly when these windows open. However, organizing such opportunities is challenging; it requires a deep understanding of the political environment, typically gained through personal relationships and experience, as well as the internal mechanisms and leadership to respond quickly. Even the use of professional intermediaries like lobbyists cannot fully eliminate the uncertainty of the policy environment (Drutman, Reference Drutman2015; Hadani et al., Reference Hadani, Bonardi and Dahan2017; Hart, Reference Hart2004). This phenomenon may help explain our overall findings, as political investments are often made well in advance of policy opportunity windows, making it difficult to link political investments with market strategies if both are examined within the same period.
Our lagged results also support this idea of policy opportunity windows. Specifically, when comparing the overall pattern of the nine interaction terms in the concurrent analyses (Table 2) with the lagged market strategies analyses (Table 4), we observe that more interaction terms are positive and significant in the lagged analysis (5) than in the concurrent analyses (3). This is most apparent for R&D and merger and acquisition attempts; for example, the interaction terms with revolving door connections became more positive as we moved from concurrent (R&D) to lagged measures (R&Dt-1 and R&Dt-2). Additionally, the long-term nature of R&D in many industries makes it more likely that R&D investments occur during political opportunity windows than investments in marketing or mergers and acquisitions. Overall, we propose that when firms’ market activities allow sufficient time for CPA to capitalize on open political opportunity windows, integration is more likely to lead to positive financial outcomes.
In other words, our analysis highlights the importance of not only prioritizing market activities but also adopting a “wait-and-see” approach to assess, in advance, when political opportunity windows may arise. This would enable leveraging existing R&D efforts in the political sphere. It emphasizes both the potential for positive integration and the contingent nature of aligning R&D with certain aspects of CPA, which may not always be relevant. Additionally, it stresses the need to clearly identify how firm-level market activities could benefit from CPA. This level of strategic nuance is critical for effectively capitalizing on opportunity windows, yet current NISM-related research does not adequately address it.
Visibility via market activities
Research has long recognized that visibility can undermine the effectiveness of firms’ CPA (Bonardi & Keim, Reference Bonardi and Keim2005). Various scholars attribute this to the social stigma associated with CPA, which leads firms to keep such activities low-profile or even covert (Katic & Hillman, Reference Katic and Hillman2022; McDonnell & Werner, Reference McDonnell and Werner2016; Shanor, McDonnell, & Werner, Reference Shanor, McDonnell and Werner2021). Other studies (Hadani, Aksu & Combes, 2020) have shown that visibility diminishes a firm’s ability to leverage its CPA. They observe, “Our findings not only emphasize the need to explore the forces to which firms and their political counterparts are susceptible but also more generally reflect a need to consider how political marketplace exchanges are impacted by the broader information environment in which firms operate” (Hadani et al., 2020, p. 22).
Ramanna (Reference Ramanna2015) introduced the concept of “thin political markets” to describe situations where political activities go unnoticed and are more likely to yield positive outcomes for the interest group. Similarly, Culpepper (Reference Culpepper2021, p. 134) writes that “business continues to punch above its weight in all democracies, and the recesses in which quiet politics has long operated continue to facilitate that undue degree of influence.” Conversely, when public attention rises, corporate political activities are less likely to produce favorable results. Building on this logic, we argue that market activities that increase a firm’s visibility may negatively impact the effectiveness of its CPA, which typically benefits from secrecy and opacity (Smith, Reference Smith2000); Smith notes that as the circle of participants in the firm–government interaction grows, the ability of firms to leverage their political activities declines. The same activities that enhance a company’s visibility and attractiveness to consumers may make it less appealing to policymakers, especially elected officials. Among the market activities, marketing is the most explicitly visible by design. Our findings indicate that, in most cases, firms that spent considerable amounts on marketing and CPA did not see improvements in market value; in fact, many experienced declines. Marketing, by its nature, makes a firm more visible to the public, potentially increasing scrutiny on political transactions, thereby diminishing any benefits of integration (Hadani et al., Reference Hadani, Aksu and Coombes2021).
Visibility might also explain why, compared to other market strategies, R&D was beneficial when integrated with CPA. As shown in the summary table, most interactions between R&D and CPA positively affect market value, even in the deviation analyses. This may be because firms hide many elements of R&D as trade secrets and, even when disclosed through patents, remain less visible to outsiders.
NISM mimicry
Another explanation for the lack of a statistically significant positive relationship between integration and financial performance is that any “abnormal” returns from integration may dissipate as competitors adopt similar strategies. As competitors imitate what they perceive to be successful nonmarket and market strategy (NISM) integrations, the abnormal returns may revert toward zero, thereby obscuring the link between integration and performance.
We addressed this concern directly in our post hoc analysis using QCA and found no evidence supporting this notion, either across or within industries. The lack of support might not be surprising. As noted earlier in this paper, there is significant causal ambiguity affecting firms’ ability to integrate CPA and market activities (e.g., Funk & Hirschman, Reference Funk and Hirschman2017; Holburn & Vanden Bergh, Reference Holburn and Vanden Bergh2014). Additionally, there is uncertainty about whether CPA can generate beneficial policy outcomes (Baumgartner, Berry, Hojnacki, Kimball, & Leech, Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009; Drutman, Reference Drutman2015; Hadani, Bonardi & Dahan, Reference Hadani, Bonardi and Dahan2017; Keim, 2001) or improve firm performance (Hadani et al., Reference Hadani, Bonardi and Dahan2017; Katic & Hillman, Reference Katic and Hillman2022; Kersch, Reference Kersh2002). Research suggests that when uncertainty and ambiguity are high, as they are in the challenge of integrating the political and market environments, firms will struggle to learn from their peers (Lieberman & Asaba, Reference Lieberman and Asaba2006). This is because ambiguity not only makes causal learning more difficult (Levitt et al., Reference Levitt and March James1988; Strang & Still, Reference Strang and Still2006) but also complicates the benchmarking of other firms’ behavior under such conditions. Thus, we cannot conclude that the lack of statistically significant results across most of our models stems from the imitation of successful integration strategies.
Taken together our analyses elaborate on the potential dynamics explaining why for the most part NISM fails to result in performance. Extant CPA research should take heed and reconsider the past beneficial view of such integration dynamics and the executive challenges of attempting such integration. A long time ago Bhagwati (Reference Bhagwati1982) discussed the notion of directly unproductive rent seeking reflected in political activities and the economic distortions they can generate, and Lux et al. (Reference Lux, Crook and Woehr2011) raised the notion of not only tradeoffs but implied that opportunity costs exist when firms focus on political activities vis-à-vis their market activities. While some research is optimistic about the potential and ease of integration (i.e., Jia & Mayer, Reference Jia and Mayer2016) such a view may be too simplistic as it does not take into account such tradeoffs and opportunity costs associated with investments in CPA. Though our focus is slightly different than that of Bhagwati and Lux and colleagues’ work, our current analyses strongly echo the potential for a trade-off and an opportunity cost associated with different patterns of CPA and market activities’ integration. Perhaps future CPA research should focus on the primacy of market activity as not only taking precedence over CPA but also highlight the risks of relying too much on the unfulfilled promises of CPA and its questionable ability to create strategic synergies; a focus on the market first is key.
In addition to the above, it is noteworthy that Baron’s work also implies that nonmarket strategies can be effective even without being combined with market strategies. His research originated in the late 1990s, while later research strongly disconfirmed this view as a recent meta-analysis shows (i.e., Hadani et al., Reference Hadani, Bonardi and Dahan2017). We echo this view as our analysis of CPA shows they are not positively associated with performance (market value), and some are negatively associated with market value. Baron’s belief in the ability of CPA in and of itself to benefit firms’ bottom line may have contributed to the belief in the strength of the NISM.
Limitations
Like most research, our study has limitations. First, we examined integration as patterns of investments or configurations without delving into the internal decision-making processes of organizations. We assume that firms that invest heavily in both market and nonmarket areas value each and their integration. However, our research design and data do not allow us to assess this assumption directly. We do not know whether firms consciously attempt to integrate or instead segregate market decisions from political decisions. We also did not study other paths that firms use to pursue both CPA, such as through trade association political activities, and market strategies, such as through joint ventures, which might obscure the integrated strategy relationship. Researchers might use qualitative methods to collect such data and explore the assumptions underlying organizational decision-making about integration.
Second, while we focused on the most common and popular forms of CPA and market investments, we may have overlooked other relevant forms. Future studies could explore additional CPA measures, such as covert political activities. Although we cast widely to capture key market strategy variables, we recognize that we may have missed other important ones.
Third, we were unable to find significant interaction effects between market and CPA variables in models using ROA as the dependent variable. It is possible that it is better to capture the complexity of evaluating CPA and its integration with market investments through market-based measures, such as stock valuation, rather than accounting metrics. Indeed, others find positive market reactions to political activities (e.g., Green, et al. Reference Green, Schuler, Zavyalova, Swartz, Nault and Kazi2023). Investors have considerable flexibility to use big data and sophisticated analysis to estimate the value of integration, whereas accounting is constrained by professional standards and legal requirements. Given the challenges in empirically validating theories (King & Berchicci, Reference King and Berchicci2022), researchers might examine the effects of integration on other performance measures.
Fourth, we accounted for industry effects by using both dummy variables and industry deviation measures. However, it is possible that certain industries are better at achieving integration than others. Certain industries might experience clearer connections between their market investments and political activities and have more predictable life cycles of relevant political issues. Future research should explore this possibility.
Finally, although our overall findings do not support a strong positive effect on integration, we found that certain combinations of market and political variables slightly improved firms’ market value. This suggests that scholars should investigate whether specific conditions and contexts lead to more predictable outcomes from integrating political and market activities.
Conclusion
The idea that firms can integrate CPA and market activities and achieve superior performance outcomes has long held a strong and logical appeal in the management field (Baron, Reference Baron1995; Bonardi & Keim, Reference Bonardi and Keim2005; Hillman et al., Reference Hillman, Keim and Schuler2004; Jia & Mayer, Reference Jia and Mayer2016; Lenway et al., Reference Lenway, Schuler, Marens, Werner and Green2022). Yet, there has not been extensive empirical support for its thesis. Our study set out to investigate the veracity of NISM. Studying a large data set of publicly traded companies over 21 years, we discover that firms integrating CPA and market strategies are not more likely to experience superior financial performance than firms not doing this integration. With a few exceptions, such as revolving door connections and concurrent and lagged R&D investments, the integration effects of other combinations were either negative or statistically insignificant. While our study has limitations, the exhaustive analyses cannot validate the thesis that integrating market and nonmarket activities brings financial rewards to companies. Although we do not address all aspects of the NISM puzzle, this study shows that in this realm, the Emperor has no clothes.
Open access
