Revisiting the Effect of Internationalization on Firm Governance: A Replication and Extension Study.
| Verfasser | Chandler, Jeffrey A. |
| Posten | RESEARCH ARTICLE |
1 Introduction
Strategic management and international business researchers have long examined how internationalization shapes firm's governance structures (e.g., Athanassiou & Nigh, 1999, 2000; Sanders & Carpenter, 1998). According to these studies, firms cope with the heightened uncertainty and complexity of expanding into international markets by implementing better monitoring and information-processing governance structures (Sanders & Carpenter, 1998; Singla et al., 2014). The general assumption throughout the literature is that firms with greater levels of internationalization structure their governance in a way to improve the monitoring of top executives and also increase the specialized knowledge of conducting business internationally. Sanders and Carpenter's (1998) study in the Academy of Management Journal was one of the first to identify this phenomenon by examining the effects of internationalization on several important aspects of firm governance: the CEO, top management team, and the board of directors. The authors specifically found empirical evidence that firms cope with the information-processing demands and agency issues arising from internationalization using five unique approaches: (1) by granting CEOs' higher and more longer-term pay, (2) by constructing larger top management teams, (3) by constructing a larger board of directors, (4) by hiring more independent directors on their boards, and (5) by separating the CEO and chairman of the board positions.
While these unique and important insights have helped shape our understanding of firm governance structures for decades, we suggest that a reexamination of the relationships and results of their study may offer important new insights to strategy and international business research for several reasons. First, Sanders and Carpenter's (1998) study provides arguably the most foundational piece of our understanding of how internationalization shapes a firm's governance structures and research continues to build their conceptual bases on the general premise of their study (e.g., Athanassiou & Nigh, 1999, 2000; Singla et al., 2014). Accordingly, their research is the most cited study addressing the internationalization-governance relationship and has been cited more than 1420 times according to Google Scholar. Replication of their results is needed as it would provide further confidence in their general premise that firms cope with internationalization through their governance decisions. Second, as noted by recent replication studies (Chadwick et al., 2016; Ghosh et al., 2016), there is a need to reexamine previously supported relationships as data availability increases and analytical techniques advance. Sanders and Carpenter's (1998) study has certain limitations in terms of sample size, analytical methods, control variables, and generalizability. For instance, while their study offers important evidence from a cross-sectional design using a sample of 258 firms in only 1 year (1992), it is hard to discern causality in the relationship as there is no time lag to assess a temporal causal inference for their prediction that a firm's degree of internationalization influences its governance structures. This is especially important as sources of endogeneity (i.e., firms selecting certain types of executives or executives selecting into certain types of firms) is central to the hypothesized relationships in their study, which could raise questions of reverse causality.
Third, replication of these results is also needed as some of the relationships found in Sanders and Carpenter's (1998) study may have changed over time. In particular, as CEO compensation has spiked over 514% since the time of their study (Bachelder et al., 2018), research has found that compensation packages are now structured differently with an increasing number of socio-cognitive factors having stronger effects than traditional predictors of CEO pay (Hill et al., 2015; Wade et al., 2006). In addition, research has also shown that top management teams and boards have recently 'ballooned' to over double in size from the time of Sanders and Carpenter's (1998) study (Neatby, 2016) and that boards have been forced to add more independent members as a result of the number of recent crisis-induced regulatory demands placed on publicly traded organizations (Menz, 2012; Neatby, 2016). Accordingly, these developments suggest that the relationships found in Sanders and Carpenter's (1998) study may have changed over time and that firms may not still cope with international uncertainty through CEO compensation or unique construction of the top management team or board of directors, as other factors may be driving these decisions.
Lastly, replication of these results is also needed since the executives that comprise the governance structures of firms have dramatically changed since the context of their study (1992). Media articles and academic research has noted society has increasingly placed expectations that firms become more diverse in their governance structures with recent demands to have greater diversity among boards (Cutter, 2020; Sardon, 2021). As a result, firms have hired more international executives at an exponential rate over the past few decades (Gillenwater, 2020) given the expectations that international executives and board members help alleviate agency issues and improve the specialized knowledge of conducting business internationally (Bergh et al., 2016; Dalton et al., 1998; Herrmann & Datta, 2002). As such, it is important to not only examine if firms still continue to cope with internationalization through the same monitoring structures that were highlighted by Sanders and Carpenter (1998), but also to extend their framework by examining whether firms cope with internationalization by constructing their governance structures with more international executives as well.
To replicate Sanders and Carpenter's (1998) study, we undertake a quasi-replication approach (Bettis et al., 2016) that involves replication with extension (Hubbard et al., 1998; Singh et al., 2003). In particular, we use similar measures and analytical techniques as seen in Sanders and Carpenter (1998) but expand the time frame and use a substantially larger sample (1999 - 2019) with more inclusive control variables in our analysis (Bettis et al., 2016; Chadwick et al, 2016). The purpose of this replication is not to reproduce the original results reported by Sanders and Carpenter (1998) but instead see if their supported findings also hold given the recent changes to firm governance. As a result, we do not re-analyze the original study's data set. Instead, our primary objective is to employ the same specifications and tests to examine the generalizability of the original study to a different context. Importantly, non-findings in replication would not cast doubt on the original study's findings--but instead point to an interesting development that firms have adapted the ways in which they cope with internationalization through their governance structures compared to 30 years ago.
Testing the relationships with Standard & Poor's (S&P) 1500 firms between the years 1999 - 2019, our results report mixed support for the findings highlighted by Sanders and Carpenter. Specifically, our data replicates their findings that firms cope with internationalization by separating the CEO and chairman of the board positions and hiring a higher proportion of independent directors to serve on the board. Also, counter to Sanders and Carpenter's (1998) findings, we find that firms have recently been constructing smaller boards to cope with internationalization. However, our results did not find support that firms cope with internationalization through the CEO's compensation (i.e., higher and longer-term pay) or the size of the top management team. Given that our results coincide with Sanders and Carpenter (1998) central premise that firms cope with internationalization through the CEO position and elements of the board of directors (i.e., lack of CEO duality; board size; board independence), we ran a post-hoc test to examine whether or not internationalization influences firms to hire more international executives in these positions. Specifically, we found strong support that firms cope with the complexity of internationalization by (1) hiring CEOs with international experience, and (2) hiring a more internationally diverse board of directors. Lastly, we also investigate the boundary conditions of the non-findings between internationalization and CEO compensation by examining if firms cope with internationalization through CEO compensation when operating in more corrupt countries. Overall, our replication provides strong support for Sanders and Carpenter's (1998) general premise that firms cope with internationalization through its governance structures but also highlights important non-findings and new findings that reflect how the internationalization-governance relationship has changed over the past 30 years.
2 Replication Methods
Since our primary purpose is to replicate Sanders and Carpenter's (1998) study to determine the generalizability of their results, we did not re-analyze the exact data used in the original study. Instead, we followed previous replication approaches and used a two-step approach for replication (Bettis et al., 2016; Ghosh et al., 2016). First, using an extended time period and different population, we directly tested every hypothesis in their study using the same controls and estimation procedures.' Table 1 presents an overview of our results of this replication. Tables 2, 3, 4, 5, 6, 7 report the results of these tests.
Second, following previous replication studies focused on testing the generalizability of results, we then test their findings by using their same measures while also improving the rigor of the analysis through additional control variables and more advanced analytical...
Um weiterzulesen
FORDERN SIE IHR PROBEABO ANCOPYRIGHT GALE, Cengage Learning. All rights reserved.
