Locally-rooted directors

Kind, Axel; Volonté, Christophe Article — Published Version Locally-rooted directors Review of Quantitative Finance and Accounting Provided in Cooperation with: Springer Nature Suggested Citation: Kind, Axel; Volonté, Christophe (2024) : Locally-rooted directors, Review of Quantitative Finance and Accounting, ISSN 1573-7179, Springer US, New York, NY, Vol. 63, Iss. 2, pp. 633-678, https://doi.org/10.1007/s11156-024-01266-4 This Version is available at: https://hdl.handle.net/10419/315611 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) Review of Quantitative Finance and Accounting (2024) 63:633–678 https://doi.org/10.1007/s11156-024-01266-4 1 3 ORIGINAL RESEARCH Locally‑rooted directors AxelKind1· ChristopheVolonté2 Accepted: 8 March 2024 / Published online: 14 April 2024 © The Author(s) 2024 Abstract We study the influence of locally-rooted directors (LRDs)—board members with personal ties to a company’s geographic location—on firm performance. On the one hand, LRDs may provide valuable local know-how and access to local networks. On the other hand, as their appointments may go back to social ties with insiders (e.g., corporate directors, top executives, or large shareholders), LRDs may be used to extract rents and lack relevant experience, business skills, and independence. Using the directors’ alma mater as a proxy for local roots, LRDs turn out to be heavily overrepresented, making up 30% of all directors in our sample. We show that LRDs are negatively related to Tobin’s Q. However, this finding does not apply to domestically-oriented companies, i.e., firms without material foreign sales, and firms in regulated industries. Thus, while the results indicate that LRDs harm firm performance on average, their presence may be optimal in some cases. Keywords Corporate governance· Board of directors· Social ties· Firm value JEL Classification G30· G34 1 Introduction Bad corporate performance is often ascribed to weaknesses in corporate governance in general and poor board composition in particular (see, e.g., Shleifer and Vishny 1997; Daily etal. 2003). The lack of board directors’ independence and business skills may lead to weak monitoring, poor managerial advice, and suboptimal strategic decisions (see, e.g., Adams etal. 2010, and Johnson etal. 2013, for two surveys on the importance of the board of directors). In this paper, we study the directors’ local roots as an additional dimension in the composition of corporate boards. We consider directors to be locally rooted if they possess personal ties—gained via relevant life experience—to the region where a firm is headquartered. * Christophe Volonté christophe.v[email protected] Axel Kind ax[email protected] 1 University ofKonstanz, Universitätsstrasse 10, 78457Konstanz, Germany 2 University ofBasel, Peter Merian-Weg 6, 4002Basel, Switzerland 634 A.Kind, C.Volonté 1 3 Local roots may have two opposing effects on firm performance. On the one hand, according to Resource Dependence Theory (Pfeffer and Salancik 1978), locally-rooted directors (LRDs) may provide access to important local know-how and experience, as well as valuable links to the company’s external environment, such as municipal authorities, suppliers, financing institutions, and other local stakeholders, making them particularly effective and valuable board members. On the other hand, according to Agency Theory (Jensen and Meckling 1976); Fama and Jensen 1983), LRDs may be appointed because of their personal relationships with corporate insiders, such as the CEO, board members, or controlling shareholders (see e.g., Cheung etal. 2013). These social ties may prevent them from being truly independent and acting as effective monitors. Given these countervailing hypotheses, the relevance and the actual influence of LRDs on firm performance is a matter of empirical research. The paper contributes in several ways to the growing literature on (optimal) board characteristics. First, we add to this literature by proposing directors’ local roots as an additional distinctive dimension in the array of features that determine the contribution of corporate boards and their members to firm outputs. For instance, recent research has highlighted the influence of board diversity (e.g., Chen etal. 2023a; Xie etal. 2024), directors’ co-option (e.g., Chen et al. 2023b), board overconfidence (Twardawski and Kind 2023), and directors’ military experience (e.g., Nawaz and Nawaz 2024) on firm outputs. The effects of local roots on firm performance relate both to Resource Dependence Theory (positive effects due to local know-how and access to powerful local networks) and Agency Theory (negative effects due to lack of independence, commitment to local communities, and lack of monitoring skills), which represent the two most prominent and successful theories for explaining the performance of corporate boards (see, e.g., Johnson etal. 2013). Importantly, LRDs (and the reasons for choosing them) significantly differ both from (nonlocally-rooted) domestic directors and from foreign directors. In particular, the latter are known to be selected for their country-specific know-how by firms with substantial foreign operations, an international shareholder base, and cross-border acquisition intentions (Masulis etal. 2012; Miletkov etal. 2017; Xiang etal. 2023). In contrast to both domestic directors and foreign directors, LRDs may provide specific local know-how and better access to information and resources in the local community. However, due to their personal relations and their commitment to local communities, LRDs may give more attention to the interests of local stakeholders than to those of shareholders (e.g., in decisions on the relocation of production sites that involve lay-offs of the local workforce). This conflict of interest does not exist in that form with the other two types of directors. Second, we provide a simple way of measuring directors’ local roots by focusing on the match between the headquarters’ location and a director’s alma mater. The use of educational institutions as a proxy of cultural proximity is inspired by the work of Cohen etal. (2008), Nguyen (2012), Fracassi and Tate (2012), Ishii and Xuan (2014), and Schmidt (2015) who use the common educational institution as a proxy of social ties among individuals (individual-individual relations). In this paper, we acknowledge the importance of the alma mater in the personal development of individuals but use it to capture the linkages of a director to the firm’s headquarters region (individual-headquarters’ location relations). Third, we contribute to a strand of research that emphasizes that boards are endogenously formed institutions designed to deal with firm-specific challenges (Hermalin and Weisbach 2003; Pathan and Skully 2010). Similar to studies that question the idea that “one size fits all” (see, e.g., Coles etal. 2008; Lehn etal. 2009) and argue that certain board characteristics are value-increasing for some firms but not for others, we investigate whether the effect of locally-rooted directors on firm valuation varies in the cross-section, 635 Locally-rooted directors 1 3 depending on firm characteristics related to the theoretically-grounded effects of local roots. To investigate the phenomenon of locally-rooted directors (LRDs) and their influence on firm valuation, we carry out our research in Switzerland—a country that offers several unique and favorable features for the purposes of our study. First, Switzerland is characterized by a particularly pronounced cultural diversity and strong local peculiarities (often referred to as “Kantönligeist”, i.e., “cantonal spirit”). Second, due to its federal structure, decisions are often made on low hierarchical levels, which makes local roots (and thus familiarity with the local environment) a valuable asset. Third, its comparatively small geographic extension lets travel time likely play a minor role in the choice of directors: the distance between St. Gallen on the eastern border and Geneva on the western border amounts to only 360km, or 224 miles, less than four hours by either car or train. Finally, the distribution of companies’ headquarters across its main regions is remarkably even (see Sect.3.1: Sample). The cultural heterogeneity of Switzerland can be traced back to its 26 federal states (cantons), its four official languages,1 the multitude of local dialects,2 and the religious split in Catholicism and Protestantism.3 Hence, especially in Switzerland local roots of corporate directors are distinctive and well measurable while travel time and geographic distances may unlikely restrict the pool of potential corporate directors and play a major role in choosing them. Our results indicate that LRDs are highly over represented in corporate boards, making up almost 30% of board members. For example, in Hügli Holding, an international food company based in Steinach, 15km from St. Gallen, five out of seven directors graduated from the University of St. Gallen: one with a degree in banking, one with a degree in economics, one with a Ph.D. degree in strategic management, and two with a law degree. Second, and most importantly, the fraction of LRDs is negatively related to firm performance as measured by Tobin’s Q. The result is particularly strong for export-oriented firms, suggesting that LRDs generate net costs in firms where the local business is of minor importance. On the contrary, there is no significant relationship between LRDs and firm performance for companies without relevant foreign sales and for companies in regulated industries, suggesting that boards with an overrepresentation of LRDs may match the needs of those firms. The results hold even after accounting for a large set of common controls and using the University locations as instruments for the percentage of LRDs. The remainder of the paper is structured as follows. Section2 provides a review of the related literature and develops the research hypotheses to be tested. Section3 describes the data and presents the results. Section4 concludes with a summary. 1 The four official languages in Switzerland are (in decreasing order of dispersion): German, French, Italian, and Romansh. 2 There are about 1800 German local dialects in Switzerland (see Lameli etal. 2020). 3 The last civil war in Switzerland took place in 1847 and emerged from a conflict between the more rural, conservative, Roman Catholic cantons and the more urban, liberal, and mostly Protestant, cantons and ended in the Swiss Federal Constitution of 1848. Today, 36% of the Swiss are Roman Catholic and 24% are Protestant (Source: CIA––The World Factbook: https:// www. cia. gov/ libra ry/ publi catio ns/ theworld- factb ook/ index. html). 636 A.Kind, C.Volonté 1 3 2 Related literature andhypotheses In this study, we focus on the importance of LRDs and their link to firm performance. Boards of directors have different duties—most notably monitoring, advising management, and setting the strategy—in relation to which their competencies, skills, and characteristics must be defined and assessed. Therefore, shareholders should spend considerable time and resources evaluating, selecting, and (re-)electing board directors at annual general meetings. In practice, however, directors are often proposed and elected on the board for other reasons, including their relationship with the CEO, board members, and controlling shareholders, or because of their status and reputation (see, e.g., Cohen etal. 2012). Local roots may be part of the specific set of skills that matter for the ideal profile of board members. For example, locally-rooted directors may have access to valuable local networks. Alternatively, LRDs may be elected just because of their local acquaintanceships, which would reduce their social independence and, consequently, their monitoring efforts. A priori, LRDs may, therefore, have either a positive or a negative (net) influence on firm performance. 2.1 Positive aspects oflocal roots As suggested by the Resource Dependence Theory (Pfeffer and Salancik 1978), LRDs may be beneficial to firms for several reasons. First, an important feature of corporate directors is their access to networks, i.e., the number, importance, and strength of their linkages to the firm’s external environment and stakeholders (e.g., customers, suppliers, financing institutions, or governmental and regulatory institutions). As argued by Koenig and Gogel (1981), locally-rooted board members may have better access to information and resources in the local community where the company is headquartered. In this respect, LRDs may provide added value to the board by offering higher-quality advisory services to management. For example, they are likely better lobbyist because of their privileged relations with local authorities and institutions. This can be beneficial in a variety of situations, such as public tender calls, negotiations related to the expansion of plants, restructurings of operations, the agreement on severance schemes in the aftermath of layoffs, and in obtaining favorable tax treatments (Hillman et al. 1999; Faccio et al. 2006; Duchin and Sosyura 2012). In compliance with this view, Goldman etal. (2013) show that politically connected directors increase procurement contracts (see also Guo etal. 2021). Such privileged relationships may well exist thanks to LRDs. LRDs may also provide networks to local suppliers, the chambers of commerce, or even important local celebrities. Further, as legal disputes are usually resolved by local courts, knowing locally-accredited prosecutors and lawyers can be advantageous. This is especially critical in federated countries where many decisions are made at the local level. Local roots may help build up social capital within a firm and thereby positively affect firm performance. Along these lines, La Porta etal. (1997) argue that social capital contributes to firm value and arises from networks, norms, and mutual recognition. LRDs are also more aware of the country-specific expectations, duties, and responsibilities of board members (Firoozi etal. 2019). Second, an additional positive aspect of LRDs may lay in the fact that they increase mutual trust both inside and outside the board, i.e., between the firm and stakeholders (e.g., employees, state, and NGOs) (see, e.g., Westphal 1999). Mutual trust inside the board is likely to reduce monitoring costs (see Zak and Knack 2001). Trust among people who are 637 Locally-rooted directors 1 3 culturally similar is also higher than amongst culturally dissimilar people (see, e.g., Guiso et al. 2009). In this respect, common local roots decrease uncertainty and information asymmetry among board members and between the board and the CEO, thereby lowering coordination costs (Cai etal. 2017). Third, geographical closeness is another positive feature of LRDs. LRDs are likely to live close to the headquarters. Alam etal. (2014) find that the distance between the directors’ residential address and the headquarters influences their information gathering costs, leading to a trade-off between director expertise and information-gathering costs. Masulis etal. (2012) show that foreign independent directors are negatively related to both firm performance and the intensity of monitoring as measured by the attendance to board meetings, CEO compensation, and the frequency of CEO turnover. They reason that the distance of foreign directors to the headquarters generates oversight costs. For such directors, gathering information about the firm, a country’s economy and business practices, the legal environment, and the institutional environment in general is more costly. Mazur and Salganik-Shoshan (2017) show that the geographic proximity of institutional investors facilitates interpersonal connections and private communication among them, which in turn induces firms to increase incentive-based compensation. Directors with local roots may also be more accessible to inputs from coordinating institutional investors. Accordingly, Lerner (1995) shows that venture capitalists are less likely to sit on boards of distant firms as monitoring intensity is especially high in start-ups. Finally, related to the Stewardship Theory (Donaldson and Davis 1991), LRDs may also act altruistically and in the firm’s best interest because their motivation increases with their identification with the region and its stakeholders. As LRDs identify themselves with the company and the local community and likely feel obliged to help foster their region’s economic development, they should be particularly committed to the firm’s success and intrinsically motivated to exert effort in this direction. In fact, such intrinsic motives may also influence a director’s decision to join the board in the first place (De Jong etal. 2014). As pointed out by Masulis and Mobbs (2014), reputational issues may play an important role in explaining why graduates of a local university are more likely to serve on the board of a closely located firm than on more prestigious boards. 2.2 Negative aspects oflocal roots While the positive features of locally-rooted directors are mostly related to their advising role, the majority of aspects that may have an adverse effect on firm value are associated with their monitoring task. First, from an Agency Theory perspective, the board’s foremost task consists in monitoring the management in the shareholders’ best interests. For an unbiased control of the firm’s resources, directors’ independence is crucial (see Jensen and Meckling 1976; Fama and Jensen 1983). Agency costs arise from the conflict of interest between managers and shareholders. Self-interested managers may engage in a long list of activities that benefit themselves, but harm shareholders: building empires (Jensen 1986), enforcing excessive pay packages (Bertrand and Mullainathan 2001), entrenching themselves (Shleifer and Vishny 1989), shirking (Bertrand and Mullainathan 2003), or using corporate resources for private consumption (Yermack 2006). The traditional view of directors’ independence focuses on the material relationships with the firm. It defines 638 A.Kind, C.Volonté 1 3 directors as either insiders or outsiders (independent directors). Non-independent outsiders are often denominated as “gray” or “affiliated” directors. Independent boards are generally considered to be better monitors and thereby improve firm performance. However, even conventionally defined independent directors can lack true independence by having close relationships or friendships with key executives. As a consequence of weaker monitoring, executives may be replaced too late or paid too much, which may harm shareholders (Adams and Ferreira 2007). The problem of dependent, or captured, board members is particularly severe if the CEO has a strong power in the directors’ nomination process (Shivdasani and Yermack 1999). Researchers have started to investigate the presence of even more subtle social ties among corporate directors or with CEOs (see, e.g., Davis etal. 2003; Conyon and Muldoon 2006; Hwang and Kim 2009; Cohen etal. 2012; Tan etal. 2021). For example, Hwang and Kim (2009) measure the directors’ (in)dependence from the CEO by considering their social ties arising from the same alma mater, shared military service, regional origin, discipline of study, and industry experience and are able to link them to the strength of their monitoring activity. Nguyen (2012) shows that social ties between CEOs and directors decrease the probability of CEOs being dismissed after poor performance. The relevance of social ties between business actors has also been examined in other circumstances. For instance, Ishii and Xuan (2014) show that social ties between acquirers and targets have an adverse effect on the performance of mergers. Because LRDs may have social ties to other (local) board members, controlling shareholders, or the CEO, they may restrain board independence and may, therefore, harm firm performance. In particular, social ties between controlling shareholders and directors may create a certain dependency that induces the latter to help the former to extract private benefits of control. For example, directors may decide that a company shall financially support pet projects of controlling shareholders (e.g., arts or sports). In this respect, LRDs may be nominated because of their social ties rather than their capability to monitor top executives. Second, locally-rooted directors may be more committed to local stakeholders than to their fiduciary duties as directors (see Böhler etal. 2010). For example, they may refrain from closing an unprofitable plant or from switching to a better supplier. Furthermore, LRDs may lack relevant industry-specific and international experience, access to global networks, and general business skills compared to other candidates in the broader supraregional pool of potential corporate directors (see, e.g., Masulis etal. 2012; Oxelheim etal. 2013; Drobetz etal. 2018). Third, in the spirit of this paper, Knyazeva etal. (2013) use the size of the pool of local directors (measured as the number of U.S. nonfinancial firms headquartered near a given firm) as an instrument for board independence. Board independence is shown to be higher when the pool of potential directors is larger. Thus, firms that rely on the rather narrow local market of directors may miss the opportunity to find truly independent directors. Finally, from a social psychological perspective, the Similarity-Attraction Theory posits that individuals and groups have preferences for people who resemble themselves (Byrne and Griffitt 1973). Similarity can refer to psychological characteristics (e.g., shared values or mindsets) or demographic traits (e.g., gender or educational background). Top management teams are inclined to reproduce themselves (Zajac and Westphal 1996; Nielsen 2009). In fact, while new directors are ultimately elected by shareholders at general meetings, candidates are nominated by the incumbent board members. Cronyism and “homophily” within the board may hamper its effectiveness (McPherson etal. 2001). In this context, LRDs may be an important factor for boosting reproduction tendencies within boards. Directors selected on the 639 Locally-rooted directors 1 3 board because of their similarity will not raise potentially controversial opinions that are not in line with the expected view of the group. Such uniformity could harm firm performance and is an argument against the so-called “old boys network” (see, e.g., Adams and Ferreira 2009). In such situations, directors may lower their efforts, receive higher compensation, enjoy fringe benefits (e.g., by organizing board meetings in luxurious surroundings), and protect each other from critical assessments and (potential) liability claims. 2.3 Hypotheses As argued in Sect.2.1, LRDs may provide important linkages to the firm’s (local) external environment and know-how in the local economy. These positive features can lead to a high representation of LRDs on the board of directors. However, as elaborated in Sect.2.2, their appointments may also reflect, at least in some circumstances, the managerial intent to reduce boards’ monitoring and extract private benefits. Notwithstanding the motives for appointing LRDs, we expect them to be in high demand. We therefore formulate the first hypothesis as follows: H1 Locally-rooted directors are overrepresented in corporate boards. On average, we expect the benefits of LRDs (enhanced advisory skills due to local knowhow and links to local political, economic, and regulatory institutions) to outweigh their potential costs (weaker monitoring due to lower independence). Thus, we hypothesize the following: H2 Locally-rooted directors are positively related to firm performance. The board of directors has been argued to be an “endogenously determined institution” (Hermalin and Weisbach 2003, p. 9). Therefore, the optimal composition of the board depends on the firm characteristics and the business environment. In particular, in an international business domain, the benefits of local know-how and local linkages should become comparatively less relevant, while international experience and access to global networks—characteristics that LRDs are likely to lack—should become more important (Masulis etal. 2012; Oxelheim etal. 2013). For instance, as internationally-oriented firms tend to be larger and, therefore, tend to rely more heavily on non-local sources of financing, the value contribution of LRDs in this type of firms should be lower. Thus, as a subset of companies in our sample are highly active in international markets, LRDs in those companies may create costs that exceed the benefits of their local roots. We, therefore, formulate the following hypothesis. H3 In internationally-oriented firms, locally-rooted directors are negatively related to firm performance. On the contrary, the benefits of local know-how and local linkages should become comparatively more important in regulated industries, where LRDs can fully exploit their privileged access to local networks to influence the decisions of local authorities. Due to the strong dependence on regulating authorities, the superior lobbying skills of well-connected, locallyrooted directors should be of particular use in regulated industries. In fact, Helland and Sykuta (2004) find that regulated firms have more directors with a political background. Further, 640 A.Kind, C.Volonté 1 3 Goldman etal. (2009) show that politically-connected directors have a positive impact on firm value. We, therefore, formulate the following hypothesis. H4 In regulated industries, the contribution of locally-rooted directors to firm performance is particularly large. 3 Data andvariables 3.1 Sample In this study, we analyze the effect of locally-rooted directors on firm performance. We derive directors’ local roots from their educational background. The approach is similar to the one used by several scholars for measuring social ties via mutual educational institutions (see, e.g., Cohen etal. 2010; Nguyen 2012; Fracassi and Tate 2012; Ishii and Xuan 2014; Schmidt 2015). In our case, we consider a director as locally rooted if he graduated from the university closest to the company’s headquarters. It is noteworthy that for Swiss companies with multiple production sites, the official headquarters often serves as the primary physical locus of their business operations. This is the place where top management meets, major decisions are made, and the corporate culture develops. Moreover, it is common practice for annual meetings to be held in proximity to these headquarters. The location of the corporate headquarters typically reflects the company’s historical roots, frequently aligning with the site of the company’s founding and occasionally even the birthplace of its founders. These headquarters have often remained the company’s central hubs since their inception.4 Consequently, the headquarters not only symbolizes the identity of a city or town but also embodies deep-rooted traditional ties with the local community. This connection is manifested in various forms, such as sponsorship activities and lobbying efforts. Our definition of local roots offers several advantages. First, it is easily available as it can be collected from the directors’ resumes published in annual reports. Second, it represents an objective and measurable criterion. Third, while it does not consider all possibilities to build up local roots, it ensures that a director classified as locally rooted has been exposed to a certain local environment for at least three years in an age characterized by a steep learning curve. 4 In our dataset of 2035 firm-year observations, we observed 13 instances of headquarters relocations, including one case where a company returned to its original headquarters. In seven of these relocation events, the change of headquarters location did not affect their proximity to university regions, as the new locations were still near the same universities. Of the six remaining relocations, three occurred due to mergers. Typically, such mergers result in significant adaptions in board composition due to changes in ownership. In two of the other three instances, there were no locally-rooted directors on the board either before or after the relocation. The only case where a relocation materially impacted our focus variable was with Valora Holding Ltd. In 2008, the shareholders decided to move the registered office to the operational site. This atypical situation, where the historical registered office was different from the production site, led to an increase in the proportion of locally-rooted directors from 20 to 40%. 647 Locally-rooted directors 1 3 4 Empirical results 4.1 Overrepresentation oflocally‑rooted directors We start the empirical analysis by measuring the overrepresentation of locally-rooted directors who graduated from any of the seven universities and the two federal technical universities in Switzerland located in one of the seven main university regions we consider in this study. Following Grinblatt and Keloharju (2001), we calculate overrepresentation by dividing for all Swiss companies in a given region (e.g., Basel) the average number of local graduates on the board (e.g., for the board of Roche Holding AG, graduates from University of Basel) by the average number of directors who graduated from this university. Table3 shows that in all regions the directors with a local university degree are heavily 20 % 22 % 24 % 26 % 28 % 30 % 32 % 34 % Locally rooted directors Fig. 2 Development of locally-rooted directors Table 3 Summary statistics documenting over representation of locally-rooted directors (H1) The table presents the over representation of 9 Swiss university graduates and their representation on boards in different regions in Switzerland. The figures represent the difference between the number of graduates in the regions and the Swiss average Number of firm-years University Federal Technical University Region Basel Bern Fribourg Geneva Lausanne St. Gallen Zurich ETH EPF Basel 315 4.1 1.0 0.4 0.9 0.1 0.6 0.8 0.7 1.5 Bern 189 0.8 2.4 0.0 0.9 0.5 0.7 0.6 1.2 0.2 Fribourg 36 1.5 0.0 8.8 2.6 1.6 0.4 0.1 0.4 0.0 Geneva 121 0.5 0.4 0.7 5.5 1.7 0.2 0.5 0.5 1.6 Lausanne 191 0.0 0.5 1.8 2.2 6.2 0.6 0.5 0.8 7.0 St. Gallen 184 0.7 0.5 0.9 0.1 0.2 2.6 1.3 1.0 0.0 Zurich 999 0.5 1.1 0.9 0.4 0.5 1.0 1.2 1.2 0.2 648 A.Kind, C.Volonté 1 3 overrepresented (in bold). The results support Hypothesis 1, according to which locallyrooted directors are overrepresented in corporate boards. Similarly to the “home bias” in stock ownership, which depends on familiarity, distance, language, and culture (see, e.g., Coval and Moskowitz 1999; Grinblatt and Keloharju 2001), there is, potentially, also a local bias in the selection of board members. Although Switzerland is a comparatively small country and distances should likely play a minor role in director selections, the market for board directors seems to be subject to a strong cultural segmentation. In our sample, almost 30% of all directors can be defined as locally-rooted directors. As indicated in Fig.2, the proportion of locally-rooted directors on boards decreased in the last ten years. By accepting the view that boards of directors are endogenous and optimally determined (Hermalin and Weisbach 2003), this drop suggests that the value of locally-rooted directors has diminished over the years. Table 4 Tobin’s Q and locally-rooted directors (H2) The table presents regression coefficient estimates for Tobin’s Q. The sample consists of 2035 firm-year observations. Cluster-robust standard errors are reported in parentheses, and significance at the 1%, 5%, and 10% levels is indicated by ***, **, *, respectively Independent Variables Dependent variable: Tobin’s Q (I) (II) (Intercept) 1.45949 (***) 1.62315 (***) (0.395) (0.407) Locally-rooted directors −0.40381 (***) (0.118) Size −0.02539 −0.03317 (*) (0.019) (0.019) Sales growth 0.10134 0.12886 (0.191) (0.192) Firm age −0.01253 0.00127 (0.034) (0.033) Profitability 3.75637 (***) 3.80576 (***) (0.526) (0.510) Liquidity 0.72868 (***) 0.73495 (***) (0.234) (0.229) Investments 3.33021 (***) 3.07806 (***) (1.014) (0.979) Tangibility −1.11394 (***) −1.02349 (***) (0.174) (0.176) R&D 1.64828 (***) 1.65490 (***) (0.454) (0.435) Leverage 0.20420 0.14189 (0.195) (0.188) Fixed effects Industries, Years Industries, Years Adjusted R251.77% 52.92% 649 Locally-rooted directors 1 3 4.2 Locally‑rooted directors andfirm performance: baseline model Table 4 presents regression results on the relationship between Tobin’s Q and locallyrooted directors using cluster-robust Huber/White standard errors. Controlling for industry effects, time trends, and a battery of controls, the results suggest that locally-rooted directors are negatively related with firm performance. In particular, all else being equal, boards with a ten percentage points larger fraction of locally-rooted directors are associated with lower Tobin’s Q by 4.04 (Model III) to 5.55 (Model IV) percentage points. Put differently, moving from a fraction of locally-rooted directors equal to the first quartile (11%) to the third quartile (43%) lowers the predicted Tobin’s Q by 12.82 percentage points (based on Model III) to 17,61 percentage points (based on Model IV). Thus, the relation between locally-rooted directors and Tobin’s Q is significant both in statistical and in economic terms. Therefore, we have to reject Hypothesis 2, which is that locally-rooted directors are positively related to firm performance. Thus, despite the numerous good reasons for appointing LRDs to corporate boards (see Sect.2.1), the empirical findings suggest that in the average firm, the negative aspects of LRDs (most likely related to their weaker monitoring and lack of skills and experience, see Sect.2.2) dominate on the margin. 4.3 Locally‑rooted directors andfirm performance: instrumental variables approach Alongside the omitted-variable bias, reverse causation is another potential source of endogeneity. Higher firm performance may induce firms to seek directors from more distant regions (even from abroad) instead of locally-rooted directors, because the former may possess relevant experience and specialist know-how that the latter likely lack. To address this issue, we use the seven university locations in Switzerland as instruments for LRDs and estimate our model using 2SLS. We believe that University locations in Switzerland are valid instruments as they likely comply with the IV’s relevance condition and exclusion condition. According to the relevance condition, the chosen instrument—in our case the university regions—must be strongly correlated with the endogenous explanatory variable, in our case the fraction of locally-rooted directors. Otherwise, the IV estimates become biased and inefficient, and the resulting inference may be unreliable. On theoretical grounds, we expect the fraction of LRDs in a firm to be influenced by the regional environment of the companies’ headquarters for several reasons. In fact, the regions considered in the paper are characterized by distinctive cultural and geographic features that affect both the local demand and the local supply of different types of directors. On the one hand, language and religion (Mayer 1951), as well as other local peculiarities, e.g., the degree of urbanization and political orientation (Steenbergen 2010), likely influence the level of people’s trust and their openness to directors who are not familiar with the headquarters’ environment (see, e.g., Guiso etal. 2009). In some regions, the overlap of people involved in economic and non-economic activities is higher than in others, which explains the regional variations in the relevance of local roots for board appointments. On the other hand, the presence of important international airports in some of the regions considered (Basel, Geneva, and Zurich) but not in others, and the different levels of agglomerations (highest in Basel, Bern, Geneva, Lausanne, and Zurich with one-third of the overall Swiss population) likely influence the relevant supply of different types of directors. Fortunately, the relevance condition can be tested empirically, and this is usually done through appropriate tests in the 650 A.Kind, C.Volonté 1 3 first-stage regression. In our setting, the F-statistics of the joint significance of the excluded instruments (i.e., dummy variables for university locations) in the first-stage regression is 12.02. According to Staiger and Stock (1997), a value of over 10 indicates the relevance of the instruments. Second, with respect to the exclusion condition, the direct formal test of whether the instrument is uncorrelated with the error term in the outcome equation is not feasible because the error term is unobservable. Therefore, theoretical justifications are particularly important. First, the characteristics of the regions are determined outside the model and are, therefore, at least for the purposes of this study, exogenous. In particular, it is unlikely that the short-term success (or the lack of it) of the firms in a region influences the social, cultural, and geographic characteristics of that region. Second, these geographic regions are not likely to have direct effects on firm performance for reasons we do not account for in the regressions, e.g., industry affiliation. In particular, none of the listed companies in our sample is dependent on the regions’ economic conditions because they all sell their goods either in the other Swiss regions or internationally. Finally, as already argued, relocations are very rare, and the majority of companies are well-rooted in their region. The results in Table5 show that even when using the university regions as instrumental variables, LRDs are negatively related to Tobin’s Q in the second-stage IV regression, which supports a causal interpretation of the results. 4.4 Locally‑rooted directors ininternationally‑oriented firms andfirms inregulated industries The optimality of board composition depends on a firm’s external environment (Hermalin and Weisbach 2003). As the value of LRDs differs across companies that operate in different environments, in Table6, we re-run our baseline model on a number of subsamples: (i) internationally-oriented firms vs. domestically-oriented firms (Model I vs. Model II) and (ii) companies in regulated vs. non-regulated industries (Model III vs. Model IV). The last two rows of Table6 show that the proportion of LRDs is significantly higher in companies without foreign sales (31.1% vs. 27.9%) and in regulated industries (31.6% vs. 27.8%). These descriptive findings comply with the conjecture that LRDs are generally less valuable and less demanded in internationally-oriented and non-regulated firms. The regressions in Table6 analyze the relation of LRDs with firm performance across four relevant subsamples. Strikingly, LRDs are not significantly related to firm performance in domestically-oriented companies (no foreign sales, Model II) and regulated firms (Model III). Based on these results and following the view of Hermalin and Weisbach (2003) that “boards are endogenously driven institutions”, the higher proportion of LRDs chosen by firms in domestically-oriented sectors and regulated industries may be optimal. Conversely, the negative and significant coefficients of LRDs in internationally-oriented companies (Model I) and in firms in non-regulated industries (Model IV)—i.e., precisely those firms in which the benefits of LRDs are expected to be lower—show that the use of LRDs in those firms is excessive, despite being already significantly lower than in the other firms. According to Hypothesis 3, LRDs may harm firm value if a firm is internationally oriented. Our model, therefore, accounts for a firm’s internationalization profile. To examine the impact of LRDs in internationally-oriented firms, we include the interaction terms Locally-rooted directors × Foreign sales (dummy) and Locally-rooted directors × Number of geographic segments. Foreign sales (dummy) is a dummy variable and equals one if the company has positive sales abroad (and zero otherwise). Number of geographic segments 651 Locally-rooted directors 1 3 Table 5 Instrumental variables approach: University locations The table presents regression coefficient estimates for Tobin’s Q. The sample consists of 2035 firm-year observations. Dummy variables indicate the name of the closest university to the company’s headquarters (e.g., Basel). The University of Geneva (i.e., Geneva) serves as the reference group and is set to zero. Cluster-robust standard errors are reported in parentheses, and significance at the 1%, 5%, and 10% levels is indicated by ***, **, *, respectively Independent variables Dependent variables First stage Locally-rooted directors (I) Second stage Tobin’s Q (II) (Intercept) 0.29414 (*) 2.03317 (***) (0.152) (0.490) Locally-rooted directors −1.41544 (**) (0.707) Basel 0.10166 (0.066) Bern 0.07889 (0.075) Fribourg −0.00056 (0.143) Lausanne 0.08166 (0.072) St. Gallen 0.20246 (***) (0.077) Zurich 0.10874 (*) (0.058) Size −0.01777 (**) −0.05266 (**) (0.008) (0.023) Sales growth 0.04456 0.19779 (0.093) (0.236) Firm age 0.03164 (**) 0.03585 (0.015) (0.046) Profitability 0.11483 3.92950 (***) (0.124) (0.505) Liquidity 0.02455 0.75066 (***) (0.075) (0.234) Investments −0.66283 (***) 2.44639 (**) (0.206) (1.026) Tangibility 0.20568 (**) −0.79690 (***) (0.088) (0.271) R&D 0.02623 1.67150 (***) (0.151) (0.424) Leverage −0.14855 (*) −0.01421 (0.078) (0.225) Fixed effects Industries, Years Industries, Years Adjusted R217.25% 48.01% F-statistics for joint-significance of instruments (University locations) 12.02 (***) 652 A.Kind, C.Volonté 1 3 is the logarithm of the number of geographic segments as indicated by the segment information in the annual report. Traditionally, Swiss firms have a large share of export sales. Similarly, Hypothesis 4 states that in regulated industries, the contribution of locally-rooted directors to firm performance is particularly large. We, therefore, include the interaction term Locally-rooted directors × Regulated industries (dummy) to investigate this channel. Regulated industries comprise banks, insurance companies, financial services, and utilities. In Table7, Model II, the coefficient of the interaction term of locally-rooted directors with foreign sales (or, alternatively, with the number of geographic segments in Model IV) is negative and significant, while the coefficient of locally-rooted directors as a stand-alone variable ceases to be statistically significant. We cannot, therefore, reject Hypothesis 3, which posits that Locally-rooted Directors (LRDs) are negatively related to firm performance in internationally-oriented firms. Contrarily, Model VI demonstrates that the impact of LRDs in regulated industries is significantly more substantial than in the rest of the sample, thus supporting Hypothesis 4. However, it’s important to note that this effect on Tobin’s Q in regulated industries is derived by summing the coefficients of Locally-rooted directors (−0.49674) and the interaction term Locally-rooted directors × Regulated industries (0.51080). Consistent with the findings of Table6, Model III, this calculation suggests that the influence of LRDs on Tobin’s Q in regulated industries is negligible and not statistically significant. 4.5 Channels ofthenegative effect oflocally‑rooted directors onfirm performance In order to better understand the underlying mechanisms for the contribution of LRDs on firms’ valuations, we adopt the following procedure. First, we test whether, in our sample, LRDs and non-LRDs differ along selected professional characteristics that, according to past research (Agency Theory and Resource Dependence Theory), are related to the performance of board directors. Second, we test whether any of those features mediate the effects of LRDs on Tobin’s Q. Third, we focus on alternative explanations of the results by running additional regression models with appropriate interaction terms. For this purpose, we report in Table8 the differences in board directors’ features. Panel A refers to characteristics related to the directors’ (lack of) independence, e.g., board independence (Knyazeva et al. 2013) and tenure (Huang and Hilary 2018). Panel B reports information on directors’ activities that may reduce their attention to the matters of the focal firm but may also provide access to valuable resources, e.g., the number of external activities (Cashman etal. 2012), busy directors (Fich and Shivdasani 2006), external CEO positions (Fahlenbrach etal. 2010), and directors with political connections (Goldman etal. 2009). Finally, Panel C reports additional information on board directors that have attracted considerable attention in past research as drivers of directors’ performance, e.g., industry know how, CEO experience, financial know how, and international experience (e.g., Oxelheim et al. 2013; Volonté and Gantenbein 2016) and gender (Bennouri etal. 2018; Chen etal. 2023a). Strikingly, Table8 shows several significant differences between LRDs and non-LRDs, which may help explain the results presented in the previous section. Specifically, LRDs are more often former company executives, have more often business relationships with the company, and have longer tenure, on average. However, Panel A also indicates that LRDs’ stronger lack of independence is not due to their role as executive directors. Panel B shows that LRDs pursue fewer external activities, especially concerning high-level external positions as CEO and board chairman, which may indicate that they are less qualified. On 653 Locally-rooted directors 1 3 Table 6 Tobin’s Q and locally-rooted directors in different subsamples Independent Dependent variable: Tobin’s Q variables (I) (II) (III) (IV) (Intercept) 1.47931 (***) 3.27376 (***) 0.99189 (***) 1.66688 (***) (0.466) (1.030) (0.121) (0.477) Locally-rooted directors −0.47686 (***) 0.10497 0.02100 −0.48568 (***) (0.143) (0.113) (0.046) (0.141) Size −0.02516 −0.09612 (***) −0.00775 −0.03787 (0.022) (0.034) (0.005) (0.024) Sales growth 0.32881 −0.47402 0.26422 (***) 0.11202 (0.242) (0.315) (0.097) (0.241) Firm age 0.02167 −0.02267 0.00810 −0.00699 (0.039) (0.032) (0.011) (0.042) Profitability 3.66733 (***) 3.25806 (***) 2.52730 (***) 3.90743 (***) (0.503) (0.971) (0.506) (0.528) Liquidity 1.00057 (***) 0.08235 0.30747 (**) 0.82327 (***) (0.268) (0.254) (0.125) (0.257) Investments 4.69335 (***) 0.46282 −0.11533 3.24397 (***) (1.219) (0.601) (0.442) (1.047) Tangibility −1.33605 (***) −1.20271 (***) −0.14690 −1.02080 (***) (0.305) (0.355) (0.183) (0.192) R&D 1.14664 1.32604 (***) 0.24309 (**) 1.59649 (***) (0.826) (0.332) (0.114) (0.443) Leverage −0.01485 0.84937 (***) 0.84937 (***) 0.14605 (0.225) (0.259) (0.259) (0.206) Subset Foreign sales No foreign sales Regulated industries Non regulated industries Observations 1,535 500 479 1,556 Fixed effects Industries, Years Industries, Years Industries, Years Industries, Years 654 A.Kind, C.Volonté 1 3 The table presents regression coefficient estimates for Tobin’s Q. The sample consists of 2035 firm-year observations. Foreign sales is a dummy variable and equals one if the company has positive sales abroad (and zero otherwise). Regulated industries is a dummy variable and equals one if the industry is regulated (and zero otherwise). Clusterrobust standard errors are reported in parentheses, and significance at the 1%, 5%, and 10% levels is indicated by ***, **, and *, respectively Table 6 (continued) Independent Dependent variable: Tobin’s Q variables (I) (II) (III) (IV) Adjusted R251.18% 63.62% 49.20% 47.36% LRDs in the subsamples 27.9% 31.1% 31.6% 27.8% Difference in LRDs (t-test) 3.19 p. p (2.74) (***) 3.87 p. p (3.43) (***) 655 Locally-rooted directors 1 3 Table 7 Tobin’s Q and locally-rooted directors in internationally-oriented firms (H3) and regulated industries (H4) Independent variables Dependent Variable: Tobin’s Q (I) (II) (III) (IV) (V) (VI) (Intercept) 1.61874 (***) 1.56839 (***) 1.62676 (***) 1.58390 (***) 1.62315 (***) 1.62807 (***) (0.406) (0.407) (0.411) (0.406) (0.407) (0.409) Locally-rooted directors −0.40312 (***) −0.04633 −0.40717 (***) 0.00912 −0.40381 (***) −0.49674 (***) (0.118) (0.112) (0.119) (0.117) (0.118) (0.140) Foreign sales (dummy) 0.01693 0.15420 (*) (0.059) (0.085) Number of geographic segments (log) 0.01673 0.13529 (0.059) (0.083) Regulated industries (dummy) −0.24722 −0.34605 (*) (0.203) (0.203) Locally-rooted directors × Foreign sales (dummy) −0.45826 (***) (0.171) Locally-rooted directors × Number of geographic segments (log) −0.40246 (***) (0.133) Locally-rooted directors × Regulated industries (dummy) 0.51080 (***) 656 A.Kind, C.Volonté 1 3 The table presents regression coefficient estimates for Tobin’s Q. The sample consists of 2035 firm-year observations. Foreign sales is a dummy variable and equals one if the company has positive sales abroad (and zero otherwise). Number of geographic segments is the number of geographic segments as indicated by the segment information in the annual report. Regulated industries is a dummy variable and equals one if the industry is regulated (and zero otherwise). Cluster-robust standard errors are reported in parentheses, and significance at the 1%, 5%, and 10% levels is indicated by ***, **, *, respectively Table 7 (continued) Independent variables Dependent Variable: Tobin’s Q (I) (II) (III) (IV) (V) (VI) (0.153) Size −0.03417 (*) −0.03894 (*) −0.03501 −0.04210 (*) −0.03317 (*) −0.03299 (*) (0.020) (0.020) (0.023) (0.023) (0.019) (0.019) Sales growth 0.12872 0.13518 0.13150 0.14067 0.12886 0.13342 (0.192) (0.194) (0.191) (0.191) (0.192) (0.193) Firm age 0.00088 0.00366 0.00140 0.01183 0.00127 −0.00185 (0.033) (0.033) (0.033) (0.034) (0.033) (0.033) Profitability 3.80726 (***) 3.81245 (***) 3.80922 (***) 3.82460 (***) 3.80576 (***) 3.81411 (***) (0.509) (0.506) (0.508) (0.504) (0.510) (0.507) Liquidity 0.74139 (***) 0.73388 (***) 0.74184 (***) 0.75953 (***) 0.73495 (***) 0.72290 (***) (0.227) (0.225) (0.227) (0.228) (0.229) (0.228) Investments 3.05931 (***) 3.08101 (***) 3.05412 (***) 3.04401 (***) 3.07806 (***) 3.05358 (***) (0.991) (0.980) (0.988) (0.983) (0.979) (0.971) Tangibility −1.01228 (***) −1.01321 (***) −1.01082 (***) −1.00263 (***) −1.02349 (***) −0.99561 (***) (0.187) (0.184) (0.181) (0.179) (0.176) (0.178) R&D 1.65650 (***) 1.72663 (***) 1.64881 (***) 1.65264 (***) 1.65490 (***) 1.66629 (***) (0.434) (0.434) (0.437) (0.441) (0.435) (0.431) Leverage 0.14378 0.14053 0.14033 0.12110 0.14189 0.13750 (0.187) (0.185) (0.188) (0.185) (0.188) (0.187) Fixed effects Industries, Years Industries, Years Industries, Years Industries, Years Industries, Years Industries, Years Adjusted R252.90% 53.14% 52.91% 53.31% 52.92% 53.20% 663 Locally-rooted directors 1 3 Table 11 Channel: Interrelationship with local shareholders Independent variables Dependent variable: Tobin’s Q (I) (II) (III) (IV) (Intercept) 1.60866 (***) 1.68664 (***) 1.68680 (***) 2.08781 (***) (0.401) (0.420) (0.420) (0.486) Locally-rooted directors −0.33778 (**) −0.47795 (***) −0.47724 (***) −0.48762 (***) (0.166) (0.149) (0.148) (0.147) Locally-controlled firm 0.01115 (0.100) Long term locally-controlled firm −0.17315 (*) −0.17482 (*) −0.18158 (*) (0.104) (0.102) (0.104) Locally-rooted directors × −0.09397 Locally-controlled firm (0.228) Locally-rooted directors × 0.27616 0.27386 0.29006 Long term locally-controlled firm (0.225) (0.230) (0.231) Dual class 0.00743 0.00844 (0.081) (0.084) Shareholding directors 0.00215 (0.117) Size −0.03334 (*) −0.03211 (*) −0.03212 (*) −0.03240 (*) (0.019) (0.019) (0.019) (0.019) Sales growth 0.12456 0.11695 0.11832 0.10806 (0.192) (0.192) (0.189) (0.190) Firm age 0.00257 0.00427 0.00404 0.00126 (0.034) (0.034) (0.034) (0.034) Profitability 3.79975 (***) 3.83691 (***) 3.83660 (***) 3.86198 (***) (0.509) (0.506) (0.507) (0.515) Liquidity 0.74132 (***) 0.75346 (***) 0.75321 (***) 0.75119 (***) 664 A.Kind, C.Volonté 1 3 The table presents regression coefficient estimates for Tobin’s Q. The sample consists of 2035 firm-year observations. Locally-controlled firm is a dummy variable and equals one if the company is controlled by a Swiss shareholder with more than 20% of voting rights (and zero otherwise). Long term locally-controlled firm is a dummy variable and equals one if the company has been controlled by the same Swiss shareholder with more than 20% of voting rights over the full period (and zero otherwise). Dual class is a dummy variable and equals one if the company has two or more classes of equity issued (and zero otherwise). Shareholding directors is the proportion of directors with significant shareholding or who represent significant shareholders. Cluster-robust standard errors are reported in parentheses, and significance at the 1%, 5%, and 10% levels is indicated by ***, **, *, respectively Table 11 (continued) Independent variables Dependent variable: Tobin’s Q (I) (II) (III) (IV) (0.230) (0.230) (0.230) (0.230) Investments 3.08309 (***) 3.03299 (***) 3.03045 (***) 3.03123 (***) (0.984) (0.998) (0.996) (0.998) Tangibility −1.01980 (***) −1.01943 (***) −1.01790 (***) −1.01847 (***) (0.181) (0.175) (0.175) (0.174) R&D 1.63957 (***) 1.60226 (***) 1.60107 (***) 1.62119 (***) (0.438) (0.436) (0.435) (0.432) Leverage 0.14908 0.11743 0.11818 0.11369 (0.189) (0.185) (0.185) (0.186) Fixed effects Industries, Years Industries, Years Industries, Years Industries, Years Adjusted R252.90% 53.22% 53.19% 53.39% 665 Locally-rooted directors 1 3 Table 12 Alternative specifications of locally-rooted directors Independent variables Dependent variable: Tobin’s Q (I) (II) (III) (IV) (Intercept) 1.99702 (***) 1.99236 (***) 1.96789 (***) 1.90453 (***) (0.449) (0.442) (0.461) (0.455) Locally-rooted non-executive directors −0.45225 (***) (0.120) Locally-rooted independent directors −0.47413 (***) (0.127) Locally-rooted directors (without two federal technical universities) −0.29311 (**) (0.141) Locally-rooted directors’ overrepresentation −0.00871 (**) (0.004) Size −0.02952 −0.02786 −0.02875 −0.02540 (0.019) (0.019) (0.020) (0.019) Sales growth 0.10321 0.06304 0.10046 0.08242 (0.194) (0.196) (0.193) (0.192) Firm age 0.00133 −0.00432 −0.01079 −0.01364 (0.033) (0.033) (0.034) (0.033) Profitability 3.82493 (***) 3.84101 (***) 3.79624 (***) 3.78393 (***) (0.512) (0.515) (0.524) (0.527) Liquidity 0.73164 (***) 0.72958 (***) 0.72339 (***) 0.72929 (***) (0.227) (0.227) (0.233) (0.233) Investments 3.11832 (***) 3.20760 (***) 3.20465 (***) 3.23320 (***) (0.986) (0.977) (0.990) (1.017) Tangibility −1.01284 (***) −1.03527 (***) −1.05671 (***) −1.07355 (***) (0.174) (0.173) (0.178) (0.174) R&D 1.69260 (***) 1.68635 (***) 1.66853 (***) 1.68418 (***) 666 A.Kind, C.Volonté 1 3 The table presents regression coefficient estimates for Tobin’s Q. The sample consists of 2035 firm-year observations. Locally-rooted non-executive directors is the proportion of locally-rooted directors who are non-executive on the board. Locally-rooted independent directors is the proportion of locally-rooted directors who are independent on the board. Locally-rooted directors (without two federal technical universities) is the proportion of locally-rooted directors on the board without considering the two federal technical universities. Locally-rooted directors’ overrepresentation is the ratio of the number of locally-rooted directors to the average number of locally-rooted directors within the company’s headquarters region. Cluster-robust standard errors are reported in parentheses, and significance at the 1%, 5%, and 10% levels is indicated by ***, **, *, respectively Table 12 (continued) Independent variables Dependent variable: Tobin’s Q (I) (II) (III) (IV) (0.431) (0.438) (0.440) (0.445) Leverage 0.13019 0.13883 0.20486 0.19402 (0.189) (0.190) (0.191) (0.193) Fixed effects Industries, Years Industries, Years Industries, Years Industries, Years Adjusted R253.29% 53.12% 52.32% 52.12% 667 Locally-rooted directors 1 3 particular in domestically-oriented companies and firms in regulated industries—their presence on boards is not associated with lower valuations and may thus be value-enhanc- ing and optimal. Appendix See Table 13, 14, 15 and 16. Funding Open access funding provided by University of Basel. Table 13 Universities in Switzerland The table presents figures about Swiss universities and their graduates’ representation on corporate boards of directors in SPI firms Source: www. swiss unive rsiti es. ch (access on 04.08.2015) and own data base Sample universities University figures Sample figures Standard deviation Number of students Year of foundation Fraction of directors (%) Maximum of directors (%) University of Basel (BS) 12,982 1460 5.08 90.00 0.118 University of Bern (BE) 15,406 1834 3.35 66.67 0.080 University of Fribourg (FR) 10,084 1889 1.42 66.67 0.059 University of Geneva (GE) 15,514 1559 2.79 75.00 0.083 University of Lausanne (LS) 12,947 1537 3.41 66.67 0.093 University of St. Gallen (SG) 7809 1898 10.33 85.71 0.150 University of Zurich (ZH) 26,351 1833 10.80 75.00 0.147 EPF Lausanne (EPF) 9395 1969 1.16 42.86 0.048 ETH Zurich (ETH) 17,309 1855 13.65 80.00 0.177 Excluded universities University of Swiss Italian Region (–) 2918 1995 0.00 0.00 0.000 University of Lucerne (–) 2654 1848 0.00 0.00 0.000 University of Neuchatel (–) 4345 1838 0.28 20.00 0.020 Study sample statistics All university graduates (incl. foreign universities) 76.60 100.00 0.205 668 A.Kind, C.Volonté 1 3 Table 14 Home cantons from students and graduations from Swiss universities in 1981 U BS (%) U BE (%) U FR (%) U GE (%) U LS (%) U LU (%) U NE (%) U SG (%) U ZH (%) U SI (%) EPFL (%) ETHZ (%) 1981 Graduates per Total Population (%) Zürich 0 1 1 2 0 0 0 363 0028 0.07 Bern 1 63 2 4 2 0 4 3 4 01 16 0.05 Luzern 13 16 11 1 1 2 1 9 24 00 22 0.07 Uri 4 26 19 0 0 7 0 4 22 00 19 0.08 Schwyz 8 15 25 3 0 0 0 10 15 00 25 0.04 Obwalden 0 26 11 0 0 0 0 11 26 00 26 0.07 Nidwalden 22 39 17 0 0 0 0 0 11 00 11 0.06 Glarus 0 12 0 0 0 0 0 12 65 00 12 0.05 Zug 2 6 0 2 0 0 0 11 66 00 13 0.06 Fribourg 1 9 47 11 12 0 1 1 2 07 9 0.09 Solothurn 23 30 4 1 0 2 0 6 12 00 23 0.07 Basel-Stadt 80 1 1 3 0 0 0 1 4 00 9 0.13 Basel-Land- schaft 73 1 2 1 0 1 0 1 5 00 15 0.10 Schaffhausen 0 5 0 7 0 0 2 5 34 00 48 0.06 Appenzell A 10 7 3 3 0 0 3 14 28 00 31 0.06 Appenzell I 0 14 0 0 0 0 0 29 57 00 0 0.05 St. Gallen 5 13 7 2 0 0 1 14 33 00 24 0.07 Graubünden 6 12 6 4 2 0 0 4 45 01 21 0.07 Aargau 14 8 1 1 1 0 0 6 41 00 27 0.06 Thurgau 2 13 3 2 1 0 0 12 39 00 28 0.05 Ticino 2 9 14 21 10 0 1 3 15 03 22 0.09 Vaud 0 2 0 16 62 0 1 0 1 014 4 0.08 Valais 1 9 19 34 22 1 1 1 1 04 8 0.09 Neuchâtel 0 4 1 20 15 0 48 1 0 03 7 0.09 669 Locally-rooted directors 1 3 Table 14 (continued) U BS (%) U BE (%) U FR (%) U GE (%) U LS (%) U LU (%) U NE (%) U SG (%) U ZH (%) U SI (%) EPFL (%) ETHZ (%) 1981 Graduates per Total Population (%) Genève 0 1 0 87 10 0 0 1 07 2 0.15 Jura 2 2 6 29 25 0 21 2 0 04 10 0.08 Average 10 12 5 16 9 0 2 4 22 03 17 0.08 The table shows the proportion of graduates in each canton graduating from the various universities in Switzerland Bold indicates graduates from cantons that have a university and who graduated from these universities (i.e., canton of Basel-Stadt and University of Basel, U BS) Italic indicates the three universities (Neuchatel, Luzern, and Lugano) or an insignificant number of graduates in our sample 670 A.Kind, C.Volonté 1 3 Table 15 Firms’ headquarters and their closest university Headquarters Postcode Canton Region Closest University Travel time by car Distance in km Aigle 1860 VD Lausanne LS 36 44 Allschwil 4123 BL Basel BS 13 6 Altdorf 6460 UR Other region ZH 57 76 Arbon 9320 TG St. Gallen SG 17 16 Baar 6340 ZG Central Switzerland ZH 29 34 Bad Ragaz 7310 SG Other region SG 54 83 Baden 5400 AG Swiss Plateau ZH 28 25 Basel 4000 BS Basel BS 0 0 Bern 3000 BE Bern BE 0 0 Biel/Bienne 2500 BE Bern BE 33 41 Boudry 2017 NE Other region LS 45 65 Brusio 7743 GR Other region SG 3h01 237 Bubendorf 4416 BL Basel BS 26 23 Bubikon 8608 ZH Zurich ZH 32 28 Buchs (AG) 5033 AG Swiss Plateau ZH 26 15 Burgdorf 3400 BE Bern BE 27 25 Cham 6330 ZG Central Switzerland ZH 27 31 Cheseaux-sur-Lausanne 1033 VD Lausanne LS 12 9 Chéserex 1275 VD Geneva GE 32 29 Chur 7000 GR Other region SG 1h05 103 Dierikon 6036 LU Central Switzerland ZH 37 45 Dietlikon 8305 ZH Zurich ZH 18 13 Domat/Ems 7013 GR Other region SG 1h10 109 Dornach 4143 SO Basel BS 17 14 Dottikon 5605 AG Swiss Plateau ZH 36 36 Düdingen 3186 FR Other region FR 15 11 Eglisau 8193 ZH Zurich ZH 28 28 Emmen 6032 LU Central Switzerland ZH 40 50 Flamatt 3175 FR Other region BE 19 18 Frauenfeld 8500 TG Other region SG 36 48 Fribourg/Freiburg 1700 FR Other region FR 0 0 Apples 1143 VD Lausanne LS 27 22 Genève 1200 GE Geneva GE 0 0 Gerlafingen 4563 SO Swiss Plateau BE 28 32 Gland 1196 VD Lausanne LS 32 36 Glarus 8750 GL Other region ZH 51 70 Granges-Marnand 1523 VD Other region FR 31 24 Gränichen 5722 AG Swiss Plateau ZH 46 48 Hergiswil 6052 NW Central Switzerland ZH 41 58 Herisau 9100 AR St. Gallen SG 13 11 Hinwil 8340 ZH Zurich ZH 31 29 Hochdorf 6280 LU Swiss Plateau ZH 45 49 Horgen 8810 ZH Zurich ZH 20 21 671 Locally-rooted directors 1 3 Table 15 (continued) Headquarters Postcode Canton Region Closest University Travel time by car Distance in km Horw 6048 LU Central Switzerland ZH 43 55 Interlaken 3800 BE Other region BE 44 58 Ittigen 3063 BE Bern BE 11 6 Jona 8645 SG Zurich ZH 31 40 Kilchberg (ZH) 8802 ZH Zurich ZH 12 6 Kloten 8302 ZH Zurich ZH 17 13 Küsnacht (ZH) 8700 ZH Zurich ZH 13 7 Laufenburg 5080 AG Other region BS 33 40 Lausanne 1000 VD Lausanne LS 0 0 Lenzburg 5600 AG Swiss Plateau ZH 39 38 Liestal 4410 BL Basel BS 22 19 Locarno 6600 TI Ticino ZH 2h28 200 Lupfig 5242 AG Swiss Plateau ZH 33 32 Luterbach 4542 SO Swiss Plateau BE 35 36 Luzern 6000 LU Central Switzerland ZH 42 52 Lyss 3250 BE Bern BE 26 29 Männedorf 8708 ZH Zurich ZH 29 20 Morges 1110 VD Lausanne LS 16 14 Moutier 2740 BE Other region BS 56 54 Muttenz 4132 BL Basel BS 15 7 Neuhausen 8212 SH Other region ZH 46 52 Niederwangen 3172 BE Bern BE 10 8 Niederweningen 8166 ZH Zurich ZH 35 24 Oberdorf (BL) 4436 BL Basel BS 32 30 Oberkirch 6208 LU Swiss Plateau ZH 53 71 Olten 4600 SO Swiss Plateau BS 45 54 Perlen 6035 LU Central Switzerland ZH 36 44 Pfäffikon (SZ) 8808 SZ Zurich ZH 26 36 Plan-les-Ouates 1228 GE Geneva GE 13 5 Porrentruy 2900 JU Other region BS 1h04 69 Prilly 1008 VD Lausanne LS 6 3 Quartino 6572 TI Ticino ZH 2h25 193 Regensdorf 8105 ZH Zurich ZH 20 12 Reinach (BL) 4153 BL Basel BS 17 9 Rorschacherberg 9404 SG St. Gallen SG 19 15 Rümlang 8153 ZH Zurich ZH 19 14 S. Antonino 6592 TI Ticino ZH 2h11 184 Schaffhausen 8200 SH Other region ZH 43 52 Schindellegi 8834 SZ Zurich ZH 27 31 Schlieren 8952 ZH Zurich ZH 19 9 Sion 1950 VS Other region LS 1h05 94 St. Gallen 9000 SG St. Gallen SG 0 0 Stäfa 8712 ZH Zurich ZH 32 23 672 A.Kind, C.Volonté 1 3 Table 15 (continued) Headquarters Postcode Canton Region Closest University Travel time by car Distance in km Stans 6370 NW Central Switzerland ZH 46 64 Stein am Rhein 8260 SH Other region ZH 57 56 Steinach 9323 SG St. Gallen SG 16 15 Steinhausen 6312 ZG Central Switzerland ZH 28 30 St-Prex 1162 VD Lausanne LS 21 18 Tägerwilen 8274 TG Other region SG 52 43 Thalwil 8800 ZH Zurich ZH 15 12 Uster 8610 ZH Zurich ZH 24 24 Uznach 8730 SG Other region ZH 39 55 Vaduz 9490 FL Other region SG 45 68 Vaz 7082 GR Other region SG 1h26 125 Vernier 1214 GE Geneva GE 17 7 Vevey 1800 VD Lausanne LS 25 19 Waldenburg 4437 BL Basel BS 32 31 Wattwil 9630 SG Other region SG 40 37 Solothurn 4500 SO Swiss Plateau BE 35 40 Wetzikon 8620 ZH Zurich ZH 29 29 Winterthur 8400 ZH Zurich ZH 31 27 Wolfenschiessen 6386 NW Central Switzerland ZH 1h02 76 Yverdon 1400 VD Other region LS 33 39 Zermatt 3920 VS Other region LS 2h23 172 Zofingen 4800 AG Swiss Plateau BS 42 53 Zug 6300 ZG Central Switzerland ZH 31 34 Zürich 8000 ZH Zurich ZH 0 0 The table presents the firms’ headquarters location, its postcode, the firms’ region and its closest university, as well as the travel time and distance from the headquarters to this university