The impact of corporate governance provisions on R&D intensity: a closer look at corporate governance in an international perspective

The impact of corporate governance provisions on R&D intensity:

a closer look at corporate governance in an international

perspective

Msc. International Financial Management, DD program Faculty of Business and Economics, University of Groningen

Department of Business Studies, Uppsala Universitet

Abstract:

Using panel data analysis, this paper considers the impact of governance provisions on firm’s R&D investment decisions. The current paper also contributes to the literature on corporate governance and innovation by introducing an interaction dimension which captures the influence that internationalization of U.S. firms may have on R&D investment decisions. Based on a sample of 627 U.S. firms for the years 2008-2016 this paper’s results suggest that governance provisions do not nurture or impede R&D investments, and therefore the results casts doubt on the existence of a relationship between governance provisions and R&D intensity. The current paper’s findings also suggest that proposed effects of internationalization do not modify the relationship between entrenchment and R&D for U.S. firms.

1.

Introduction

Investments in innovative activities such as research and development (henceforth, R&D) play a critical role in the creation of new products and processes, driving firms to attain higher efficiency and long run competitive advantages (Holmstrom, 1989). Although investments in R&D are vital to the long term value of firms (Hitt et al., 1996), R&D investments bear increased risk with respect to return on investment, in particular on short run accounting performance measures (Bhagat and Welch, 1995; Hoskisson et al., 1993). As a result, managers and shareholders may have differing attitudes towards opportunities and risks associated with R&D investments.

From an agency perspective, delegation of R&D investment decisions to managers may lead to conflicts of control between shareholders and managers of the firm (Jensen and Meckling, 1976). First and foremost, shareholders and managers have different perceptions of, and attitudes towards the risk of R&D investment (Becker-Blease, 2011). Secondly, shareholders have asymmetric information compared to managers on investment in innovative activities. To align interest of managers with interests of shareholders, shareholders must address the difficulties regarding external valuation of risk and high costs of monitoring managerial efforts to innovate. The agency framework assumes that when managers are insufficiently monitored they will exhibit opportunistic behaviors, resulting in agency costs. Therefore, shareholders aim to install an effective framework of corporate governance mechanisms to counteract such agency costs. Effective corporate governance mechanisms are paramount for shareholders to mitigate agency costs associated with the separation of ownership and control, considering the vitality of R&D investment strategies for firms’ long-term success (Holmstrom, 1989).

governance, or merely view corporate governance as a control variable for exogeneity in their studies of firm performance.

This paper distinguishes itself from preceding agency studies by introducing an interaction dimension, capturing the influence that internationalization may have on firms’ managerial R&D decision-making. Arguing that internationalization both reduces the informational asymmetry and provides shareholders with a broader set of governance mechanisms to valuate and monitor managerial efforts to innovate. Moreover, the aftermath of the 2007-2008 collapse of the global banking system has overturned the corporate governance framework, leading to renewed interest in the topic from scientists and practitioners.

To do so, this paper positions itself in a strand of the corporate governance literature studying corporate governance practices that influence shareholders’ rights (Danielson and Karpoff, 2006; Gompers et al., 2003). The current paper studies the influence of a specific type of corporate governance mechanism: governance provisions. Governance provisions are defined as any governance mechanism in a firm’s corporate charter or bylaws that is aimed to limit the shareholders control over management (Bebchuk et al., 2008). This paper attempts to contribute to the corporate governance debate by exploring for U.S. firms what effect governance provisions have on investments in R&D, and whether this effect is moderated by an interactional dimension of firm internationalization.

2.

Literature review and hypothesis development

Prior literature indicates that the presence of governance provisions has implications for management’s efforts to maximize shareholder value (Gompers et al., 2003). Gompers et al., (2003) constructed a governance index containing twenty-four governance provisions and linking the presence of these provisions to changes in shareholder value. Subsequently, Bebchuk et al., (2008) empirically found that six out of these twenty-four governance provisions are the main drivers within the relationship between governance provisions and shareholder value. Bebchuk et al., (2008) identify that these six provisions all are specifically aimed at preventing shareholders from forcing a change of control. This implies that the threat of change of control has a powerful influence on the creation of shareholder value. The Entrenchment index (henceforth, E-index) is an aggregate of these six provisions ranging from 0 to 6. Consistent with the literature, this paper argues a higher E-index implies stronger protection from change in control (i.e. higher degree of entrenchment). Following the findings of the Bebchuk et al., (2008), Chakraborty et al., (2014) provide evidence that a higher E-index is both associated with lower shareholder value and lower innovative performance. This finding suggests that the inferior shareholder value of high E-index firms may be explained by a different managerial approach to innovation.

Within the scope of this paper it is examined how the presence of governance provisions comprising the E-index alter managerial efforts of firms to invest in R&D, and whether this relation is moderated by a firm’s degree of internationalization. Prior to investigating this moderating effect of internationalization, the relationship between governance provisions and the likelihood of managers to invest in R&D is examined through two contrasting strands of literature. The agency theory framework predicts that a higher degree of entrenchment could have two opposing effects on managerial incentives to invest in innovative activities (Hitt et al., 1996; Hoskisson et al., 1993; Stein, 1988).

2.1 Entrenchment view and incentives to invest in R&D

maximization will lead management to cut back on investments (Holmstrom, 1989). This is especially true for investments in innovative activities because R&D investments are unlikely to yield direct short-term benefit for managers (Shleifer and Vishny, 1989). This is caused by the fact that R&D projects have a higher probability of failure compared to routine projects (Hall, 2002; Hall et al., 2001; Holmstrom, 1989), and are characterized by long term payoff structures (Vithessonthi and Racela, 2016). This implies that if any revenues of R&D investments would surface, this would likely happen beyond tenures of management (Dechow and Sloan, 1991; Gibbons and Murphy, 1992). Independently from entrenchment, prior findings suggest that managers already have the tendency to underinvest in R&D due to lack of adequate incentives (Holmstrom, 1989). The entrenchment view predicts that the presence of governance provisions might further exacerbate this underinvestment in R&D projects through the pursuit of personal utility maximization.

Prior empirical findings underpin this negative relationship between governance provisions and R&D expenditures. For example, Mahoney et al., (1997) show that the introduction of governance provisions results in lower long-term investments. Meulbroek et al. (1990) find that firms decrease R&D expenditures when management adopts governance provisions. Seru (2014) shows that conglomerates innovate less when managerial entrenchment increases. Based on these findings, it may be predicted that the presence of governance provisions (i.e. higher degree of entrenchment) negatively affects management’s efforts to invest in R&D. This paper labels that prediction as the entrenchment Hypothesis.

H1a: Management’s R&D expenditures are negatively related to the presence of governance provisions as expressed by the E-index.

2.2 Managerial myopia view and incentives to invest R&D

term projects with more certain payoff structures (Chemmanur and Tian, 2018). Previous literature labels this phenomenon “corporate myopia”.

Corporate myopia also relates to Stein’s (1988) “managerial myopia view”. Managerial myopia view states that due to information asymmetry shareholders cannot properly assess the potential of management’s investment in long-term innovation (Meulbroek et al., 1990; Stein, 1998). It may be argued that R&D activities are specifically associated with information asymmetry for two reasons. Firstly, R&D innovations are likely to be an intangible asset that is often, compared to tangible assets, difficult to valuate for shareholders (Stein, 1988). Secondly, R&D investments are clouded by secrecy to protect their competitive advantage (Humphery-Jenner, 2014). Consequently, short-term oriented equity market investors (i.e. shareholders) will be prone to undervalue the shareholders’ value of firms’ investments in long-term R&D (Chemmanur and Tian, 2018). The potential undervaluation of innovative firms results in an increased likelihood to be targeted for hostile acquisition attempts (Meulbroek et al., 1990). Managers might try to reduce information asymmetry by cutting down on R&D investment (Meulbroek et al., 1990). By shifting efforts to routine tasks with more certain short-term returns, market value is increased through the reduction of R&D associated undervaluation (Chakraborty et al., 2014). Hence, according to the myopia view, managers attempt to reduce the threat of hostile acquisition by diminishing information asymmetry between managers and shareholders through R&D investment reductions.

In the managerial myopia view, governance provisions that hinder the change of control and insulate managers from hostile acquisition pressures will enable managers to overcome shareholder induced myopia. In diametrical opposition to the entrenchment hypothesis, this theory suggest that governance provisions will incentivize firm’s R&D investment (Humphery-Jenner and Powell, 2011; Shleifer and Vishny, 1989; Stein, 1988).

are produced by Lerner et al., (2011), finding that a higher percentage of privately owned shares - implying a more long-term orientation- as opposed to publicly owned shares leads to more innovative investment.

In short, shareholders that are characterized by longer investment horizons are more focused on innovation and less likely to pressure management to focus on short-term results, alleviating the managerial myopia problem (Chemmanur and Tian, 2018). Chemmanur and Jiao (2012) confirm this line of reasoning and show that the adoption of specific governance provisions enables managers to pursue long-term R&D projects rather than short-term projects. Hence, sheltering management from short term performance pressures exerted by shareholders alleviates managerial myopia concerns. Therefore, the presence of governance provisions may enable managers to engage in long term investments in R&D activities. This paper labels the prior argumentation as the managerial myopia hypothesis.

H1b: Management’s R&D expenditures are positively related to the presence of governance provisions as expressed by the E-index.

2.3 Interaction effects of firm internationalization on entrenchment-R&D

relationship

The focus of the current paper is to introduce internationalization as an interaction dimension within the entrenchment-R&D relationship. This paper investigates whether the degree of internationalization of U.S. firms modifies how governance provisions affect R&D

decision-making. To further examine this interaction effect, prior theoretical work is examined.

2.3.1 Interaction effects of internationalization on R&D investment within managerial entrenchment framework

convergence towards this composition of superior corporate governance mechanisms (Denis and McConnell, 2003). Prior works explored and proposed various mechanisms through which practices could converge around the globe.

For example, Aggarwal et al. (2011) find that international (institutional) investors are a vehicle through which corporate governance practices migrate. Foreign investors were shown to transfer superior governance mechanisms to local firms with inferior governance mechanisms. Therefore, international investors could be associated with corporate governance migration and convergence towards a set of global standards. Firm internationalization, similarly to international investments, may be a channel through which governance mechanisms migrate across borders. More specifically, through internationalization U.S. firms become increasingly subjected to foreign legal environments that influence shareholder protection (Porta et al., 1998). Also, internationalization exposes U.S. firms to non-U.S. institutional environments that influence the shareholders’ capability of monitoring (Aggarwal et al., 2011). U.S. firms potentially voluntarily or coercively import and adopt ‘exotic’ non-U.S. corporate governance mechanisms into their U.S. corporate governance framework. In other words, through internationalization foreign corporate governance mechanisms may migrate and get adopted into U.S. corporate governance practice. Thus, the degree of internationalization may affect the governance policies of a given firm, modifying managerial R&D investment outcome.

With regard to external pressure, internationalization may lead firms to interact with different types of markets for corporate control (Jensen and Ruback, 1983). In particular, it is found that the threat of hostile acquisitions (i.e. strength and activity of the market for corporate control) differs between countries and corporate governance frameworks (Denis and McConnell, 2003; O’Sullivan, 2001; Prowse, 1994). Moreover, Aguilera and Jackson (2003) argue that the greatest threat of a change in control through hostile acquisitions exists in the U.S. market-based economy. Essentially, within the U.S. markets-market-based corporate governance framework managers are argued to be incentivized mainly by the threat of hostile acquisitions (Koke, 2001).

frameworks the governance mechanism of threat of hostile acquisitions is partly substituted by other governance mechanisms (e.g. long-term debt financing, ownership by families and large block-holders and shareholders litigations) (Becht et al., 2003; Porta et al., 1998). Therefore, a higher degree of internationalization by U.S. firms suggests exposure to an overall weaker market for corporate control by inclusion of non-U.S. corporate governance frameworks. Altogether, when viewing the effect of internationalization through the lens of the entrenchment view, the effect of governance provisions on R&D expenditures may be influenced by three factors related to internationalization. Namely, migration of governance mechanisms, altered monitoring abilities and varying external pressure exerted by the market for corporate control. Through these factors, the effect of governance provisions examined in this study possibly exert a reduced effect on the entrenchment-R&D relationship for firms with higher degree of internationalization. Hence, this paper predicts that a higher degree of internationalization has a negative interaction effect on R&D investments.

H2a: The negative relationship between management’s R&D investment and governance provisions is negatively moderated by a higher degree of internationalization

2.3.2 Interaction effects of internationalization on R&D investment within managerial myopia framework

Alternatively, it may be argued that when firms internationalize, they will face increased competition from an increased number of foreign competitors. It is argued that an increased (international) market competition causes firms to disclose more and to be more open to stakeholders (Nickell et al., 1997). This increased openness is especially beneficial for firms with high levels of information a-symmetry, such as R&D intensive firms. As explained earlier, the monitoring and valuation of R&D intensive firms by shareholders is relatively complex. Consequently, equity market investors are less equipped to appreciate the R&D investments conducted, and will be prone to undervalue the shareholders’ value of firms investing in long-term innovative activities (Chakraborty et al., 2014).

to more accurately assess the potential of a manager’s investment in innovation, then the undervaluation effect of information asymmetries will be mitigated by internationalization in the managerial myopia view. The managerial myopia resulting from the threat of hostile acquisition and subsequent the underinvestment in R&D are decreased by a higher degree of firm internationalization. Thus, internationalization dampens short term performance pressures on management from hostile acquisitions, easing the managerial myopia problem. As a consequence, governance provisions that protect managers from myopia problems that are associated with informational a-symmetries exert a reduced influence on the entrenchment-R&D relationship. Hence, internationalization has a negative interaction effect on entrenchment-R&D expenditures.

H2b: The positive relationship between management’s R&D investment and governance provisions is negatively moderated by a higher degree of internationalization.

3.

Sample description and methodology

3.1 Sample description

The sample examined in this paper includes U.S. publicly traded firms during the period 2008– 2016. The base sample is constructed by combining data from two different sources. Using CUSIP company identifiers the governance provisions data from the ‘Institutional Shareholder Services-governance database’ (formerly RiskMetrics) is merged with financial data regarding innovation, internationalization and other firm-level control variables from the Compustat Capital-IQ annual database.

The ISS-governance database covers all firms included in the S&P-1500 index. The S&P 1500 index combines the S&P 500 index with the S&P mid- and the small-Cap indices, covering approximately 90% of the U.S. market capitalization (Becker-Blease, 2011). Within the ISS-governance database the S&P 1500 index firms are accompanied with other publicly traded firms that are primarily selected on market capitalization and high institutional ownership

levels1. Furthermore, Baumol (2004) argues that large established firms, such as largely

represented by the S&P 1500 index, are responsible for the vast majority of investment in R&D. Whereas this paper’s results indicate that firms in the ISS-Governance database are keenly adopting governance provisions. Therefore, the ISS-Governance database offers a suitable

environment to further examine the relation between governance provisions and investment in R&D.

This paper adopts a panel dataset to examine the moderating effect of internationalization on the relationship between governance provisions and investment in R&D for two principal arguments. First, by estimating a 9-year period this study can apprehend changes in firm’s degree of internationalization and managerial entrenchment enabling inferences whether these changes influence firm’s R&D expenditures. Second, by taking a longitudinal instead of a cross-sectional perspective this study overcomes potential concerns of time specific business cycle characteristics biasing the regressions analysis.

In order to obtain this paper’s sample, all accessible data on governance provisions within the ISS-Governance database for U.S. publicly traded firms were obtained for the period between 2007-2016. Resulting in 2,323 unique U.S. firms with 10-year data observations regarding governance provisions. For all 10 years in this sample these 2,323 firms are supplemented with firm specific financial data regarding innovation, internationalization and other firm-level control variables from the Compustat Capital-IQ annual database. Resulting in an unbalanced panel sample. Preceding the regression analyses, this unbalanced panel sample is trimmed by means of three limiting factors. First, through introduction of lagged and time-dummy variables, the observed sample period is shortened by one year. All individual firm observations from the year 2007 are dropped from the final sample, resulting in a sample spanning a nine-year period between 2008 and 2016. Despite the reduced number of firm-nine-year observations, none of the firms were dropped as a result of this data trimming measure. The reasoning behind the introduction of these variables is further discussed in the methodology section. Secondly, for the estimation of the fixed effects regression models the statistical software package selected 632 out of the 2,323 firms with sufficient data across all variables and years. This is, all firms with sufficient data on the variables and years are selected to econometrically responsibly estimate the fixed effects model. Finally, consistent with the majority of empirical studies on finance and innovation, firms in finance, banking, insurance and real estate industries are

removed from the final sample2. Reducing the final sample to 627 unique firms with 3,530

firm-year observations. On average, the sample contains 5.6 firm-years of observations per firm. The descriptive statistics and correlation matrix is shown in table 2 in the results section.

3.2 Methodology

3.2.1 Measurement of variables

3.2.1.1 Measuring independent variable of Entrenchment: Governance Provisions

Following the methodology described in Bebchuk et al. (2008), this paper constructs an Entrenchment-index (henceforth, E-index) based on six out of twenty-four six governance provisions included in the ISS – Governance database measuring the strength of shareholder rights. Bebchuk et al. (2008) discuss that all six provisions within their E-index are aimed at hindering change of control for incumbent management. A higher E-index implies stronger entrenchment concerning change in control and vice versa.

The provisions within the E-index could be separated into two distinct types. The first type of provisions (staggered boards, limits to shareholder amendments of the bylaws, limits to shareholder amendments of the charter, and supermajority requirements for mergers) pose statutory constrains on voting power for shareholder to force a change in control at annual meetings (Bebchuk et al., 2008). The second type (poison pills and golden parachutes) are designed to increase cost for the acquirer in case of a hostile acquisition attempt, and thus, insulate management from a change of control as a result of acquisitions attempt (Bebchuk et al., 2008). Definitions of governance provisions encompassing the E-index are provided in Appendix II.

All the governance provisions in the E-index are assigned identical weights. Hence, when a firm introduces or dissolves a governance provision a single point is added or removed from the E-index. The easy to use and widely-adopted E-index of Bebchuk et al. (2008) owes it popularity to the fact that their subset of governance provisions has shown to be the strongest driver for a range of different performance measures (Bebchuk et al., 2013). Since this paper is interested in the effect of governance provisions on a specific measure of performance this index is a suitable measure for managerial entrenchment regarding change in control.

3.2.1.2 Measuring dependent variable: R&D expenditure ratios

innovation and therefore R&D expenditures reflect the intensity with which management is pursuing the potential future benefits of innovation (Danielson and Karpoff, 2006; Lhuillery, 2011). Therefore, R&D expenditure based ratios provide a direct measure for management’s efforts to innovate. Two common R&D based proxies for innovation are employed: R&D-intensity ratio and R&D-sales ratio, the former is defined as R&D expenditures in million dollars divided by total book value of the assets in million dollars and the latter is defined as R&D expenditures in million dollars divided by total sales in million dollars.

Utilizing R&D expenditures based ratios as proxy for innovation is appealing since under ‘United States Generally Accepted Accounting Principles’ (hereinafter, GAAP), R&D expenditures must be expensed instantaneously and therefore are well documented and observable. Additionally, R&D expenditures could be effectuated reasonably quickly after managerial decisions to invest or disinvest in innovation and are therefore clearly linked to managerial intent to innovate. So, by its very nature, firm’s R&D expenditures have the potential of producing a well-defined causal relation between governance provisions and innovation efforts.

3.2.1.3 Measuring of interaction effect: degree of internationalization

To examine whether internationalization moderates the entrenchment–innovation expenditure relationship, this study estimates a measure of entrenchment that is weighted by the degree of internationalization. Prior theoretical contributions indicate that internationalization is a multidimensional construct with multiple complexities in estimating a firms’ degree of internationalization (Kotabe et al., 2002), and no consensus has emerged among researchers regarding the measurement of degree of internationalization (Burgman, 1996).

overcome the potential mixture of trade and internationalization, and to further ensure robustness of internationalization results, this paper uses foreign income ratios as a proxy for internationalization.

3.2.1.4 Control variables

This study also controls for other variables that may affect the relation between managerial entrenchment and R&D. Three firm specific variables which previous research indicate that could be related to R&D are examined. First, previous contributions show that controlling for firm size is crucial, because it suggested that larger firms have lower returns on innovation (Humphery-Jenner and Powell, 2011). It is argued, that smaller firm structures are better adapted to an innovative perspective and have higher tolerance of risk taking behavior (Holmstrom, 1989). Therefore, firm size could affect management’s efforts to invest in R&D. This present paper includes firm size measured by the natural logarithm of total assets in million dollars to control for these size effects, and expects firm size to be negatively related to innovation. Second, previous literature indicate that a firm’s operating performance may also influence innovation. For example, Eberhart, Maxwell, & Siddique (2004) argue operating returns are positively related to innovation performance. Therefore, firms with higher operation performance might be prone to invest more in innovation than firms with lower operating performance. This paper includes return on assets (ROA), defined as operating income in million dollars dived by total asset in million dollars, as a control measure of firm operating performance. The expectation is that return on assets is positively related to R&D. Third, previous literature suggests that financial constraints resulting from interest and principal payment of debt, may force management to forsake profitable investment (Tallman and Li, 1996). Therefore, firms with higher debt may invest less in promising R&D projects than firms with lower debt. This paper includes leverage measured by book value of long-term debt over total value of the assets (in millions of dollars) to control for financial constraints, and expect leverage to be negatively related to R&D expenditures.

3.2.2 Empirical methodology

significant difference across firm’s could be observed. The null is rejected (Prob > F = 0.000 < 0.05), thus the sample contains a panel data effect and either a random or fixed effects model is selected.

Thereafter, the current paper conducts the Hausman-test to decide whether the fixed or random effects model is more appropriate estimation technique. Whereas the null hypothesis is that the random effects model must be selected versus the alternative hypothesis that the fixed effects. The Hausman-test essentially controls whether the unique errors are uncorrelated with the regressors, against the alternative that the unique errors are correlated with the regressors. The null hypothesis is rejected (Prob > F = 0.0462 < 0.05), thus the errors are correlated and the fixed effects model is applied.

Finally, it controls for the presence of time fixed effects in the fixed effect model by means of the Chow-test. The Chow-test jointly tests an auxiliary regression with a dummy variable for all years, controlling for time-variation within the parameters and specifying whether the parameters have changed once during the sample. In this sample the Chow-test rejects the null hypothesis that the coefficients for all years are jointly equal to zero (Prob > F = 0.0174 < 0.05), therefore time fixed effects are applied in this study. Results of the three tests indicate that fixed time and entity effect model is the appropriate estimation technique. A comparison between different estimation techniques is provided in Appendix III.1. The results of Breusch-Pagan Lagrange multiplier, Hausman test, and Chow-test are provided in appendix III.2 to appendix III.4.

Second, Wooldridge test for autocorrelation in panel data is conducted to control whether there are patterns in the error terms. Under the null hypothesis, the error term has no first order autocorrelation, and the alternative hypothesis states that the error term has first order autocorrelation. For the studied sample the null hypothesis is rejected (Prob > F = 0.000), thus the error term has first order autocorrelation (i.e. there are patterns in the error terms). To ensure the robustness of the standard errors, the standard errors are corrected with the Newey-West HAC (heteroscedasticity and autocorrelation consistent standard) procedure.

3.2.3 Model estimation

To test Hypotheses 1a and 1b, the current paper employs a fixed time and entity effect model that controls for the relationship between two defined R&D ratios and managerial entrenchment and other control variables. The models’ variables and their definitions are discussed in appendix I.

!&# %&'(&)%'*

+,-!&# )./()0,1 = 𝛽 𝐸𝑁𝑇𝑅%,'89+ 𝛽 𝑆𝐼𝑍𝐸%,'89+ 𝛽 𝑅𝑂𝐴%,'89 + 𝛽 𝐿𝐸𝑉%,'89+ α%+ 𝜆'+ 𝑢%,'

[Equation 1] Where:

𝑅&𝐷 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦%,' or 𝑅&𝐷 𝑠𝑎𝑙𝑒𝑠%,' are the dependent variables, where i = entity and t = time.

𝛽 𝐸𝑁𝑇𝑅_%,'89 represents the independent variable of managerial entrenchment 𝛽 𝑆𝐼𝑍𝐸_%,'89 represents the control variable of firm size

𝛽 𝑅𝑂𝐴%,'89 represents the control variable of firm return on assets

𝛽 𝐿𝐸𝑉_%,'89 represents the control variable of firm leverage

α% represents the entity fixed effect

𝜆' represents the time fixed effect

𝜇%' represents the error term

!&# %&'(&)%'*0,1 !&# )./()0,1 = 𝛽 𝐸𝑁𝑇𝑅%,'89+ 𝛽 𝐼𝑁𝑇𝐸𝑅 %,'89+ 𝛽 𝑆𝐼𝑍𝐸%,'89 + 𝛽 𝑅𝑂𝐴%,'89 + 𝛽 𝐿𝐸𝑉%,'89+ 𝛼%+ 𝜆' + 𝑢%,' [Equation 2] !&# %&'(&)%'*0,1 !&# )./()0,1 = 𝛽 𝐸𝑁𝑇𝑅%,'89+ 𝛽 𝐼𝑁𝑇𝐸𝑅 %,'89+ 𝛽 𝐸𝑁𝑇𝑅%,'89∗ 𝛽 𝐼𝑁𝑇𝐸𝑅%,'89+ 𝛽 𝑆𝐼𝑍𝐸%,'89+ 𝛽 𝑅𝑂𝐴%,'89 + 𝛽 𝐿𝐸𝑉%,'89+ 𝛼% + 𝜆'+ 𝑢%,' [Equation 3] Where:

𝛽 𝐼𝑁𝑇𝐸𝑅%,'89 represents the independent variable of firm internationalization

𝛽 𝐸𝑁𝑇𝑅%,'89∗ 𝛽 𝐼𝑁𝑇𝐸𝑅%,'89 represents the interaction variable of firm entrenchment and firm

internationalization

Equivalently, the entity and time fixed effects regression model can be written in terms of a least squares dummy variable regression model with inclusion of a single intercept. The specifications of the rewritten time and firm fixed regression models that is utilized for further

estimation is discussed in appendix V3.

4.

Results

4.1 Descriptive statistics

3 _{For all models of R&D intensity the measurement of internationalization variable and the interaction effect is based on the INTERrdi, while}

for all models of R&D sales the measurement of internationalization and the interaction effect is based on the INTERrds. Definitions of all proxies could be found in Appendix table I.

Table 1: Descriptive statistics and correlations.

Variable Mean Min Max Std dev. 1 2 3 4 5 6 7 8 9 10

1. R&D intensity 3,530 0.051 0 0.601 0.057 1 2. R&D sales ratio 3,530 0.075 0 3.358 0.109 0.779*** 1 3. Managerial entrenchment 3,530 3.535 0 6 1.067 -0.022 -0.022 1 4. Dummy staggerd board and poison pill 3,530 0.124 0 1 0.330 0.0611*** 0.0385* 0.498*** 1 5. Dummy golden parachute and poison pill 3,530 0.158 0 1 0.364 0.014 0.018 0.474*** 0.586*** 1 6. Internationalization 3,530 0.486 -56.981 83.026 3.242 0.007 0.013 -0.011 0.004 0.004 1 7. Managerial entrenchement * internationalization 3,530 1.678 -184.321 249.077 10.701 0.008 0.014 0.032 0.031 0.030 0.974*** 1 8. Firm size 3,530 3.405 1.759 5.902 0.691 -0.253*** -0.142*** -0.163*** -0.142*** -0.107*** 0.032 0.023 1 9. Return on assets (ROA) 3,530 0.057 -1.541 0.783 0.096 -0.133*** -0.183*** -0.024 -0.019 -0.011 -0.010 -0.012 0.138*** 1 10. Firm leverage 3,530 0.193 0 1.929 0.177 -0.016 -0.026 0.0588*** 0.033 0.028 -0.009 -0.005 0.030 -0.0336* 1 See Appendix I. for variable definitions. * p<0.05, ** p<0.01, *** p<0.001

Table 1 displays the descriptive statistics of the characteristics of the 627 firms examined in this study. Table 1 indicate that the firms in the sample, on average, adopt 3.54 governance provisions and have yearly R&D investments equal to 5.10 per cent of firm’s book value of the assets. Similarly, it is depicted that the sample firms obtain on average 48.6 per cent of firm’s total income from foreign operations. Additionally, from the frequency statistics in table 2, it could be observed that the adoption of governance provisions is a widely-spread practice for the U.S. firms in the sample, for example 99.83 per cent of firms adopt at least 1 provision whereas 34.03 per cent adopts the median of 3 governance provisions. The frequencies of observed normal distributed E-scores (Shapiro-Wilk W test for normality prob. > Z = 0.063) is graphically depicted in figure 1.

From the correlation matrix reported in table 1, it could be observed that the independent variables of managerial entrenchment, internationalization and the moderating interaction factor exhibit no significant correlation with the dependent variables R&D intensity and R&D sales. This is derived from the Pearson correlation coefficient, that ranges from 1 to +1. This result indicates no clearly distinguishable trend between the regressors and the dependent variable. Confirming this paper’s predictions, the correlation matrix in table 1 shows that two control variables, respectively firm size and firm ROA show a weak to moderate negative correlation with R&D intensity and are statistically significant.

E-score Frequency Per cent Cumulative

0 6 0.170% 0% 1 69 1.955% 2% 2 482 13.650% 16% 3 1,201 34.020% 50% 4 1,106 31.330% 81% 5 577 16.350% 97% 6 89 2.521% 100% Number of Firms 627 Total No. Of observations 3,530

Table 2: Frequency score table E-scores with per centages

0 500 1000 1500 F re q u e n cy 0 2 4 6 E-Score

4.2 Results fixed time and entity effect regression models

Table 3 presents three fixed time and entity effect models that control for the hypothesized relationships. The R-squared from all models indicate that the explanatory power of the independent variables on R&D intensity ratio is superior to the explanatory power of the independent variables on R&D sales ratio. Therefore, the interpretation and discussion of the results will be merely focused on the models that explain the dependent variable of R&D intensity.

The beta coefficient of entrenchment in model 1 indicate that R&D intensity ratio increases with one per mille when E-score increases by one. On that premise, the model suggests that this causal link between entrenchment and R&D intensity is significant on 5 a per cent confidence interval. However, the size-effect is so small (one per mille), that the effect does not appear to be practically relevant. Furthermore, results of model 2 indicate that beta coefficient of internationalization has a very small, one per ten mille, non-significant effect on R&D intensity

RDI RDS

Variable Model 1 Model 2 Model 3 Base model Model 1 Model 2 Model 3 Base model

Managerial entrenchment 0.00140** 0.00141** 0.00132* 0.00132* 0.000830 0.000822 0.000952 0.000952 (0.000696) (0.000696) (0.000717) (0.000717) (0.00140) (0.00142) (0.00133) (0.00133) Internationalization 0.000109 -0.000595 -0.000595 -8.13e-05 0.00100 0.00100 (0.000124) (0.000973) (0.000973) (0.000213) (0.00121) (0.00121) Entrenchment * 0.000216 0.000216 -0.000333 -0.000333 (0.000311) (0.000311) (0.000414) (0.000414) Firm size -0.0242*** -0.0243*** -0.0244*** -0.0244*** -0.00219 -0.00209 -0.00195 -0.00195 (0.00790) (0.00791) (0.00792) (0.00792) (0.0147) (0.0148) (0.0148) (0.0148) Firm leverage -0.0142** -0.0141** -0.0142** -0.0142** -0.0492*** -0.0492*** -0.0492*** -0.0492*** (0.00558) (0.00558) (0.00559) (0.00559) (0.0168) (0.0168) (0.0167) (0.0167) Return on assets (ROA) 0.00563* 0.00571* 0.00565* 0.00565* 0.00353 0.00347 0.00356 0.00356 (0.00309) (0.00310) (0.00309) (0.00309) (0.00541) (0.00544) (0.00542) (0.00542) 2009 -0.00628*** -0.00631*** -0.00630*** -0.00630*** -0.00372 -0.00370 -0.00372 -0.00372 (0.00215) (0.00215) (0.00215) (0.00215) (0.00374) (0.00374) (0.00374) (0.00374) 2010 -0.00823*** -0.00823*** -0.00827*** -0.00827*** -0.00504 -0.00504 -0.00498 -0.00498 (0.00226) (0.00226) (0.00227) (0.00227) (0.00703) (0.00703) (0.00709) (0.00709) 2011 -0.00607*** -0.00608*** -0.00602*** -0.00602*** -0.00959** -0.00958** -0.00967** -0.00967** (0.00201) (0.00201) (0.00200) (0.00200) (0.00420) (0.00419) (0.00424) (0.00424) 2012 -0.00443** -0.00444** -0.00441** -0.00441** -0.00644* -0.00643* -0.00648* -0.00648* (0.00202) (0.00202) (0.00201) (0.00201) (0.00387) (0.00386) (0.00390) (0.00390) 2013 -0.00349* -0.00351* -0.00349* -0.00349* -0.00286 -0.00285 -0.00287 -0.00287 (0.00205) (0.00205) (0.00205) (0.00205) (0.00370) (0.00368) (0.00368) (0.00368) 2014 -0.00255 -0.00254 -0.00251 -0.00251 -0.00185 -0.00186 -0.00190 -0.00190 (0.00214) (0.00214) (0.00214) (0.00214) (0.00299) (0.00300) (0.00301) (0.00301) 2015 -0.00283 -0.00282 -0.00280 -0.00280 -0.00284 -0.00285 -0.00288 -0.00288 (0.00217) (0.00217) (0.00217) (0.00217) (0.00341) (0.00342) (0.00345) (0.00345) 2016 -0.00288 -0.00286 -0.00282 -0.00282 -0.00127 -0.00128 -0.00135 -0.00135 (0.00224) (0.00224) (0.00224) (0.00224) (0.00439) (0.00442) (0.00447) (0.00447) Constant 0.132*** 0.132*** 0.133*** 0.133*** 0.0850* 0.0847* 0.0838* 0.0838* (0.0267) (0.0267) (0.0268) (0.0268) (0.0479) (0.0481) (0.0483) (0.0483) Observations 3,530 3,530 3,530 3,530 3,530 3,530 3,530 3,530 Number of firms 627 627 627 627 627 627 627 627 R-squared 0.040 0.041 0.042 0.042 0.008 0.008 0.008 0.008 Adj. R-squared 0.0371 0.0372 0.0378 0.0378 0.0045 0.0043 0.0042 0.0042 Newey-West Standard robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

when holding all the other independent variables constant. Model 3 produces a non-significant beta coefficient regarding the interaction effect of two per ten mille. Hence, no statistical support is found that the hypothesized internationalization interaction effect significantly moderates the entrenchment–R&D intensity relationship, holding all the other independent variables constant.

All R&D intensity models show to have a R-squared of 4.2 per cent or lower. The R-squared explains the fraction of the variation in R&D intensity that is accounted for by the regressors. In addition, this paper considers the adjusted squared. The advantage of the adjusted R-squared is that it resolves various problems associated with the regular R-R-squared, but more importantly the adjusted R2 penalizes the model for insertion of additional independent variables. All models R-squared are between 3.7 and 3.8 per cent. This indicates that the practical implications and the generalizability of the results should take into account the limited explanatory power of the models.

4.3 Robustness checks and further analyses

In this section, this paper performs additional analysis and robustness checks to control whether the primary findings are affected by applying alternative measures and it addresses the potential concerns of endogeneity.

4.3.1 Controlling for unequal potency of governance provisions

To do so, this paper employs two additional models with dummy variables to capture the effect of the described combinations on the relationship between managerial entrenchment and investments in innovative activities. This is reflected by the fourth regression model, that controls for whether the presence of the combination of staggered board with poison pill and the fifth regression model that that controls for the presence of the combination of golden parachutes with poison pill. The models’ variables and their definitions are discussed in appendix I. !&# %&'(&)%'*0,1 !&# )./()0,1 = 𝛽 𝐷𝑠𝑝𝑖,𝑡−1+ 𝛽 𝑆𝐼𝑍𝐸𝑖,𝑡−1+ 𝛽 𝑅𝑂𝐴𝑖,𝑡−1 + 𝛽 𝐿𝐸𝑉𝑖,𝑡−1+ α𝑖+ 𝜆𝑡+ 𝑢𝑖,𝑡 [Equation 4] !&# %&'(&)%'*0,1 !&# )./()0,1 = 𝛽 𝐷𝑔𝑝𝑖,𝑡−1+ 𝛽 𝑆𝐼𝑍𝐸𝑖,𝑡−1+ 𝛽 𝑅𝑂𝐴𝑖,𝑡−1 + 𝛽 𝐿𝐸𝑉𝑖,𝑡−1+ α𝑖+ 𝜆𝑡+ 𝑢𝑖,𝑡 [Equation 5] Where,

𝑅&𝐷 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦%,' or 𝑅&𝐷 𝑠𝑎𝑙𝑒𝑠%,' are the dependent variable, where i = entity and t = time.

𝐷𝑠𝑝%,'89 1 when both staggerd board and poison pill provisions are present_{0 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒}

𝐷𝑔𝑝%,'89 1 when both golden parachute and poison pill provisions are present_{0 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒}

4.3.2 Concerns of endogeneity

One of the major issues in empirical corporate finance studies is the concern of endogeneity, that is, broadly defined as a correlation between the explanatory variables and the error term in a regression (Roberts and Whited, 2013). Through an econometric lens, endogeneity is an issue in a panel dataset when there is a violation of one or more out of three assumptions. Namely, the sample has no omitted variables bias, the sample has no measurement error or the sample has no reverse causality. By adopting the fixed effects model this paper controls for all time-invariant differences between the firms, so the estimated coefficients of the fixed-effects models cannot be biased because of omitted time-invariant characteristics. Therefore, the fixed effect model approach offers a remedy to the endogeneity concern of omitted variable bias.

Furthermore, considering the potential measurement error, this paper is confident that the measurements used are comprehensively described and concisely applied. Addressing another common panel data concern regarding the simultaneous occurrences of changes of between variables and possible biases introduced by reverse causality, this paper follows a similar approach as Phillips and Zhdanov (2012), and scales the dependent variable R&D

Variable Model 1 Model 4 Model 5

0.00140** (0.000696) 0.00181 (0.00277) 0.000854 (0.00216) -0.0242*** -0.0239*** -0.0239*** (0.00790) (0.00792) (0.00786) -0.0142** -0.0141** -0.0140** (0.00558) (0.00561) (0.00562) 0.00563* 0.00576* 0.00579* (0.00309) (0.00316) (0.00314) 2009 -0.00628***-0.00655*** -0.00654*** (0.00215) (0.00220) (0.00213) 2010 -0.00823***-0.00785*** -0.00803*** (0.00226) (0.00226) (0.00225) 2011 -0.00607***-0.00564*** -0.00584*** (0.00201) (0.00205) (0.00197) 2012 -0.00443** -0.00394* -0.00416** (0.00202) (0.00207) (0.00196) 2013 -0.00349* -0.00308 -0.00331* (0.00205) (0.00206) (0.00197) 2014 -0.00255 -0.00228 -0.00252 (0.00214) (0.00217) (0.00209) 2015 -0.00283 -0.00261 -0.00285 (0.00217) (0.00221) (0.00213) 2016 -0.00288 -0.00275 -0.00300 (0.00224) (0.00229) (0.00222) Constant 0.132*** 0.135*** 0.136*** (0.0267) (0.0272) (0.0267) Observations 3,530 3,530 3,530 Number of firms 627 627 627 R-squared 0.040 0.039 0.039 Adj. R-squared 0.0371 0.0355 0.0352

Newey-West Standard robust standard errors in parentheses: *** p<0.01, ** p<0.05, * p<0.1

Managerial entrenchment

RDI

Firm size

Table 4: Regression results further analysis and robustness check

Firm leverage Return on assets (ROA)

intensity and R&D sales ratio at time t whilst all independent and control variables are scaled at time t−1.

5.

Conclusion and discussion

Corporate governance mechanisms exist to safeguard the responsible governance of firms, by both management and shareholders. One aspect of corporate governance is to protect against short term opportunism that may be harmful to a firm on the long-term. Firms and shareholders employ governance provisions to obtain this goal, and in this study are observed to be very widely used. Given the high prevalence of governance provisions, research into their effect on firm’s performance is highly relevant. Shareholder value is a common proxy for firm performance, however it represents a high level of abstraction, providing little insight into the underlying performance drivers of the firm.

Therefore, this study looks at a crucial pillar for long-term firm performance, that is R&D (Chakraborty et al., 2014). Firms that do not invest in R&D tend to lose their competitive advantage over time (Atanassov, 2013; Chemmanur and Tian, 2018). Moreover, the present paper reports a significant portion (± 5 per cent) of firm value to be allocated to R&D expenditures. Notwithstanding the relative size and importance of R&D investments, it is widely argued that the separation of ownership and control may induce both shareholders and managers to shirk away from investments in R&D (Holmstrom, 1989; Meulbroek et al., 1990). This is caused by the nature of R&D investments, since it is a present-day effort with uncertain results in the future. Therefore, R&D investments are likely target to be replaced by short-term opportunistic activities. Governance provisions are believed to serve as a tool to counteract this opportunism. This study endeavours to investigate if this perceived effect can be observed in a large sample of U.S. firms.

Firms in the modern economy are highly internationalized as illustrated by the high percentage of foreign income in this study (± 49 per cent). In contrast with past era’s, modern firms have globalized to such a degree that the isolated domestic perspective in finance research is no longer suitable. The factor of internationalization is therefore of fundamental importance when studying the effect of governance provisions on R&D investments. This led to the main question of this study: What effect do governance provisions have on investments in innovation, and is

Two theoretical strands of literature were identified through which the relationship between entrenchment and R&D investment may be viewed. Regression results suggest that the effect of entrenchment on R&D intensity is significant and positive. However, the size-effect is so small (one per mille), that the effect does not appear to be practically relevant. Thus, the absence of a concrete effect of governance provisions on R&D intensity calls into question the relevance of this specific corporate governance mechanism framework in the U.S. context. In addition, this paper finds no significant moderating interaction effect of internationalization on the entrenchment-R&D relationship. Therefore, this paper’s findings suggest that proposed effects of internationalization do not modify the relationship between entrenchment and R&D for U.S. firms.

between 2007 and 2010 a steep increase of the average adoption of governance provisions could be observed succeeded by a post-crisis gradual decrease and return to pre-crisis level between 2009 and 2016. The E-score heterogeneity across years is depicted in appendix VII. Thus, the non-reproducibility of prior results may be explained by post-crisis changes to corporate governance mechanisms employed by firms. The results from this paper suggest that succeeding studies could contribute by exploring qualitatively how different corporate governance mechanisms are interrelated, and how they develop over time. It could further the comprehension of corporate governance to quantify these complementary governance mechanisms and compare the prior-crisis framework with the post-crisis framework.

Another difference between this study and most previous research is that it examines R&D intensity instead of R&D performance measures. The strength of applying R&D intensity as a proxy for innovation is that it reflects managerial intention to innovate. Compared to outcome measures (e.g. number of patents and number of patents citations) which are dependent on success of the investments, it is a more appropriate measure for managerial intentions. This is suitable when studying the nexus between managerial incentives and R&D intensity (Hoskisson et al., 1993; Lhuillery, 2011; Vithessonthi and Racela, 2016), but a downside may be that the difference in approach here could obstruct the generalizability and comparison of findings based on outcome measures (Atanassov, 2013; Becker-Blease, 2011; Chakraborty et al., 2014) to R&D intensity based inferences.

dynamics, geographic locations differences or other firm specific characteristics. Therefore, it might be the case that firms with characteristics that are associated with higher corporate disclosure are overrepresented in the final sample. Third, this paper employs a fixed effect model and controls for unanticipated time effects. The upside of this approach is that the fixed effect model is superior over the random effect model for explaining inferences within the sample. However, by adopting the fixed effect, this paper has limited generalization power for the results beyond the sample. Notably, the result of the Hausman-test (see appendix III.1) are only borderline in favor of the fixed effects model. Thus, generalizability of this paper’s conclusions must be considered with caution.

All in all, this study’s findings cast doubt on the existence of a relationship between governance provisions and R&D intensity for U.S. S&P 1500 firms. It is conceivable that the two diametrically opposing effects as proposed in the entrenchment view and myopia view coexist and exert equally weighted effects, cancelling each other out. This might explain why the effects of the entrenchment on R&D are close to zero in this study.

This paper could be a stepping stone towards a better theoretical and practical understanding of how corporate governance practices influence decision-making of innovation for internationalized firms. To the author’s best knowledge, however, no data is currently available to researchers for non-U.S. firms on the Bebchuk et al., (2008) inspired E-index. Therefore, it is still largely unknown how governance provisions affect corporate governance and innovation outside the U.S domicile. It would be fascinating to gather this data and compare the effects governance provisions around the globe.

References

Aggarwal, Reena, Isil Erel, Miguel Ferreira, and Pedro Matos, 2011, Does governance travel around the world? Evidence from institutional investors, Journal of Financial Economics 100, 154-181.

Aghion, Philippe, John Van Reenen, and Luigi Zingales, 2013, Innovation and institutional ownership, American economic review 103, 277-304.

Aguilera, Ruth V., and Gregory Jackson, 2003, The cross-national diversity of corporate governance: Dimensions and determinants, Academy of management Review 28, 447-465.

Atanassov, Julian, 2013, Do hostile takeovers stifle innovation? Evidence from antitakeover legislation and corporate patenting, The Journal of Finance 68, 1097-1131.

Bae, Sung C., and Seungwook Noh, 2001, Multinational corporations versus domestic corporations: a comparative study of R&D investment activities, Journal of

Multinational Financial Management 11, 89-104.

Bebchuk, Lucian A., and Alma Cohen, 2005, The costs of entrenched boards, Journal of

Financial Economics 78, 409-433.

Bebchuk, Lucian, Alma Cohen, and Allen Ferrell, 2008, What matters in corporate governance? The Review of financial studies 22, 783-827.

Bebchuk, Lucian A., Alma Cohen, and Charles C. Wang, 2013, Learning and the disappearing association between governance and returns, Journal of Financial

Economics 108, 323-348.

Becht, Marco, and Ailsa Röell, 1999, Blockholdings in Europe: An international comparison, European Economic Review 43, 1049-1056.

Becht, Marco, Patrick Bolton, and Ailsa Röell, 2003, Corporate governance and control, Handbook of the Economics of Finance 1, 1-109.

Becker-Blease, John R., 2011, Governance and innovation, Journal of Corporate Finance 17, 947-958.

Bhagat, Sanjai, and Ivo Welch, 1995, Corporate research & development investments international comparisons, Journal of Accounting and Economics 19, 443-470. Burgman, Todd A., 1996, An empirical examination of multinational corporate capital

structure, Journal of International Business Studies 27, 553-570.

Chakraborty, Atreya, Zaur Rzakhanov, and Shahbaz Sheikh, 2014, Antitakeover provisions, managerial entrenchment and firm innovation, Journal of economics and business 72, 30-43.

Chemmanur, Thomas J., and Xuan Tian, 2018, Do antitakeover provisions spur corporate innovation? A regression discontinuity analysis, Journal of Financial and Quantitative

Analysis, 1-32.

Danielson, Morris G., and Jonathan M. Karpoff, 2006, Do pills poison operating performance? Journal of Corporate Finance 12, 536-559.

Dechow, Patricia M., and Richard G. Sloan, 1991, Executive incentives and the horizon problem: An empirical investigation, Journal of Accounting and Economics 14, 51-89. Denis, Diane K., and John J. McConnell, 2003, International corporate governance, Journal

of financial and quantitative analysis 38, 1-36.

Eberhart, Allan C., William F. Maxwell, and Akhtar R. Siddique, 2004, An examination of long-term abnormal stock returns and operating performance following R&D

increases, The Journal of Finance 59, 623-650.

Gibbons, Robert, and Kevin J. Murphy, 1992, Optimal incentive contracts in the presence of career concerns: Theory and evidence, Journal of political Economy 100, 468-505. Gompers, Paul, Joy Ishii, and Andrew Metrick, 2003, Corporate governance and equity

prices, The quarterly journal of economics 118, 107-156.

Hall, Bronwyn H., 2002, The financing of research and development, Oxford review of

economic policy 18, 35-51.

Hall, Bronwyn H., Adam B. Jaffe, and Manuel Trajtenberg, 2001, The NBER patent citation file: lessons, insights, and methodological tools, Patents, citations and innovations.

Hitt, Michael A., Robert E. Hoskisson, Richard A. Johnson, and Douglas D. Moesel, 1996, The market for corporate control and firm innovation, Academy of management

journal 39, 1084-1119.

Holmstrom, Bengt, 1989, Agency costs and innovation, Journal of Economic Behavior &

Organization 12, 305-327.

Hoskisson, Robert E., Michael A. Hitt, and Charles W. Hill, 1993, Managerial incentives and investment in R&D in large multiproduct firms, Organization Science 4, 325-341.

Humphery-Jenner, Mark L., and Ronan G. Powell, 2011, Firm size, takeover profitability, and the effectiveness of the market for corporate control: Does the absence of anti-takeover provisions make a difference? Journal of Corporate Finance 17, 418-437.

Humphery-Jenner, Mark, 2014, Takeover defenses, innovation, and value creation: Evidence from acquisition decisions, Strategic Management Journal 35, 668-690.

Jensen, Michael C., and William H. Meckling, 1976, Theory of the firm: Managerial

behavior, agency costs and ownership structure, Journal of Financial Economics 3, 305-360.

Jensen, Michael C., and Richard S. Ruback, 1983, The market for corporate control: The scientific evidence, Journal of Financial Economics 11, 5-50.

Kabir, Rezaul, Dolph Cantrijn, and Andreas Jeunink, 1997, Takeover defenses, ownership structure and stock returns in the Netherlands: An empirical analysis, Strategic

Management Journal , 97-109.

Kotabe, Masaaki, Srini S. Srinivasan, and Preet S. Aulakh, 2002, Multinationality and firm performance: The moderating role of R&D and marketing capabilities, Journal of

International Business Studies 33, 79-97.

Lerner, Josh, Morten Sorensen, and Per Strömberg, 2011, Private equity and long-run investment: The case of innovation, The Journal of Finance 66, 445-477.

Lhuillery, Stephane, 2011, The impact of corporate governance practices on R&D efforts: a look at shareholders’ rights, cross-listing, and control pyramid, Industrial and Corporate

Change 20, 1475-1513.

Mahoney, James M., Chamu Sundaramurthy, and Joseph T. Mahoney, 1997, The effects of corporate antitakeover provisions on long-term investment: empirical

evidence, Managerial and Decision Economics 18, 349-365.

Manso, Gustavo, 2011, Motivating innovation, The Journal of Finance 66, 1823-1860. Meulbroek, Lisa K., Mark L. Mitchell, J. H. Mulherin, Jeffry M. Netter, and Annette B.

Poulsen, 1990, Shark repellents and managerial myopia: An empirical test, Journal of

Political Economy 98, 1108-1117.

Nickell, Stephen, Daphne Nicolitsas, and Neil Dryden, 1997, What makes firms perform well? European Economic Review 41, 783-796.

O'Sullivan, Mary, 2001, Contests for corporate control: Corporate governance and economic performance in the United States and Germany, Oxford University Press, number

9780199244867.

Porta, Rafael L., Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny, 1998, Law and finance, Journal of political economy 106, 1113-1155.

Prowse, Stephen D., 1994, Corporate governance in an international perspective: a survey of corporate control mechanisms among large firms in the United States, the United Kingdom, Japan and Germany, Bank for International Settlements

Roberts, Michael R., and Toni M. Whited, 2013, Endogeneity in empirical corporate finance, Vol. 2. Elsevier, 2013. 493-572.

Seru, Amit, 2014, Firm boundaries matter: Evidence from conglomerates and R&D activity, Journal of Financial Economics 111, 381-405.

Shleifer, Andrei, and Robert W. Vishny, 1989, Management entrenchment: The case of manager-specific investments, Journal of Financial Economics 25, 123-139. Stein, Jeremy C., 1988, Takeover threats and managerial myopia, Journal of Political

Economy 96, 61-80.

Sullivan, Daniel, 1994, Measuring the degree of internationalization of a firm, Journal of

International Business Studies 25, 325-342.

Tallman, Stephen, and Jiatao Li, 1996, Effects of international diversity and product diversity on the performance of multinational firms, Academy of Management journal 39, 179-196.

Vithessonthi, Chaiporn, and Olimpia C. Racela, 2016, Short-and long-run effects of

internationalization and R&D intensity on firm performance, Journal of Multinational

Appendices

Appendix I: Variables and definitions.

Variable Definition

Dependent variables

R&D intensity RDI R&D expenditures / total book value of the assets R&D sales ratio RDS R&D expenditures / total sales

Independent variable

Managerial entrenchment ENTR E-Score measured scaling from 1 to 6, lagged by 1 year Dummy variable for measuring presence combination staggerd board and poison pill DSP Dummy variable lagged by 1 year

Dummy variable for measuring presence combination golden parachute and poison pill DGP Dummy variable lagged by 1 year

Firm internationalization RDI INTERrdi Internationalization measured by Foreign income ratio:foreign income / total income, lagged by 1 year Firm internationalization RDS INTERrds Internationalization measured by Foreign sales ratio:foreign sales / total sales lagged by 1 year Moderating variable

Degree of Internationalization (DOI) ENTR * INTER Control variables

Firm size SIZE Natural logarithm of total assets, lagged by 1 year Firm leverage LEV Book value of total debt / total assets, lagged by 1 year Return on assets (ROA) ROA Operating income / total assets, lagged by 1 year

E-Score measured scaling from 1 to 6, lagged by 1 year multiplied by Foreign sales ratio measured by foreign income / total income, lagged by

!1 $ℎ&' ()*ℎ +*,--&./ (),./ ,'/ 0)1+)' 0122 ,.& 0.&+&'* 0 )*ℎ&.$1+&

!1 $ℎ&' ()*ℎ -)2/&' 0,.,4ℎ5*& ,'/ 0)1+)' 0122 ,.& 0.&+&'* 0 )*ℎ&.$1+&

Appendix II: ISS Governance Database definitions of the six Provisions within Bebchuck et al., (2008) E index

Variable Definition

Staggered board

Limitation on amending bylaws Limitation on amending the charter Supermajority to approve a merger Golden parachute

Poison pill

a board in which directors are divided into separate classes (typically three) with each class being elected to overlapping terms.

a provision limiting shareholders' ability through majority vote to amend the corporate bylaws. a provision limiting shareholders' ability through majority vote to amend the corporate charter. a requirement that requires more than a majority of shareholders to approve a merger (=> 66.66 per cent) a severance agreement that provides benefits to management/board members in the event of firing, demotion, or resignation following a change in control

Appendix III.1 Comparing different linear regression models Dependent variable: RDI

Variable Managerial entrenchment -0.00347*** 0.00103 0.00134* (0.000907) (0.000695) (0.000718) Internationalization -0.000223 -0.000569 -0.000594 (0.00238) (0.000978) (0.000972) Entrenchment * Internationalization 0.000144 0.000205 0.000216 (0.000785) (0.000312) (0.000310) Firm size -0.0209*** -0.0188*** -0.0242*** (0.00125) (0.00634) (0.00791) Firm leverage -0.0596*** -0.0105* -0.0145*** (0.0132) (0.00537) (0.00559) Return on assets (ROA) -0.00308 0.00485 0.00554* (0.00534) (0.00305) (0.00308) Constant 0.138*** 0.111*** 0.132*** (0.00608) (0.0214) (0.0267) Observations 3550 3550 3550 R-squared 0.081 0.024 0.041 Number of firms 2323 627 627

Firm FE YES YES

Year FE YES

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Firm and time fixed OLS With dummies Firm fixed effects

Appendix III.2 Breusch and Pagan Lagrangian multiplier test for random effects

Estimated results:

Variance Std dev. H0: Var(u) = 0 RDI 0.0032522 0.0570279 H1: Var(u) ≠ 0

Appendix III.3 Hausman fixed random effect test b-B difference Managerial entrenchment 0.0010252 0.0008882 0.0001371 0.0001254 Internationalization -0.0005712 -0.0005171 -0.0000541 0.0000482 Entrenchment * Internationalization 0.0002053 0.0001892 0.0000162 0.0000136 Firm size -0.0191632 -0.0206265 0.0014633 0.0016887 Firm leverage -0.0101922 -0.0115436 0.0013514 0.0004872 Return on assets (ROA) 0.0049941 0.004739 0.0002551 0.0003444 H0: Difference in coefficients not systematic

H1: Difference in coefficients is systematic

Chi^2(6) = = 12.81 Prob>Chi^2 = 0.0462 Coefficients FE estimator (b) RE estimator (B ) 2009 = 0 2010 = 0 2011 = 0 2012 = 0 2013 = 0 2014 = 0 2015 = 0 2016 = 0 F( 8, 626) = 2.34 Prob > F = 0.0174 Appendix III.4: Chow-test For Time fixed-effects

H0: σ(i)^2 = σ^2 for all i H1: σ(i)^2 ≠ σ^2 for all i

Chi^2 (632) = 1.6e+35 Prob>Chi^2 = 0.000

Appendix IV.1: Modified Wald test for

H0 : no first order autocorrelation H1 : first order autocorrelaion

F( 1, 490) = 526.126 Prob > F = 0.0000

Appendix IV.2: Wooldrige test for

autocorrelation in panel data

Appendix V: Specification of rewritten time and firm fixed regression models utilized for further

estimation

𝑅&𝐷 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦%,'= 𝛼 + 𝛽9𝑋%,'89+ ⋯ + 𝛽m𝑋%,'89+ 𝛾o∗ 𝐷o,%+ ⋯ + 𝛾&∗ 𝐷&,%+ 𝛿o∗ 𝐵o,'+ ⋯ +

𝛿&∗ 𝐵r,'+ 𝑢%,'

𝑅&𝐷 𝑠𝑎𝑙𝑒𝑠%,' = 𝛼 + 𝛽9𝑋%,'89+ ⋯ + 𝛽m𝑋%,'89+ 𝛾o∗ 𝐷o,%+ ⋯ + 𝛾&∗ 𝐷&,%+ 𝛿o∗ 𝐵o,'+ ⋯ + 𝛿&∗

𝐵r,'+ 𝑢%,'

With n − 1 dummy variables and T − 1 binary variables whereas,

D_o,t u1 when i = 2

0 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 ,..., Dw,t u1 when i = n_{0 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒} & Bo,' u1 𝑤ℎ𝑒𝑛 𝑡 = 2_{0 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒} ,..., Br,' u1 𝑤ℎ𝑒𝑛 𝑡 = 𝑇_{0 𝑂𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒}

Where: 𝑅&𝐷 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦%,' or 𝑅&𝐷 𝑠𝑎𝑙𝑒𝑠%,' is the dependent variable where i = entity and t = time.

𝛽 𝑋_%,'89 represents independent variables, up to 𝛽 𝑋𝑘_%,'89 𝛼 represents the constant

𝛾o ,…, 𝛾& represents the coefficient for the binary entity regressors

do ,…, do represents is the coefficient for the binary time regressors

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Classified Board Limit Ability to Amend ByLaws

Limit Ability to Amend Charter

Supermajority - mergers in percent

Golden Parachutes Poison Pill Dummy Stag And Poison Dummy Golden And Poison Pe rc en ta ge o f f ir m s t ha t a do pt ed a n E -in de x go ve rn an ce p ro vis io n

Appendix VI: Presence of E-index governance provisions for sample firms shown in percentages

3.3 3.4 3.5 3.6 3.7 Me a n E-Sco re s 2008 2010 2012 2014 2016 Year