Master Thesis
Board Political Orientation and Organizational Innovativeness.
by
PATRICK DUDINK
S2539845
Rijksuniversiteit Groningen
Msc Business Administration – track: Change Management
Words: 10.068
Semester 2 2018-2019
Supervisor: dr. B.C. Mitzinneck
Co-assessor: dr. I. Maris-de Bresser
ABSTRACT:
Political orientation of corporate decision makers has emerged as an increasingly relevant topic within management. This paper proposes a mediated effect of board room political ideology, specifically conservatism, on quantity of innovation, as well as different types of innovation. I theorize that board room conservatism invokes risk avoidance and short-termism, which in turn results in more
conservative R&D policies that dampen innovation. I find support in a sample of S&P500 firms for an effect of conservatism on R&D policy and organizational innovativeness. Results show that board room conservatism is correlated with lower levels of organizational innovativeness and lower R&D spending. Conversely, liberal boards are expected to have more R&D spending and thus higher innovativeness. No support, however was found for the proposed mediation effect as R&D spending did not result in significant changes in innovation. No support was found for the influence of political ideology on types of innovation either.
Introduction
Innovativeness serves for many companies as an important weapon to compete and, according to some, even to survive (eg. Cefis & Marsili, 2006; Ortiz-Villajos & Sotoca, 2018). For many years innovativeness is well-sought after by many practitioners and theorists alike (Ruvia, Shoham, Vigoda-Gadot & Schwabsky, 2014). More recently, political orientation of corporate decision makers has emerged as an increasingly relevant topic within management (Bermiss & McDonald, 2018). Corporate political ideology has been shown to influence corporate policies and R&D spending (Hutton, Jiang & Kiamar, 2014). The choice of lower levels of R&D due to financial conservatism is expected to dampen innovation (Hutton, Jiang & Kiamar, 2014). However, so far this relation remains empirically untested. Therefore, this paper addresses the question: What is the effect of board room political composition on organizational innovativeness?
Upper Echelon Theory (UET) states that organizational outcomes are partially predicted by the background characteristics of the top ranks within an organization (Hambrick & Mason, 1984). Meta-analysis shows that CEO characteristics indeed reliably predict future firm performance (Wang, Holmes, Oh & Zhu, 2016). Big amounts of UET research has mainly focused on the more superficial demographic upper echelon characteristics (Wang, Holmes, Oh & Zhu, 2016). Only recently, attention is spent on the deeper underlying values that determine the behavior of corporate decision makers.
In this trend, still rooted in UET, a recent stream of literature examines the relation between political ideology in board rooms and organizational outcomes (eg. Chin, Hambrick, and Trevino, 2013; Christensen et al., 2015). This stream builds on the Upper Echelon Theory by introducing political ideology as an Echelon factor: a characteristic that explains organizational outcomes. I will join this stream of research by using political ideology to help better understand and predict
lower levels of R&D (Hutton, Jiang & Kiamar, 2014), which in turn decreases innovative output (Kor, 2006).
This paper adds to UET by giving insights into the effect of personal values of board members on corporate outcomes. It explores the novel and promising link between board room political
ideology and organizational innovativeness. Previous UET research has explored the effects on innovativeness (eg. Bantel & Jackson, 1989; Sperber & Linder, 2016). Those researches however, focused on individuals or demographic variables and disregarded the values of the board as a whole. My exploration of values on a board level will help UET literature as both groups (Hambrick, 2007), and values (Hambrick & Mason, 1984) are expected to have better predictive power on organizational outcomes than individual demographics.
This paper also adds to the literature on organizational innovativeness. The combination of politics and innovation (or values and innovation) is novel and promising. This new information can work as a new step towards understanding the mechanisms accompanied with organizational innovativeness. Practitioners in particular might benefit from this research, as insights in the antecedents of innovation can help them achieve innovativeness and competitive advantage.
The present study continues with an elaboration on the current literature, which will result in the proposed hypotheses. In the methods section the procedures, data and variables are explained. Then, the hypotheses will be tested and the results will be discussed. The paper will close with a discussion of the results as well as an account on the implications and limitations of the study.
Literature Review
Upper Echelon Theory and Political Ideology
According to UET, organizational outcomes are viewed as reflections of the values and cognitive bases of powerful actors in the organization (Hambrick & Mason, 1984). Bounded rationality and strategic decision making have been presented as the explaining factors in how CEO characteristics might determine the organizational outcomes. (Hambrick & Mason, 1984)
individual behavior and organizational outcomes. This is also reflected in the high amount of citations in modern articles (Yamak, Nielsen, Escriba-Esteve, 2014).
CEO characteristics (such as extraversion, age and charisma) and CEO experience (education, and tenure) all have positive significant effects on firm performance. An important note to make, is that the effect sizes found are mostly relatively small (Wang, Holmes, Oh & Zhu, 2016). This critique is countered by the argument that although the effect sizes of the relationships might be small in magnitude, they are significant in practice (Wang, Holmes, Oh & Zhu, 2016). As Hambrick (2007) explains, it is important to focus on what part of the variance you can explain, and explaining 5% of profitability can be the difference between success and failure.
Variance in outcomes can stem from the level of analysis, such as using either highly ranked individuals or the board of directors as a whole. A combination of important corporate decision makers, such as the board, is considered to be better in explaining organizational outcomes than just the CEO (Hambrick, 2007). The board is assumed to have larger effect on the outcomes than
individuals since a collective can better influence the strategic decision making (Hambrick & Mason, 1984). For example a study by Bantel and Jackson (1989) found that the organizational outcome of innovation was associated with the demographic characteristics of the top management team. Additionaly, in a supplemental analysis they found that team characteristics were more strongly related to innovation than were CEO characteristics. As this paper focuses on the same outcome variable: Organizational innovativeness, I choose to focus on board characteristics over individual differences as otherwise important predictive power is disregarded.
moreover, earlier treatises on executive leadership prominently proposed that values enter into managerial actions (Chin, Hambrick & Trevino, 2013). Many academics seem to agree that the subjective values, beliefs and attitudes are more predictive of future behavior than are demographic variables (eg. Theory of Planned Behavior (Ajzen, 1985) & Behavioral theory of the firm (Cyert & March, 1963).).
Political ideology can be considered a deep lying value. Whereas regular values operate at a more basic psychological level, individuals use political ideology to organize and structure their values and beliefs (Jost, 2006). As values determine organizational outcomes and political ideology is an important value, it will be interesting to explore the possible effects of political ideology on organizational outcomes. Before I do so, it is important to have a clear definition of what political ideology entails.
Political ideology can be defined as: a set of beliefs about the proper order of society and how it can be achieved (Erikson & Tedin, 2003). Most researchers assume that ideology is represented in memory as a kind of schema, i.e., a learned knowledge structure consisting of an interrelated network of beliefs, opinions, and values” (Jost, Federico & Napier, 2009). Ideologies are overarching
sentiments that structure opinions. Specific ideologies crystallize and communicate the widely (but not unanimously) shared beliefs, opinions, and values of an identifiable group, class, constituency, or society (Knight, 2006).
In classifying political ideologies, the liberal-conservative distinction has been the single most useful and parsimonious way to classify political attitudes for more than 200 years (Jost 2006:654). In the United States and elsewhere, it is becoming increasingly common to substitute liberal and
conservative for left and right (Jost, Federico & Napier, 2009). Both concepts are associated with different preferences and values. In a very broad sense, the continuum shows the distinction
solidarity, protest, opposition, radical, socialism and communism(Fuchs & Klingemann 1990, pp. 213–214).
In the United States party system, conservatives are considered to most likely be Republicans and the liberalists to be Democrats (Hutton, Jiang & Kiamar, 2014). The two major political parties have taken opposing stands on economic, cultural and moral issues (Layman, Carsey & Horowitz, 2006). With on the left side the Democrats favoring more wealth redistribution through the
government and less government influence in promoting traditional morality and social order. And on the right side of the spectrum the Republicans favoring less wealth redistribution through the
government and greater government influence in promoting traditional morality and social order. Over the last years, both parties are polarizing and the disagreement among them seems to grow (Layman, Carsey & Horowitz, 2006). The Republicans are becoming more conservative, and the Democrats are becoming more liberal (Layman, Carsey & Horowitz, 2006).
The association between political ideology and organizational outcomes has recently come to the attention of a growing group of authors and has found promising results. Some examples of promising directions in which board room political ideology has been explored are: tax avoidance (Christensen et al 2015), CEO pay (Gupta & Wowak, 2017), corporate financial policies (Hutton, Jiang & Kumar, 2014) and CSR (Chin, Hembrick & Trevino, 2013).
These papers share their focus on donations by board directors of S&P 500 firms. The body of literature is still limited and the proposed mechanisms are either unclear or untested. Board room ideology is promising, but requires more exploration. Also, the authors strongly differ in their explanations of how individual ideologies result in organizational behaviors. This paper combines these explanations into an own, and follows up with an exploration of the influence of board room political ideology in the specific field of innovation.
ideologically heterogeneous. The finding that within the same board, individual directors have donated to candidates with opposed policy views suggests that personal preferences, rather than the interests of their firms, largely dictate which candidates they support (Bonica, 2015). This paper assumes that corporate boards reflect the underlying relatively stable political preferences of its members. Boards ideologies change as board configurations change over time. The in- and outflux of board members can make for fluctuations in aggregated board room political ideology.
I pose that boards act according to their ideologies. Political orientations on the board level serve as an indicator of underlying shared values (Jost, 2006; Knight, 2006). For example, one such value that is central here is the preference for executive conservatism (Christensen et al., 2015). In line with these preferences, boards will enact governance policies that align with their directors’ prevailing ideologies. Directors will inherently be drawn to value-congruent alternatives that, in their view, represent the most rational, value-maximizing choices (Gupta & Wowak, 2017). Directors will bring their preexisting ideologies to bear on the decisions they face in the boardroom, and their preferences will manifest in how they discuss, debate, and vote on key issues. Accordingly, board’s decisions will reflect the aggregate ideological composition of its members (Gupta & Wowak, 2017). Among the many not yet explored outcomes resulting from board room ideology is organizational innovation. In the next section I will first introduce the concept of innovation and then explain its relation with ideology.
Board Room Political Ideology and Innovation
There are many differing conceptualizations of innovativeness depending on the level of analysis used (Subramanian & Nilakanta, 1996). The concepts of creativity and innovation are often used interchangeably in literature (Martins & Terblanche, 2003). Organizational creativity is the creation of a valuable, useful, new product, service, idea, procedure, or process by individuals working together in a complex social system (Woodman, Sawyer & Griffin, 1993). In line with that,
organizational innovativeness can be seen as creativity that has been brought to practical use (West & Farr, 1989). Truly innovative organizations are considered those that exhibit innovative behavior consistently over time (Subramanian & Nilakanta, 1996). For this paper I refer to the
innovate continually over time as a five-dimensional construct of creativity, openness, future orientation, risk-taking, and proactiveness (Ruvia, Shoham, Vigoda-Gadot & Schwabsky, 2014).
In addition to age, tenure, educational level and functional background that have already been found to have effect on organizational innovativeness (Sperber & Linder, 2016), political ideology is expected to influence innovation. Firstly, because UET states that board room values can determine organizational outcomes (Hambrick & Mason, 1984) and political ideology is an important bearer of values (Jost, Federico & Napier, 2009). Secondly, as political ideology possibly influences R&D policies, which then result in an increase or decrease in organizational innovativeness. Below will follow a further account on the relation between ideology (specifically conservatism) and
innovativeness through R&D spending policy changes.
Board Room Political Ideology and Innovation through Research & Development
Boards have considerable influence over R&D spending policies (Osma, 2008). R&D budgets are long-run plans of the multiyear spending needed to acquire and develop firm technologies (Osma, 2018). Both relative and absolute R&D spending tends to differ between companies (Chrisman & Patel, 2012), yet not so much within companies over time (Souder & Bromiley, 2012). In particular, the allocation of resources in one year depends largely on the expenditures in the same category of the previous year. With only minor adjustments being made through a budgeting process. Which implies that, without any clear reason (such as a switch in CEO), R&D spending is relatively stable, unless due to obvious causes (Souder & Bromiley, 2012).
underlying tendencies to be conservative. Secondly, by the finding that Republican managers have lower levels of corporate debt, lower capital and R&D expenditures, and less risky investments (Hutton, Jiang & Kiamar, 2014).
These risk-avoiding tendencies are to be expected to be visible in R&D policies. As R&D investment is, for multiple reasons, particularly risky. Investments in R&D are often irreversible and accompanied by uncertainties in achieving technical goals, commercialization, and appropriation (Kor, 2006). Furthermore, R&D investments are inherently risky in nature as R&D does not directly result in gains (Chrisman & Patel, 2012). It has been estimated that only 12% to 20% of R&D projects result in products that get to the market; the rest are failures (Jones, 2009). As R&D investment is highly risky, and conservative boards are expected to be risk-averse, I pose that risk avoidance makes conservative boards engage in less R&D intensive policies.
In addition to the risk-avoiding tendencies of conservatives, also short-termism of firms plays an important role in explaining differences between R&D policies. Corporate investments vary in payoff horizon: the amount of time until the expected returns of an investment will exceed (Souder & Miles Shaver, 2010). When corporate decision makers hold an excess preference for returns sooner rather than later, that mechanism is called short-termism (Laverty, 1996). Previously, pay horizon has been used as a proxy to risk, where long-term investment is seen as high risk (eg. Palmer & Wiseman, 1999). But new insights show they are two separate measures, both explaining their own variance independently from one another (Souder & Bromiley, 2012). In other words: A long project can still be low risk and a short project can be high risk. In addition to risk-avoidance, also pay horizon of the investment can explain differences in investment policy. Which is in its place explained by political preference, as I will explain now.
performance during the period before the value of long horizon investments is realized (Narayanan, 1985).
Short-termism is particularly likely to be found among conservative boards. Research by Gupta and Wowak (2017) showed that conservative and liberal boards differ in how they reward and penalize their CEOs for prior performance. They found that conservative boards more tightly link CEO pay to recent financial performance. This is explained by their tendency towards person-based attributions. The authors suggest that this relationship is a direct consequence of the conservative preference for proportionality in rewards, which conservatives believe to be an inherently desirable and fair arrangement and therefore instrumental in retaining CEO talent. The higher degrees of short-termism among conservative boards, will likely decrease their R&D spending. R&D investments will likely occur less often, in smaller amounts or be avoided altogether in more short-term focused conservative boards as R&D generally take a considerable amount of time before it results in gains (Chrisman & Patel, 2012). Given that both risk-avoidance and short-termism affect conservative boards, I pose that board room conservatism decreases R&D spending. Which results in hypothesis 1a: Hypothesis 1a: The more conservative a board is, the smaller the amount of R&D spending.
This increase or decrease in R&D spending results in changes in the amount of innovative output. R&D activities, both external and internal, are widely recognized as being the drivers of technological advancements and the levels of growth of R&D expenditures are considered to be reliable indicators of innovative capacity (Martin & Nguyen-Thi, 2015). Research shows that R&D spending is strongly positively associated with the probability of introducing a new product and R&D can affect productivity growth by facilitating the absorption of new technologies (Parisi, Schiantarelli, & Sembenelli, 2006). Also, R&D and patents are positively related both across firms at a point in time and across time for given firms. (Klette & Kortum, 2002). Moreover, in technologically intensive industries, investments in R&D have been the primary source of product innovation and superior returns (Kor, 2006). All of the above implies that R&D spending is positively associated with organizational innovation, making for hypothesis 1b.
Conservatism results in less R&D spending, and less R&D means less innovation. In this paper, R&D is posed as the main mediator between conservatism and innovation. But this is only one mechanism, and as innovation and ideology are complex and multi-faceted concepts (Knight, 2006; Subramanian & Nilakanta, 1996), other mechanisms might exist too. Therefore this relation is only partially mediated by R&D spending as it is not the only mechanism. Individuals use political ideology to organize and structure their values and beliefs (Jost, 2006). These values and beliefs can concern a wide range of subjects. Conservative values can be reflected in many different ways, such as preferences of conservative boards towards diversity and assets (discussed in the methods), culture (discussion) or innovation altogether. What is central in this story is the following. As explained, the general distinction between ideologies is based on a preference for either change (liberal) or stability (conservative) (Jost, Federico & Napier, 2009). As most conservative directors probably value stability over change, they are less likely to advocate innovation, as it inherently means change. This means that conservatives will likely hold negative attitudes towards change and will therefore not support any initiatives towards innovativeness, be it through spending, policy, culture or any other form of change promotion. Resulting in the last sub- as well as the main hypothesis.
Hypothesis 1c: The more conservative a board is, the lower the organizational innovativeness. The combination of the previous three sub hypotheses makes for an expected mediation effect. With conservatism explaining differences in organizational innovativeness as mediated by the R&D policy. Therefore I pose.
Hypothesis 1: The negative relation between board conservatism and organizational innovativeness is partially mediated through organizational innovativeness.
Board Room Political Ideology and Types of Innovation
Not only the amount of spending and innovation will likely differ with the board's ideology, also the type of innovation pursued is influenced by board room ideology. The coming part discusses the effect of ideology on a specific dimension of organizational innovativeness: relatedness of innovation.
innovation concerns introducing totally new concepts (Kobarg, Wollersheim & Welpe, 2019; Xu & Jang, 2014). The terms incremental and convergent are used interchangeably, and in this context radical innovation is synonymous for divergent innovation. The two sides of continuum are captured in the concept of relatedness of innovation.
I pose that the relatedness of innovation will likely be influenced by ideology through some of the same mechanisms of how the amount innovation is affected by conservatism. In the same way that risk-aversion and pay horizon is expected to predict the quantity of innovation, these effects are likely equally, if not more, present in the type of innovation that is pursued. I will now explain how,
according to theory explained below, I expect that conservatism will result in more related innovation. In doing so I will first revisit risk-avoidance and payoff horizon and then introduce the concept of status quo bias.
As per definition, radical innovations are associated with higher amounts of risk (Kaluzny, Veney & Gentry, 1972). Divergent, radical or unrelated innovations are generally more risky than incremental innovation as it involves not yet explored areas and it assumes higher costs. Also in terms of payoff horizon divergent changes are expected to take longer to result in gains (Souder & Miles Shaver, 2010). Many opportunities have benefits that are deferred, because it takes time for an organization to become efficient in unfamiliar activities, incorporate them into existing routines, and cultivate a presence in new product markets (Singh, 1986). Given that innovation in unrelated areas is a particularly risky and late pay-off investment, it is expected for reasons mentioned previously, that conservative boards engage in more related (convergent) rather than unrelated (divergent) innovation. In addition to this, insights from status quo bias theory can be applied to explain why
conservative boards show different types of innovativeness than liberal boards. Status quo bias theory aims to explain people's preference for maintaining their current status or situation over change (Samuelson & Zeckhauser, 1988). Status quo bias is explained in terms of three main categories: rational decision making, cognitive misperceptions, and psychological commitment, with the latter two being the more subjective categories.
Within status quo bias theory, the main mechanism of cognitive misperception is loss
in that losses loom larger than gains in value perception (Kahneman & Tversky 1979). An individual's innate conservatism (Hirschheim & Newman 1988), combined with characteristics like rigidity, inertia, and attributional style contribute to the cognitive misperception of loss aversion (Kim & Kankanhalli, 2009). High degrees of loss aversion are to be expected in more conservative boards as they have per definition a more innate conservatism. Loss aversion can result in status quo bias
because even small losses of changing from the current situation could be perceived as larger than they actually are (Kim & Kankanhalli, 2009).
The conservatives preference for the status quo over change or revolution (Jost, Federico & Napier, 2009) inhibits the companies from experimenting due to loss aversion and status quo bias. This attitude shows in higher risk-avoidance in amount of innovation (h1), but also for more conservative types of R&D and innovation. An example of these effects can be found in aging and retiring CEOs. In retiring CEOs growing conservatism can be found. These CEOs tend to prefer incremental innovations as they are becoming more and more risk-averse over time and wish to remain things as they are (Barker & Mueller, 2002; Xu & Jang, 2014). Therefore I pose that conservative boards prefer (and thus pursue) innovation closer to the core business rather than more divergent innovations. This is reflected in hypothesis 2. All hypotheses are combined in figure 1.
Hypothesis 2: The more conservative a board is, the lower the relatedness among innovations.
Methods
Approach
The hypotheses are tested through quantitative analysis. The data is collected from secondary databases. For the political ideology data, computer aided matching procedures are performed to allow for comparison of ideology on board levels. The resulting board ideology scores were then matched to existing patent data and analyzed. The collection process is further described under the header
collection. Any incomplete or obviously skewed data has been eliminated. Where possible, mostly best practices in terms of operationalization and analysis have been derived from existing literature. The applied conceptualizations and a description of the analyses can be found in respectively the parts on measures and analysis.
Collection
This paper uses data from multiple sources. For each of the factors, where possible, annual data has been gathered. The sample consists of S&P 500 firms in the period of 2000-2006. For all variables, data has been found for each of the years and most of the firms. All variables used are continuous measures. Below follows an account on which variables have been collected from which sources.
Political ideology, indicated through personal donations, has been collected from the DIME database (Bonica, 2016). I utilized the DIME database for political ideology as it comprises a thorough body of information consisting of multiple millions of individual donations (Bonica, 2014). This information is converted to CF (Campaign Finance) scores, which have shown great internal and external validity in predicting political preferences (Bonica, 2018)1. DIME data has been mapped through fuzzy matching techniques. The computer-aided matching procedure coupled the donations to S&P 500 board directors (retrieved from BoardEx). To ensure the validity of the process manual checks of the matching procedures have been performed by two independent coders. Disagreements in the coding are discussed and resolved.
Innovation (patent) data is recovered from the NBER PDP data project2. Patents are used as a measure of innovativeness. The information is used from the years 2000 through 2006. Data from the PDP is used as it covers over 3 million patents, from which, through careful computer-aided matching, 1,1 million of these patents were identified to be owned by S&P 500 firms (Bessen, 2009). As this data is (1) dynamically assigned to companies by combining USPTO patent data with Compustat firm data (Bessen, 2009) and (2) thoroughly and professionally cleaned (Bessen, 2009), this dataset is chosen to be most appropriate for the current scope and timeframe of this research.
In order to make for correct matching and consistency throughout the dataset, mostly
Compustat data has been used to gather the data on R&D and the control variables. The only exception is the diversity measure, which is created by combining a variety of KLD3 measures.
Measures
Dependent variables
Organizational innovativeness. Innovativeness is measured on a continual scale measuring the amount of annual patents that are granted for each company. Patent data has often been used as a measure of technological innovation (Haščič, & Migotto, 2015), whereas a higher amount of annual patents is considered to be indicative of a higher degree of innovativeness. Patents have long been recognized as a very rich and potentially fruitful source of data for the study of innovation and technical change (Hall, Jaffe & Trajtenberg, 2001). Among the listed benefits of using patent are: highly detailed information, large number of patents, voluntary base of disclosure and presence of citations. A downside however, is that due to regulations or strategic choice not all innovations are patented (Hall, Jaffe & Trajtenberg, 2001).
Relatedness of innovations. Relatedness of innovations is proxied by the degree of self-citing behavior within the patents. Self-citations represent the extent to which companies attempt to exploit their previous inventions through renewed investments. They are an indication of the extent to which organizations follow up to previous invention (Trajtenberg, Henderson & Jaffe, 1997). Companies
2 No clear source reference available for the database as it represents the culmination of a long-term research and data-creation effort that involved a wide range of researchers (Hall, Jaffe & Trajtenberg, 2001). Database is accessed through https://sites.google.com/site/patentdataproject/Home
who show high levels of self-citation, focus more on their own (more related) inventions than the less related inventions of others. For this reason, self-citation has often been used as a proxy to measure exploitation intentions (eg. (Kim, Song, & Nerkar, 2012; Rosenkopf & Nerkar, 2001)).
Independent variables
Political ideology. Political ideology is measured along the continuum of conservative-liberal. Although this distinction is not all-encompassing, it has been the single most useful and parsimonious way to classify political attitudes for more than 200 years (Jost, 2006). Moreover, it has found
resonance in almost every cultural context in which it has been introduced (Jost, 2006). Therefore also in this research, Political ideology will be presented on a continuous scale in which -2 is liberal and +2 is conservative.
To identify board room ideology, I used political donations by board members to the two major parties in the U.S.: The Democrats and the Republicans as a proxy to board room ideology. More specifically, I use the CF-score (Campaign Finance score) devised by Bonica (2014). This individual CF score is aggregated on a board level. In doing so, this article is in line with prior research conducted in this stream of literature (eg. Christensen et al., 2015; Gupta & Wowak, 2017). In my sample, 62.99% of the board members have been identified. The remaining directors have been interpreted as missing values. Resulting in the situation where a not identified person does not
influence the overall board opinion.
Research in political science and sociology has shown that the differences between the Democratic Party and the Republican Party largely reflect differences in stable political ideologies with Democrats espousing more liberal beliefs and Republicans espousing more conservative beliefs (Knight, 2006; Gupta & Wowak, 2017). Therefore, the decision to donate to one versus the other party is strongly indicative of an individual’s personal ideological orientation (Hutton, Jiang & Kiamar, 2014; Bonica, 2018).
R&D spending. R&D spending behavior is analyzed by comparing two separate
spending, as the percentage captures the intention and policy towards R&D within a company rather than the size of it.
Control variables
Diversity. One mechanism linking innovation to ideology could be through diversity.
Schwartz (1996) argued that classical liberalism is concerned with civil rights and that people who are more liberal in political ideology are likely to be sensitive to social issues in general and to such specific issues as diversity, social change, human rights, and the environment. Liberal boards are expected to have hiring policies that result in more diversity than the policies of the conservative counterparts (Carnahan & Greenwood, 2018). Diversified hiring automatically results in more diversity within the firm, whether in terms of age, nationality, background or gender.
Within companies and teams, both racial and gender diversity have been found to induce innovativeness (Miller & Triana, 2009). Milliken and Martins (1996) propose that a management team composed of executives with diverse educational backgrounds is highly likely to embrace diverse mental templates to solve complex problems. Their study reveals that the diversification of top executives’ educational backgrounds can promote discussions within boards of different problem-solving tactics and contribute to endogenous managerial learning within management teams (Milliken & Martins, 1996). Such managerial learning may increase the likelihood of devising innovative methods such as establishing synergies among old and new operations (Chen, Kang & Butler, 2019). Resulting in more diverse boards (liberal boards) having more diversity, which leads to more
innovation. Similar research showed that upper echelon diversity in banking firms resulted in higher innovative output (Bantel & Jackson, 1989)
To control for the possible mediation effect of diversity, a combined measure of KLD diversity indicators is added to the analysis. The indicators highlight companies that have made notable progress in the promotion of women and minorities in important positions in the organization, extra points have been rewarded for companies with CEOs from minority groups and with progressive policies towards minority employees.
explained because industries can differ in their position in and length of their respective product life cycles. In for example the PC and Software industries, new products can get obsolete in a very short time, which requires extremely high levels of continuous innovation (Jones, 2009: 215). Whereas in other industries like steel or electricity, change in product technology is limited and more incremental since it is near or in the decline stage (eg. cassette recorders or typewriters), or the industry has a long base rate product life cycle (eg. raw materials industry) (Jones, 2009: 215). Moreover, average
political opinion seems to differ between industries. For example oil and manufacturing industries tend to be more conservative, but media and tech show more liberal leanings (Bonica, 2015). As industry type might explain differences in both ideology and innovation, it is important to take industry type into consideration for both. This is done by calculating the average annual patent rate for each of the SIC2 industries, and controlling for them in the analyses.
Industry political ideology. For the reasons mentioned above, for each SIC2 industry the average ideology has been computed and controlled for in the analysis.
Organization size. Existing theories strongly differ in their opinions on whether organization size positively or negatively affects innovativeness (Jaskyte, 2013). According to Jaskyte (2013) smaller organizations are more flexible than the bigger and more rigid organizations and are thus more flexible. While others predict that larger organizations have more financial and human means, and are thus more capable of being innovative. Meta-analysis by Lee & Xia (2006) shows the same mixed results. Researchers have found positive, negative, and non-significant effects. These mixed results are mainly due to differences in operationalizing organizational size and not taking into consideration the right control factors (Lee & Xia, 2006). This paper defines organization size as the relative size compared to the S&P500 average, as linear measures do not take into account the notion of competition and diminishing returns that are present in the innovation race (Gottinger, 2003). Analysis
three years on the patent counts and industry average patent counts relative to board ideology. For an overview of the time lags and time spans used in the analyses, see appendix C. Three years have been chosen as it generally takes considerable time for changes in R&D policy to fully take effect in the form of granted patents. R&D efforts show lagged results both in terms of short-term operational lag as well as long-term distant future knowledge culmination (Wang & Hagedoorn, 2014). Additionally, R&D spending and assets are lagged one year relative to ideology (two years ahead of patents). A lag of one year has been chosen, as R&D policy is most often determined in annual budgeting process (Osma, 2018). Ideally, a longer time lag would have been ideal. But, as this dataset is big in size, yet moderate in its timeframe (2000-2006), three years has been chosen as an optimal agreement between sample size and time lag.
An additional list of robustness tests have been performed to address alternative explanations and increase confidence in the empirical results. The first test performs longitudal analyses over the period of 2000-2003 in the form of FEM models to address issues concerning causality and third variable problems. The second round of tests checked different types of operationalizations to assure validity of the operationalization. Detailed information on the findings of the robustness procedures can be found in Appendix B.
Results
Descriptive statistics
20
spending. Also, R&D and innovativeness are positively related (all with p<0.01). Both the control variables of organization size and industry mean ideology are negatively correlated with ideology and R&D, but not with patent counts. Diversity seems to affect or be affected by ideology, R&D and innovation. More in-depth analysis will follow in 1. the analysis of the separate hypothesis (up next), 2. the discussion of the results and 3. The robustness checks (appendix B).
Test of Hypotheses
Table 2a shows an excerpt of the findings from the SEM procedure. This table will be used to test hypotheses 1a, 1b and 1c. The full results of the SEM procedure can be found in table 2b in appendix D.
Table 2a shows that the analysis is consistent with the argument that board room conservatism results in a lower degree of R&D spending (b = -2.899 ; z = -1.79). The model predicts that fully conservative boards (CF-score = 1) spend around 2,89% of their assets less on R&D than neutral boards (CF-score = 0). Given that the mean R&D spending in this sample is 6,8%, extremely
conservative companies are expected to do less than half the spending on R&D (3,9%) of what liberal boards do (9,7%). The effect is big, yet not very convincing as the model is significant only on the level of p < 0.1). Among the control variables only size and innovativeness are significant, indicating that bigger companies from more innovative industries invest relatively more in R&D. Overall, there
Table 2a. SEM of Ideology and R&D-spending on Innovation
n 439
r2 0.4453 Prob > F 0.00
Direct effects
Coef. Std. Err. z P>z
R&D spending (y+1)
Conservatism -2.899 1.616 -1.790 0.073 -6.066 0.267
Diversity 0.039 0.027 1.450 0.147 -0.014 0.092
Industry innovativeness (y+3) 0.050 0.006 7.930 0.000 0.038 0.063 Industry ideology -4.374 3.986 -1.100 0.273 -12.186 3.439
Size -0.944 0.397 -2.380 0.017 -1.723 -0.166
Innovativeness (y+3)
R&D spending (y+1) 0.020 0.015 1.310 0.190 -0.010 0.051 Conservatism -0.837 0.374 -2.240 0.025 -1.571 -0.103
Diversity 0.021 0.006 3.450 0.001 0.009 0.033
Industry innovativeness (y+3) 0.012 0.002 6.980 0.000 0.008 0.015 Industry ideology 1.786 1.023 1.750 0.081 -0.219 3.790 Size -0.009 0.044 -0.210 0.833 -0.095 0.076 Indirect effects Coef. Std. Err. z P>z Innovativeness (y+3) Conservatism -0.059 0.056 -1.050 0.295 -0.169 0.051 Diversity 0.001 0.001 0.970 0.333 -0.001 0.002
Industry innovativeness (y+3) 0.001 0.001 1.270 0.203 -0.001 0.003 Industry ideology -0.089 0.106 -0.840 0.402 -0.297 0.119
Size -0.019 0.018 -1.060 0.290 -0.055 0.016
Total effects
Coef. Std. Err. z P>z
R&D spending (y+1)
Conservatism -2.899 1.616 -1.790 0.073 -6.066 0.267 Diversity 0.039 0.027 1.450 0.147 -0.014 0.092 Industry innovativeness (y+3) 0.050 0.006 7.930 0.000 0.038 0.063 Industry ideology -4.374 3.986 -1.100 0.273 -12.186 3.439 Size -0.944 0.397 -2.380 0.017 -1.723 -0.166 Innovativeness (y+3)
R&D spending (y+1) 0.020 0.015 1.310 0.190 -0.010 0.051 Conservatism -0.896 0.371 -2.410 0.016 -1.624 -0.168 Diversity 0.022 0.006 3.590 0.000 0.010 0.034 Industry innovativeness (y+3) 0.013 0.001 8.600 0.000 0.010 0.016 Industry ideology 1.697 1.019 1.660 0.096 -0.301 3.694 Size -0.028 0.040 -0.710 0.478 -0.107 0.050
Notes: Proportion of total effect mediated 6.58%
[95% CI]
[95% CI]
is some support for hypotheses 1a, as a negative relation between board room conservatism and one year lagged R&D policy has been found.
Against expectations, table 2a did not find support for hypothesis 1b. Organizational R&D spending had a small and insignificant effect on the lagged patent counts (b = 0.020; z = 1.31). The findings do not give support for the hypothesis that relative R&D spending positively affects organizational innovative output. Hypothesis 1c, which predicted that an increase in board room conservatism is associated with lower organizational innovativeness, has been supported by the results in table 2a (b = -0.837; z = -2.24). Of the control variables, diversity, industry innovativeness and industry mean ideology managed to explained variance in innovativeness. With the former two being highly significant (p < 0.01), and the latter being almost significant (p = 0.081).
To test hypothesis 1, which predicted that the relation between the political orientation of the board and innovativeness will be positively mediated by the R&D spending policy, the full structural equation model in table 2b in appendix D is consulted. Consistent with the predictions, there seem to be signs of partial mediation. There is a small, not significant mediation effect (r2 = 0.45). The direct of ideology on innovation significant (h1c) (b = -0.837; z = -2.24), the indirect effect is not (b = -0.059; z = -1.05). The indirect effect weakens the direct effect, indicating that R&D spending explains a part (6,58 %) of the total effect of ideology on company innovativeness (b = -0.896; z = -2.41).
If we compare these results with the four steps of causal analysis (Baron & Kenny, 1986), we can see what obstructs the mediation. In short, for mediation to occur, three paths have to be
Figure 2 – Conceptual Model with indications for paths, effect sizes and significance.
Notes: * : p<0,10; ** : p<0,05; *** : p<0,01.
Lastly, Table 2c presents the results for hypothesis 2, which poses that higher board room conservatism is related with higher relatedness of innovation. The results do not support the hypothesis. Although a negative main effect has been found, it did not prove to be significant (b = -0.653, t = -0.31). Showing that ideology might have a small effect on the type of innovation pursued, but that this effect is not convincing enough to be considered significant. Also, the directionality of the effect indicates that conservative boards perform less self-citations, hence have less related
innovations. Moreover, none of the control factors seem to be significant in predicting relatedness either, indicating that there might be something wrong with the conceptualization of relatedness. More on this in the next section on robustness checks.
Robustness of Findings
Table 2c. Regression of Self-citations on Relatedness of innovations
n 159
r2 0.03000 Prob > F 0.4462
Relatedness (y + 3) Coef. Std. Err. t P>t
Conservatism -0.653 2.138 -0.310 0.760 -4.877 3.571
Diversity 0.016 0.035 0.460 0.649 -0.053 0.084
Industry innovativeness (y+3) 0.004 0.009 0.460 0.647 -0.014 0.022
Industry ideology 7.362 6.679 1.100 0.272 -5.833 20.558
Size 0.740 0.446 1.660 0.099 -0.141 1.621
constant 9.289 1.975 4.700 0.000 5.387 13.192
Table 3a and 3b show the results of the different robustness checks that are performed to double check the models(3a) and operationalizations (3b) chosen. Some notes and insights per manipulation are discussed below.
Robustness of Models. In addition to lagging the data to control for reverse causality
problems, I have chosen to do a robustness check in the form of longitudinal fixed effects analyses on the panel data. This way multiple years can be analyzed without having to deal with autocorrelation over years. And as fixed effects analysis only looks at within-subject variation, in case of significant findings, many third variable problems can be excluded, increasing the validity of the results.
Given that political ideology is a relatively stable phenomenon (Christensen et al., 2015) little longitudinal effects are to be expected and not finding any fixed effects will be unsurprising. This is however not the case. What I found is that in this longitudal sample, board ideology significantly predicted R&D policy the year after (b = 5.05; t = 2.08). The increase in significance from SEM to FEM is explained as firstly, the sample size in the second test is bigger and secondly, because apparently changes in ideology reliably predict R&D spending a year later. Unsurprisingly, the non-significant relation between R&D and innovativeness was not non-significant in the FEM either. Thus, finding no support for hypothesis 1b. The results in table 3a also show that the previously significant hypothesis 1c has not been confirmed by the FEM. Indicating that differences in ideologies between companies are related to innovativeness, but no such relations could be found within companies using a three year time lag of innovations measured at four different years. Lastly, no support was found for any relation between ideology and type of innovation. Therefore, this sample does not confirm hypothesis 2.
The relation in case of hypothesis 1a remains solid indicating that ideology and R&D are truly temporally related factors, for the other hypotheses, the case is more complicated. The lack of
Table 3a. Robustness checks for alternative models
Models\Path a>b (h1a) b>c (h1b) a>c (h1c) a>c’ (h1) n r2 a>d (h2) n r2 No modifications -2.899 -0.020 -.837 -0.589 439 0.445 -0.653 159 0.030
z / t & sig. -1.79 * 1.31 -2.24 ** -1.05 -0.31
Fixed Effects 5.048 -0.001 0.065 -0.029 828 0.078 -0.412 616 0.002
t & sig. 2.08 ** -0.59 0.40 -0.17 -0.27
significant findings leaves us with unanswered questions on causality and temporality of the findings. This is further discussed in the discussion section.
Robustness of concepts. A second set of robustness checks are performed to check the validity of the data. Table 3b shows the effects of the different checks on the effect sizes and significance of the hypothesized effects. All of the modifications confirm the findings in hypotheses 1a and 1c as these remained significant throughout. Hypothesis 2 did not turn significant when
relatedness of innovations was proxied by the NBER PDP originality score. This finding indicates that operationalization is likely not the limiting factor. The lack of findings is elaborated on in the
discussion, the other robustness checks are further discussed in Appendix B.
Discussion
This paper addressed the question of how and if board room political ideology affected innovative outcomes in organizations. I argued that directors hold political preferences which are reflected in their approach towards risk. Risk-avoiding tendencies and pay horizon are expected to lower R&D spending in more conservative boards resulting in relatively lower innovative outputs. Empirically, I found that the statement above is not supported in the dataset. There seem to be clear relations between conservatism and innovativeness and conservatism and ideology. Also, there are signs of an indirect path of conservatism on innovativeness through R&D, these results are however modest (6,58%) and do not take up a large part of the variation in innovativeness. This is explained by the finding that R&D policy did not predict innovativeness. Introducing the longitudinal analysis did not change the conclusion that there was no support for a mediation effect. What it did reveal was that
Table 3b. Robustness check for alternative operationalizations
Modifications\Path a>b (h1a) b>c (h1b) a>c (h1c) a>c’ (h1) n r2 a>d (h2) n r2
No modifications -2.899 .020 -.837 -0.589 439 0.445 -0.653 159 0.034
z / t & sig. -1.79 * 1.31 -2.24 ** -1.05 -0.31
Elimination maybe matches -3.00 0.027 -1.037 -0.081 357 0.437 -2.922 135 0.042
z / t & sig. -1.69 * 1.68 * -2.59 ** -1.18 -1.34
Originality as relatedness -0.035 202 0.018
z / t & sig. -0.92
Regular patent counts 0.216 -52.953 -0,613 439 0.378
z / t & sig. 0,14 -1.45 -0,14
Not identified as CF=0 -6.619 0.019 -1.39 -.124 439 0.456 -0.508 159 0.034
z / t & sig. -2.63 ** 1.20 -2.34 ** -1.08 -0.15
conservatism is likely predicting R&D policy in a causal manner, and that conservatism and innovativeness are only related. As for type of innovation, the hypothesis that more conservatism resulted in higher relatedness was not reflected by the data. A not significant negative relation has been found, which indicates that board room conservatism either has no effect or a negative effect on innovation relatedness.
The findings seem to be in line with previous findings stemming from Upper Echelon Theory that indicate that board room ideology has significant effect on organizational outcomes and
specifically innovativeness (eg. Hambrick & Mason, 1984; Bantel & Jackson; 1989). Also, the
supported relation between conservatism and R&D spending is in line with the predictions from theory on short-termism(eg. Souder & Bromiley, 2012) and risk-avoidance (eg. Hutton, Jiang & Kiamar, 2014). More surprising, in contrast to recent findings (eg. Martin & Nguyen-Thi, 2015; Wang & Hagedoorn, 2014), the analyses did not find convincing evidence for a temporal relation between R&D spending and innovation. This is probably attributable to the small sample size and short period of analysis, which was reduced even more after integrating a time lag. Also, as no entries have been deleted, a very wide variety of organizations and industries have been used, with vastly different business models and probably different processes of R&D, innovation and innovation protection. Lastly, the role of R&D in explaining the effect of ideology on innovativeness is relatively small (6,58% of the total effect). This (lack of) result, nonetheless helps make for interesting discussion.
As this research is of quantitative nature, it is hard to make undeniable statements about all the underlying mechanisms. Especially since not all fixed effects analysis did not return significant findings, I cannot exclude that outside interference has occurred. The only exception here is the relation between ideology and R&D spending, interpretation of the other relations are unsure. What is feasible, however, is theoretically explore three alternative explanations that could not be ruled or controlled for in this research.
Alternative explanations
This paper will now address possible alternative explanation that could not be controlled for which serve as input for future research, followed up by limitations of the study in terms of
One possible mechanism that interacts between ideology and innovation that is not measured yet is culture. According to Tushman and O'Reilly (1997) organizational culture lies at the heart of organizational innovativeness. Some of the many aspects of organizational culture that promote innovativeness are risk-taking, open communication and support for change (Hogan & Coote, 2013; Martins & Terblanche, 2003). Liberal boards can be expected to promote exactly these cultures that welcome innovativeness, change and openness, as this is in line with the preferences that liberal people tend to hold (Jost, Federico & Napier, 2009). Indicating that ideology might affect culture, which in place affects the innovative output. For example, in companies that hold mission statements explicitly aimed at innovativeness, curiosity, exploration and open mindedness. You would expect more experimentation (exploration) and therefore more innovation. Also the direction of causality could be questioned here, as it is equally possible that the relation is reversed or reciprocal. This could for example be the case if innovative companies were to be considered more attractive to liberal employees. For example, because of an underlying culture of innovativeness and openness to new ideas that attracts young and liberal employees open for change (Zaharee, Lipkie, Mehlman & Neylon, 2018). As a supportive culture is hard to synthesize into concrete measures, this could not be
controlled for. Future research using more qualitative approaches might give more insight into the role of culture.
Also, organization structure might play a role. Although successful innovation begins in R&D, the way the efforts are coordinated with the activities of others is crucial (Burgelman & Maidique, 1996). According to the authors, proposed options to maximize R&D efficiency are cross-functional teams, skunk works and new venture divisions. It might be expected that boards with different
ideologies prefer different types and forms of organization (eg. in relation to power distance) resulting in changes in the innovativeness. This has not been controlled for as the concept of structure is too broad and appropriate measures are ill-defined. Nonetheless is this a promising direction for future research to look into.
the members themselves are more innovative. This is however, unlikely to have a substantial effect on the total innovative output. Given that boards are only a very small part of bigger companies, so their own contributions are relatively small. Especially considering that it is not their main activity or responsibility to come up with innovations.
Limitations and future research
As shown in the robustness checks, the way different variables are defined and measured can have a big impact on the final result. Where possible I tried to incorporate best practices well suited for the situation, but inherent to measuring abstract constructs there is always a certain degree of
approximation and theorizing, which inevitably means that you possibly not fully measure the desired construct or that you measure something else.
To show this, I demonstrate the use of R&D spending as a proxy. I used it as indicative of the R&D policy, but others have used it with convincing theory as a for temporal orientation, risk
preference, asset specificity and barrier to entry (Souder & Bromiley, 2012). It is used as a proxy to many different latent concepts, and it is unclear which of these concepts explained the findings. For this reason I will also discuss the conceptualization of innovativeness. I have chosen, as many others did, to use patent data as a proxy to innovativeness (Hall, Jaffe & Trajtenberg, 2001). However, using patents does not always fully capture innovation. Three reasons are: not all innovations can be patented, not all patentable innovations are patented and not all patents have the same quality of innovation (Haščič & Migotto, 2015).
dynamics associated with group level innovativeness (West & Anderson, 1996) or group-level conservatism such as contagiousness of risk-aversion (Suzuki, Jensen, Bossaerts & O'Doherty, 2016). Lastly, many directors in our sample sit on multiple boards simultaneously, it is unlikely that these directors have equal say and affection with all of these boards. Future research will have to unravel if and how board members’ tenure, function and power relative to others will mediate the results.
Sperber and Linder (2016) state that in practice the influence of the decision makers on firm innovativeness is manifold, and heterogeneity of board demographic characteristics, as well as its effects on firm innovativeness, is a complex phenomenon. And they add that most methods used to quantify this particular effect are blind for this complexity, and therefore current research lacks a deeper understanding of the process within the board room (Sperber & Linder, 2016). Their critique also seems to hold for this research, as linear and correlational research does not fully cover the complexity of board room ideology and innovation. In the same line of reasoning, Pettigrew (1992) questions whether demographic-based upper echelon research can be conducted without direct contact with top managers: “No one has ever been anywhere near a top team in an organizational setting, either to directly observe a team in action, or to interview the members about the links between their characteristics and structure, processes of communication and decision making and their impact and performance” (Pettigrew, 1992: 175). Some other indicators of the complexity of the researched relationship will now be discussed.
The proposed linear relations within the conceptual model might not fully reflect the actual situation. Firstly the assumption that bigger donations represent more extreme ideologies in
individuals does not necessarily hold. Donations are dependent on the amount of funds you have. What for a rich person might be a sign of mild support, can be a significant donation for others. Likewise in R&D spending more is not always better, R&D is often related to competition and diminishing returns (Gottinger, 2003). An overlooked, yet important aspect is efficiency and
timeliness of the investment: In technology races, the first mover tends to get the advantage (eg. in the form of patents), whereas the others do not (Gottinger, 2003).
logical and defendable, it does change the interpretation of the results. Adding mean measures of the dependent and the independent variable accounts for large parts of the variance normally explained by the independent variable. Especially as in this sample the sample size (n) is low and the amount of industry types (k) is relatively high, there is little left to be explained. Therefore, the results might actually reflect the effects on or by the outliers rather than the full range of the variables. Further research with novel operationalizations, bigger samples or more complex statistical models (such as panel SEM) might help to overcome these limitations and possibly find stronger results.
Another interesting scope for further research would be to follow up with qualitative research trying to establish the exact drivers and mechanisms of the relation between ideology and
innovativeness in organizations, this could then be input for new and stronger quantitative input to explore the expected relations. Specifically, research could assess the mechanisms of how exactly board room ideology is affecting organizations, also timeliness and group dynamics are aspects that might affect the relation between ideology and innovation, also, organizational culture and structure seem to be interesting and promising mechanisms. Lastly, the possible effect of board power will be discussed.
Board room political opinion might as well be a (skewed yet considerable) sample of the companies' general opinion. In this research ideology is measured only on the board level, but the degree of alignment with the company as a whole, might have a big effect on the influence of the ideology. The power of the board and the amount of support a board receives from its organization can affect its effectiveness. In general, I advise to also look into the effects of the board in relation to the company overall aggregate opinion and actions, to see if this hinders or promotes the boards ideology and policies.
Implications
Based on the findings academics can learn that board room ideology is associated to R&D spending and innovativeness. This thesis serves as additional proof that political ideology is a promising echelon factor. Also, a new factor in understanding the between-organizations variance in R&D policy and innovativeness has been identified in the form of board room conservatism.
influence innovativeness and R&D policies. The political orientation of board members predicts R&D policy as well as the strategy concerning risk and pay-off horizon a company pursues. Innovativeness remains a complex concept, and although ideology is related to innovativeness, no causality could be established. Indicating that this finding can be aid managers in understanding differences between companies in terms of innovativeness through the political preferences of the decision makers. I do not advise practitioners seeking to improve innovation to start hiring liberal individuals as causality has not (yet) been proven.
Conclusion
REFERENCES
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. Action control (pp. 11-39) Springer.
Bantel, K. A., & Jackson, S. E. (1989). Top management and innovations in banking: Does the
composition of the top team make a difference? Strategic Management Journal, 10(S1), 107-124. doi:10.1002/smj.4250100709
Barker, V. L., & Mueller, G. C. (2002). CEO characteristics and firm R&D spending. Management Science, 48(6), 782-801.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173.
Bermiss, Y. S., & McDonald, R. (2018). Ideological misfit? political affiliation and employee departure in the private-equity industry. Academy of Management Journal, 61(6), 2182-2209. doi:10.5465/amj.2016.0817
Berry, M. M. J., & Taggart, J. H. (1994). Managing technology and innovation: A review. R&D Management, 24(4), 341-353. doi:10.1111/j.1467-9310.1994.tb00889.x
Bessen, J. (2007). Programmer documentation on PTO assignee: Compustat matching.
Bessen, J. (2009). NBER PDP project user documentation: Matching patent data to compustat firms. Version: Beta, may,
Bonica, A. (2017). Database on ideology, money in politics, and elections (DIME).
Bonica, A. (2014). Mapping the ideological marketplace. American Journal of Political Science, 58(2), 367-386. doi:10.1111/ajps.12062
Bonica, A. (2016). Avenues of influence: On the political expenditures of corporations and their directors and executives. Business & Politics, 18(4), 367-394. doi:10.1515/bap-2016-0004 Bonica, A. (2019). Are donation-based measures of ideology valid predictors of individual-level
Burgelman, R. A., Maidique, M. A., & Wheelwright, S. C. (1996). Strategic management of technology and innovation Irwin Chicago.
Butros, A., & Taylor, S. (2011). Managing information: Evaluating and selecting citation management software, a look at EndNote, RefWorks, mendeley and zotero.
Carnahan, S., & Greenwood, B. N. (2018). Managers’ political beliefs and gender inequality among subordinates: Does his ideology matter more than hers? Administrative Science Quarterly, 63(2), 287-322. doi:10.1177/0001839217708780
Cefis, E., & Marsili, O. (2006). Survivor: The role of innovation in firms’ survival doi: 10.1016/j.respol.2006.02.006
Chen, W., Kang, M., & Butler, B. (2019). How does top management team composition matter for continual growth? reinvestigating Penrose’s growth theory through the lens of upper echelons theory. Management Decision, 57(1), 41-70. doi:10.1108/MD-02-2017-0147
Chrisman, J. J., & Patel, P. C. (2012). Variations in r&d investments of family and nonfamily firms: Behavioral agency and myopic loss aversion perspectives. The Academy of Management Journal, 55(4), 976-997.
Christensen, D. M., Dhaliwal, D. S., Boivie, S., & Graffin, S. D. (2015). Top management conservatism and corporate risk strategies: Evidence from managers' personal political orientation and corporate tax avoidance. Strategic Management Journal, 36(12), 1918-1938. Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm. Englewood Cliffs, NJ, 2(4),
169-187.
Erikson, R. S., & Tedin, K. L. (2015). American public opinion: Its origins, content and impact Routledge.
Finkelstein, S., & Hambrick, D. C. (1989). Chief executive compensation: A study of the intersection of markets and political processes. Strategic Management Journal, 10(2), 121-134.
Gupta, A., & Wowak, A. J. (2017). The elephant (or donkey) in the boardroom: How board political ideology affects CEO pay. Administrative Science Quarterly, 62(1), 1-30.
doi:10.1177/0001839216668173
Hall, B. H., Jaffe, A. B., & Trajtenberg, M. (2001). The NBER Patent Citation Data File: Lessons, Insights and Methodological Tools,
Hambrick, D. C. (2007). Upper echelons theory: An update. Academy of Management Review, 32(2), 334-343. doi:10.5465/AMR.2007.24345254
Hambrick, D. C., & Mason, P. A. (1984). Upper echelons: The organization as a reflection of its top managers. Academy of Management Review, 9(2), 193-206. doi:10.5465/AMR.1984.4277628 Haščič, I., Johnstone, N., Watson, F., & Kaminker, C. (2010). Climate policy and technological
innovation and transfer.
Hascic, I., & Migotto, M. (2015). Measuring environmental innovation using patent data. OECD Environment Working Papers, (89), 0_1.
Hirschheim, R., & Newman, M. (1988). Information systems and user resistance: Theory and practice. The Computer Journal, 31(5), 398-408.
Hogan, S. J., & Coote, L. V. (2014). Organizational culture, innovation, and performance: A test of schein's model doi:10.1016/j.jbusres.2013.09.007
Hult, G. T., Hurley, R. F., & Knight, G. A. (2004). Innovativeness: Its antecedents and impact on business performance. Industrial Marketing Management, 33(5), 429-438.
doi:10.1016/j.indmarman.2003.08.015
Jaskyte, K. (2013). Does size really matter? organizational size and innovations in nonprofit organizations. Nonprofit Management & Leadership, 24(2), 229-247. doi:10.1002/nml.21087 Jones, G. R. (2010). Organizational theory, design, and change (6th ed., global ed. ed.). Upper Saddle
River, NJ: Pearson/Prentice Hall. Retrieved from https://rug.on.worldcat.org/oclc/316524826 Jost, J. T. (2006). The end of the end of ideology. American Psychologist, 61(7), 651.
Jost, J. T., Federico, C. M., & Napier, J. L. (2009). Political ideology: Its structure, functions, and elective affinities. Annual Review of Psychology, 60(1), 307-337.
Jost, J. T., Nosek, B. A., & Gosling, S. D. (2008). Ideology: Its resurgence in social, personality, and political psychology. Perspect Psychol Sci, 3(2), 126-136.
doi:10.1111/j.1745-6916.2008.00070.x
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263-291. doi:10.2307/1914185
Kim, C., Song, J., & Nerkar, A. (2012). Learning and innovation: Exploitation and exploration trade-offs doi:https://doi.org/10.1016/j.jbusres.2011.07.006
Kim, H., & Kankanhalli, A. (2009). Investigating user resistance to information systems implementation: A status quo bias perspective. MIS Quarterly, 33(3), 567-582.
Klette, T., & Kortum, S. (2004). Innovating firms and aggregate innovation. Journal of Political Economy, 112(5), 986-1018. doi:10.1086/422563
Knight, K. (2006). Transformations of the concept of ideology in the twentieth century. American Political Science Review, 100(4), 619-626. doi:10.1017/S0003055406062502
Kobarg, S., Stumpf-Wollersheim, J., & Welpe, I. M. (2019). More is not always better: Effects of collaboration breadth and depth on radical and incremental innovation performance at the project level. Research Policy, 48(1), 1-10. doi:10.1016/j.respol.2018.07.014
Kor, Y. Y. (2006). Direct and interaction effects of top management team and board compositions on R&d investment strategy. Strategic Management Journal, 27(11), 1081-1099.
Laverty, K. J. (1996). Economic “Short-termism”: The debate, the unresolved issues, and the implications for management practice and research. Amr, 21(3), 825-860.
doi:10.5465/amr.1996.9702100316
Layman, G. C., Carsey, T. M., & Horowitz, J. M. (2006). PARTY POLARIZATION IN AMERICAN POLITICS: Characteristics, causes, and consequences
doi:10.1146/annurev.polisci.9.070204.105138
