The influence of corporate political activities on financial performance : a few new insights

(1)

The influence of corporate political activities on financial

performance: a few new insights

Author: Casper America - 10502394

Msc. Business Administration - Strategy track University of Amsterdam

Supervisor: Pushpika Vishwanathan

(2)

2

Statement of originality

This document is written by Student Casper America who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

Earlier studies have shown an ongoing debate about the relationship between corporate political activities (CPA) and financial performance (FP). This study tries to extend the literature through integrating the market and non-market strategies and the resource dependency theory (RDT). This study will investigate the direct relationship between CPA and FP, and if this relationship will be positively moderated by environmental munificence, environmental dynamism and government dependency. Consequently, I expected that firms their political activities will have different levels of financial outcomes under the different environmental conditions. Moreover, I expected that

government dependent firms would have a higher return on their political activities compared to firms that do not depend on the government. Conclusions were based on the biggest S&P 500 firms from 2010. Analysis revealed an non-significant relationship between CPA and FP. Moreover, contradictory to the hypothesis I found a negative moderating effect of environmental munificence on the lobbying expenditures – FP relationship. Furthermore, environmental dynamism and

government dependency didn't show a significant moderating effect on the CPA – FP relationship. Results indicate that there is no reason to believe that CPA has a relation with financial performance, and that the environmental conditions and government dependency positively moderate the CPA – FP relation.

(5)

5

1. Introduction

According to Baron (2000), a firm's performance depends on their operations and strategies in market and non-market environments. Thus far research conducted on the non-market research field have shown significant insights into the ways that firms shape and interact strategically with the external environment (Holburn & Vanden Berg, 2013).

One major element of a firm's non-market strategies are corporate political activities (CPA) which can be seen as a non-market strategy that firms use to shape government policies to their advantage (Hillman, Keim & Schuler, 2004). Government policies are a source of risk and uncertainty for organizations as these have the ability to shape the competitive environment through critical resources. For instance a decision maker could charge certain taxes, set pollution standards or insert regulations that affect the consumers consumption behaviors (Gale & Buchholz, 1987). As a consequence, engaging in CPA is a way for firms to control and steer regulators into similar directions and simultaneously reduce own risk and uncertainties (Hillman, 1999). An example of an industry which effectively engaged in CPA is the tobacco industry. In the 90’s the tobacco industry was facing serious problems due to tax regulations that negatively affected the consumer behavior in the U.S. As a consequence the tobacco firms started engaging in political strategies in other markets, such as Asia to avoid that these countries issued the same regulations as the U.S. (Hillman, 1999).

According to Hillman, Keim and Schuler (2004), not only the U.S. but also countries within the European union, Russia and other countries have been dedicating scholars to investigate the effects of CPA. Moreover, research into CPA is popular among different disciplines such as sociology, political science, finance and strategic management. As a result, research on CPA is globally analyzed and from a wide variety of research fields.

The different research fields have been trying to get a better grip on CPA by researching with one or several theoretical perspectives (Mellahi et al, 2016). The theoretical perspectives could be distinguished by five main theories when studying the relationship between CPA and financial performance; the resource-based view, agency theory, institutional theory, stakeholder theory and the resource dependence theory (Mellahi et al, 2016). These theories have different perspectives and some of them state that CPA has a negative effect on financial performance and others vice versa. Previous literature focused on one theoretical perspective and therefore didn’t integrate or studied the outcomes of multiple theoretical perspectives. However, a trend can be identified in which more recent literature researched CPA with multiple theoretical perspectives (Mellahi et al, 2016).

(6)

6 Since past studies into the CPA – FP relationship have been showing mixed results it is difficult to draw an overall today. Some scholarships show that CPA has a positive effect on financial performance (Bonardi, Holburn, & Vanden Bergh, 2006; Lux, et al, 2011; Shaffer et al, 2000). On the other hand, different studies argue that there are many challenges changing the policies. Those studies state that there are multiple different interest groups that make it difficult to change policies and achieve the desired result (Hersch, Netter & Pope, 2008; Baumgartner et al, (2009). Finally, there are also studies that find a negative relationship between CPA and financial performance (Aggrawal, Meschke & Wang, 2012; Faccio, 2006; Coates, 2010). As a result of these mixed results it seems that the relationship between CPA and FP has not yet been completely understood. The misunderstanding of this relationship could result in firms making resources available for activities that might not even be beneficial for the company. Consequently, more research on this topic is necessary to have a better understanding of CPA.

Despite the direct effect of CPA on financial performance there are also external conditions that influence this relationship (Hillman, 2009). The resource dependence theory (RDT) states that a corporation could be considered an open system, which depends on different contingencies from the external environment (Hillman, 2009; Pfeffer & Salancik’s, 1978). This theory indicates that factors such as government dependence and environmental uncertainty could be considered important and potentially have an effect on the firms their political activities (Hillman et al, 2004; Pfeffer & Salancik, 1978).

According to Milliken (1987), environmental uncertainties can be seen as external environmental conditions which have an impact on the managers decisions concerning the consequences of the firms’ behaviors and the competitive situations in the future. Moreover, most of the literature concerning environmental uncertainty also suggest that firms their environment affect the availability of resources for the firm (Dess & Beard, 1984; Aldrich 1979).

Thus far, empirical studies acknowledge the importance of external conditions such as the industry and environmental specific characteristics and consequently they studied the relationship it has with CPA (Hillman, et al, 2004). Most of these studies intended to test if structural variables such as the concentration and number of firms would have an impact on the firm’s ability to invest in CPA (Schuler et al, 2002; Schuler, 1996). According to Schuler et al (2002), firms that are positioned within a concentrated industry have a higher tendency to participate in political activities than firms that are positioned within a more fragmented industry. Furthermore, some scholarships even studied the effects of competitors participating in political activities (Gray & Lowery, 1997). Whereas many studies acknowledge the importance of industry specific characteristics on CPA, most of the studies overlook the moderating effect of Dess and Beard’s environmental conditions. Different

(7)

7 industries experience different levels of munificence (growth), dynamism and complexity and understanding the effects of these conditions on CPA is important.

Government purchases within the U.S. have been steadily growing within recent years and as a result, rivalry among sellers arises as the government expenditures amount up to 33.1 percent of the GNP (Zardkoohi, 1985). Therefore, some firms become government dependent because they receive resources directly from the government through contracts which are covering a significant amount of the firm's total revenues (Schuler, Rehbein & Cramer, 2002).

In relation to CPA, multiple studies have been conducted to see if there is a relationship between government dependence and CPA. Most of the empirical evidence seem to support a positive relationship (Meznar & Nigh, 1995; Birnbaum, 1985). On the other hand, more recently Blumentritt (2003), didn’t find a relationship. According to Hillman, et al (2009), most of the past studies concerning CPA – government dependence were conducted within other research fields, such as political science or political economies. Most of these past studies didn’t used the RDT in a meaningful way since these research fields have different theoretical foundations in comparison to the management research field (Hillman et al, 2009). Therefore, integrating the RDT in a meaningful way would be necessary to investigate the effect of government dependency and CPA. Since recent studies, show mixed results regarding the effect of CPA on financial performance this research will first of all examine the direct effects. Furthermore, according to the RDT external factors might affect the relationship between CPA and financial performance and thus we extend the current literature by looking at the moderating effects of government dependence and environmental uncertainty. This results into the following research question of this study:

To what extent does environmental uncertainty and government dependency have a moderating effect on the relationship between corporate political activities and financial performance?

This research looked at CPA as a non-market strategy and consequently I examined the effects of lobbying expenditures and PAC contributions on firm’s financial performance. Moreover, I used the RDT as an overarching framework to explain our moderators environmental munificence, environmental dynamism and government dependence. As a result, I found that both lobbying expenditures and PAC contributions didn’t have a significant relationship with financial performance. Therefore no evidence has been found concerning the CPA – FP relationship. Moreover, contradictory to our hypothesis munificence has a negative moderating effect on the relationship between lobbying expenditures and financial performance. Interestingly, no significant effect has been found between munificence and PAC contributions. Finally, both environmental dynamism and government dependency showed an insignificant moderating effect on the CPA – FP relationship.

(8)

8 This study will contribute to current literature in several ways. First of all this study contributes to the ongoing debate about the non-market strategy CPA and their relation to financial performance. Moreover, this study will integrate the RDT in a meaningful way to extend the current literature on the external firm conditions environmental munificence and environmental dynamism, and their relation to CPA. (Hillman, et al, 2004). Finally, this study will build on previous studies that examined the effects of government dependency and CPA. This will be done through looking at the moderating effect of government dependency on the CPA – FP relationship

2. Literature review

This section will be structured as follows. First I will explain the theory about market and non-market strategies. Then I will look at the research to date on corporate political activities (CPA) and the relation it has with financial performance (FP). Within this section CPA will be considered to be a non-market strategy. Moreover, I will elaborate on the resource dependence theory and review the existing literature on environmental uncertainty and government dependence. I will explain, why according to the RDT there might be a moderating effect by these conditions on the CPA – FP relationship. Environmental uncertainty will be evaluated by looking at two separated components of this variable and their individual relationship on the CPA – FP relationship. Those separated components are environmental munificence and environmental dynamism.

2.1 Non-market strategies

According to Baron (1995), strategy formulation should not only be focused on activities within the firm but also outside the firm. This is due to the fact that the firms environment consists of market and non-market components. Market activities are those activities in which a firm has interactions with other firms through participating on a certain market in which firms make personal agreements. The agreements consist of economic transactions between parties that exchange (monetary) resources and are aimed to improve the performance (Baron, 1995). For example, a bus touring company agrees on a contract to buy five new busses in the upcoming years from automobile manufacturer Mercedes-Benz. On the other hand, non-market activities are voluntary and more concerned with stakeholders such as the public institutions, the media or governments. The non-market environment is focused on political, legal forces and social interactions that are not part of the direct market, but instead it has an effect on the overall performance (Mahon, Heugens

(9)

9 & Lambertz, 2004; Baron, 2002). Consequently, by participating in non-market activities firms can shape the environments in a matter that would be beneficial for themselves. For example, a firm decides to build up relationships with governmental policy makers. According to Hillman, et al (2004), non-market strategies and market strategies could be carried out complementary of each other and even be substitutes. However, effective implementation is necessary. Past research have shown that the integration of political and market strategies in certain optimal ways could have a significant effect on the firm’s financial performance (Shaffer et al, 2000; Mellahi et al, 2016; Bonardi, 2004). In this study the variable CPA will be considered to be a part of the non-market strategies that firms could conduct.

2.2 Corporate political activity and financial performance

One common source of risk and uncertainty for firms are government entities that are in a position to shape the competitive environments through legislation and other policies. Through corporate political activities firms try to reduce the risks and uncertainties and try to steer the government entities into favorable directions for the firm (Hillman, Keim & Schuler, 2004). The most common view concerning CPA within the strategy literature is that CPA could be seen as an important non-market strategy for firms to improve their financial performance (Bonardi, Hillman, & Keim, 2005; Hillman, Keim, & Schuler, 2004). According to Oliver and Holzinger (2008), firms are able to manage politics strategically through multiple CPA actions which will focus on maximizing the returns of the political environment. As a consequence of CPA a firm's financial performance might be affected directly and indirectly. The direct effect of CPA is a result of several firm opportunities such as: limiting competitor’s ability to participate in a certain product market, securing contracts with the government and the permission to operate (Hadani & Schuler, 2013). Furthermore, CPA could also indirectly affect the firm through lowering the effects of policies that are harmful or influencing the policy environment in favor of the firm and its activities (Epstein, 1969).

However, despite the possible positive effects of CPA, the decision of a firm to participate in CPA is still important because of several reasons. First of all, participating in CPA could be very expensive. For example, within the steel industry a firm has to spend at least 1.5 million dollar for economic and legal consultants for every unfair trade petition (Schuler, 1996). Secondly, according to Lenway and Rehben (1991) firms could also decide to become a free rider within an industry. Free riders could be considered firms within an industry that incur no costs and don’t take any political actions. As a consequence of free riding firms could still have the benefits because other firms within

(10)

10 the industry do participate in political actions and as a result of their actions, there are mutual benefits for all the firms within the industry. On the other hand, when the majority of the firms become free riders this could also result in no collective benefits at all (Olson, 1965; Yoffie, 1987). Thirdly, nowadays it is more complex to use political strategies than it used to be in the past. This is because of three reasons; the decentralization of power into subcommittees, the increasing availability of information through technology and firm reputation might be harmed when they are associated within the policy arena (Schuler, 1996).

Therefore, assigning resources to CPA won’t pay off for every player within an industry and most of the time only the bigger players will participate in CPA (Lenway & Rehbein, 1991). The differences of CPA engagement are, according to Hillman (2013), based on the industry, the firm, institutional and issue. This would for example mean that a firm within the oil industry is more likely to engage in political activities then a firm that is located within the financial services. This could be a consequence of activist groups pressuring legislators into making new policies against CO2 emissions that would be negative for the firms within the oil industry.

Overall, this study suggest that there is a positive relationship between CPA and financial performance. This is because multiple scholars conducted research on non-market strategies and how they contribute to the organizational outcomes that ultimately affect the financial performance (Mellahi et al, 2016). The majority of the studies showed a significant relationship between important non market strategies and positive organizational outcomes (Mellahi et al, 2016). More specific, research into non market activities had several reasons to believe that there is a positive relationship. First of all, Hillman and Keim (2004), stated that non market strategies will result in stronger relationships with the primary stakeholders. Secondly, Madsen & Rodgers (2015) state that non market strategies will increase the access to finance. Finally, Frynas et al (2006), states that non market strategies will ensure preferential access to political resources over competitors. Lux, et al (2011), did a meta-analysis on CPA and summarized the CPA – FP literature to conclude that there indeed is a positive relationship between CPA and financial performance. Consequently, the majority of past research on non-market strategies has proven to be positively associated with financial performance and thus our first hypothesis indicates:

(11)

11 2.2 Resource dependency theory (RDT)

According to the resource dependency, theory firms are embedded in an external environment and their performance depend on the interaction with actors within this external environment. Firms performance depend on the relationships with actors which will grant them access to certain resources that could be considered necessary for the success and survival of the firm (Pfeffer & Salancik, 2003). For example, within the gas and oil industry firm’s production depends on the government. This is because they are the owners of the necessary resources and they are the only one that could provide them with the institutional legitimacy to operate. As a result, firms need to anticipate on the external environment to reduce their dependency and the uncertainties that they encounter.

Firms are embedded within an environment which has a degree of uncertainty. Uncertainty could be considered the difficulty for firms to predict the position of the environment in the future. According to the RDT, when the uncertainty is related to their core resources, firms need to avoid or reduce the uncertainty since it could potentially harm the firms operations (Nienhuser, 2008). Consequently, if managers want to survive, they have to successfully shape their strategies in a matter that they maintain and obtain the resources that are necessary for the firm (Pfeffer & Salancik, 2003). As a result, for firms to create a sustainable competitive advantage (CA), they would have to ensure that they combine the internal that they have with the external resources (Hart, 1995). Due to the point of view of the RDT, in this study I expect that the variables; government dependence, environmental dynamism and munificence will be relevant regarding the relationship between CPA and financial performance.

2.4 Effect of environmental munificence on the relationship between corporate political activities and financial performance

According to Dess and Beard (1984), environmental munificence represents the extent to which an environment is capable of supporting sustainable growth. The degree of munificence within an environment is based on its growth or decline rate and shows the change of the demand within a certain marketplace (Shou, Yang, Zhang & Su, 2013). High munificence within an industry could result in firms having more resources because of higher margins and increased sales growth (Cyert & March, 1963). The additional resources could be useful for firms since they can be used as a

(12)

12 buffer in times when there is scarcity. Furthermore, according to Dwyer and Oh (1987), additional resource availability to firms result into firms using those resources on other activities than the production activities. For example, a firm is located in an environment with high munificence and as a consequence they have extra monetary resources. The monetary resources aren’t necessary for the production process so as a consequence the firm uses it for R&D expenditures.

The RDT main proposition is that firm survival is based on the ability of the firm to obtain important resources from their external environment (Hillman et all, 2009). As a consequence, firm’s will use a wide variety of tactics to reduce the uncertainties when they try to obtain their necessary resources. In this study I look at the moderating effect of munificence by the application of the RDT. Consequently, firms in high munificent environments, according to the RDT perspective would have additional resource which they would spend on non-market activities to reduce uncertainties and

ensure their profits. One

commonly known tactic that reduces uncertainty is CPA, and environments that are characterized by high munificence will more likely use the additional (monetary) resources to invest in political activities, this is because political activities are a common way for firms to reduce uncertainties (Hillman, et all 2009). Managers of firms operating in a low munificence environment will most likely don’t have the additional resources and invest the available (monetary) resources on core market activities instead of nonmarket activities. Market activities such as the production process or advertising will be prior activities compared to nonmarket activities such as CPA and corporate social responsibility (CSR). Consequently, political activities will be more beneficial within high munificent markets compared to low munificent markets.

Furthermore, According to Dwyer and Oh (1987), high munificence within markets will also be accompanied by firm opportunities in profit and sales. Due to this, there will be a bigger threat of governmental policies that control this growing market compared to a stable market. As a result, the interaction with the government through political activities will be more valuable when there is a growing market because political activities can help steer the market into favorable directions for the firm.

Additionally, according to Dess and Beard (1984), firms competing in environments with low munificence have a higher tendency to commit illegal acts. As a result, firms within environments with high munificence, on average, have better reputations then firms within low munificence environments (Dess & Beard, 1984). According to Hillman et al (2004), policy makers prefer to be associated with firms that have a good reputation and a good reputation has been linked to the success of the firm’s CPA. Therefore, the returns of political activities for firms will, on average, be higher within high munificent environments. For example, automobile firm Volkswagen publishes

(13)

13 wrong numbers about their cars CO2 emission rates which results into an environmental scandal. Publicly makers now doubt their integrity and don’t want to be associated with Volkswagen because it might cost them votes or they might provide them with untrustworthy information. Consequently, Volkswagen there political activities are less beneficial nowadays as compared to before the scandal.

In short, high environmental munificence results into additional resources for firms, that according to the RDT, should be used to spend on CPA to reduce uncertainties. Also, higher growth market will result in more governmental involvement to control the growth, and thus political activities will be more valuable within munificent markets. Finally, low munificence result into more illegal acts within an industry. This simultaneously result in lower firm reputations which directly has a negative effect on the CPA – FP relationship. Therefore the following hypothesis indicates:

H2: The relationship between CPA and financial performance will be stronger within firms with high environmental munificence

2.5 Effect of environmental dynamism on the relationship between corporate political activities and financial performance

According to Aldrich (1979), Dynamism is considered to be the instability or turbulence within an industry. Dynamism is characterized by the needs of customers, industry technologies and how often products change. High dynamism often result into making it harder for firms to predict the happening of certain events in the industry and as a result the risk and uncertainty increases.

The RDT looks at how a firm can control the environment to reduce uncertainties through market and non-market tactics (Hillman et al, 2009). From the RDT perspective I would assume that a firm within a highly dynamic environment would participate in CPA to reduce the increased uncertainties that are caused by the changes in technology, products and customer demands.

As a result, I propose that in times of high dynamism, CPA has a bigger impact on financial performance and thus could be seen as a motivator. This is because of several reasons. First of all, in highly dynamic environments changes occur more often and having a competitive advantage will become more important (Porter, 1980). Through CPA firms can obtain competitive advantages over their competitors through several agreements or new policies (Hadani & Schuler, 2013). For example, Phillip Morris operates in a dynamic E-smoke market and develops a new very innovative safe and healthy E-smoker. However, the sales are disappointing because other manufacturers used new technologies and unhealthy substances to develop a new E-smoker which has a better flavor. Consequently, Phillip Morris lobbies for more restrictive health policies which resulted in the banning of the unhealthy substances within the competitors new E-smokers. As a result, Phillip

(14)

14 Morris steered the dynamic environment to go in the direction of healthy E-smokers in which they are now leading the game.

Secondly, CPA could create firm value through building strong relationships with policy makers (Hillman, 1999). The strong relationships with policy makers established by interacting with them through CPA could ensure that the policy makers are more concerned with the firm, trust the firm and make them have a better idea of the firms integrity (Hillman, 1999). Policy makers could benefit from firms that are integer and trustworthy because they could provide them with useful knowledge on certain relevant topics or other information from which they know that it is reliable (Hillman, 1999). Consequently, within dynamic industries, the higher amount of instability and uncertainty leads to information and knowledge that is more complex and it becomes harder for policy makers to process information, make rational decisions and support those decisions with evidence (Eisenhardt & Bourgeois, 1988). Therefore, within dynamic environments trustworthy firms their political activities could become more valuable because policy makers need more reliable knowledge about new technologies and products. The knowledge provided by firms makes it easier for them to process information and make rational decisions (Hillman, 1999). Contradictory, stable environments have stable technologies and products and the studies and knowledge provided by firms through CPA will have less impact on policies. As a result, the knowledge and studies provided by firms within dynamic environments compared to stable environments will have more relevant information about industry factors such as new technologies or products which will give more opportunities for beneficial conclusions for themselves and consequently steer the policy makers into making favorable decisions.

Thirdly, according to Gary Castrogiovanni (2002), in the past decades the interdependence of environmental elements have increased and resulted into more dynamic industries. One important aspect of a high dynamic industry is that it increases the anxiety of managers and political decision makers, they are afraid to make bad decisions that affect the firm or government in a negative matter (Schimmer & Brauer, 2012). As a result of the anxiety, managers and political decision makers want to avoid negative outcomes and become more risk averse (Audia & Greve, 2006). According to Dess and Beard (1984), risk averse behavior could result into long term contracts to reduce the uncertainty and create a steady flow of cash or production goods for a period of time. Consequently, within high dynamic environments the political activities could become more valuable for firms due to the fact that they could more likely contribute to long term contracts which reduces their uncertainty. For example, in ten years from now the aircraft industry becomes very dynamic and uncertain. As a result, the minister of defense signed a five year long contract with Boeing to avoid possible negative consequences of fluctuations within the industry. As a result, Boeing now

(15)

15 has to worry less about the environment since they will have a contract with steady sales for five years.

In short, the competitive advantages that firms obtain through political activities become more valuable within dynamic environments. Moreover, the developed relationships between legislators and firms become more beneficial within dynamic environments. Finally, within dynamic environments the relationships between firms and policy makers will be more beneficial due to the fact that policy makers are more interested in long term contracts. Therefore, our third hypothesis indicates:

H3: The relationship between corporate political activities and financial performance will be stronger in firms with high environmental dynamism

2.6 Effect of government dependence on the relationship between corporate political activities and financial performance

Since the government has become a large and steady economic buyer within the current US market there has been a lot of increased rivalry among selling firms (Zardkoohi, 1985). As a consequence, the competing firms try to ensure that they distinguish themselves through common strategies such as high quality or low prices. Alternatively, according to Schuler and Rehbein (1997), quality and pricing strategies are not the only strategies available for government dependent competing firms. Firms that receive a significant amount of their revenues through government entities could engage in corporate political activities for an attempt to proactively shape their political environment (Schuler & Rehbein, 1997). Consequently, resource dependency theorist have stated that when the government is an important buyer from a selling private firm, the firm will become proactive in shaping their political environment and thus, government dependence could be seen as a CPA predictor (Pfeffer & Salancik, 1978; Hansen & Mitchell, 2000).

In this study I propose that the effect of CPA on financial performance will be higher for firms that are government dependent. This is because of several reasons. First of all, the amount of years that firms participate in CPA is on average longer for government dependent firms compared to firms that are not (Schuler, Rehbein & Cramer, 2002). As a consequence of the increased CPA time span, the benefits of CPA will be higher because of better partnerships between firms and legislators

(16)

16 (Schuler, Rehbein & Cramer, 2002). For example, a firm that participates in CPA for multiple years build better and more relationships compared to a firm that only participate once in a while.

Secondly, according to Shaffer (1995), government policies could be considered an important determinant in a firms financial performance. Policies have a certain amount of uncertainty because they can shape the firms environment. According to the RDT, firms want to decrease the risk and uncertainty of the environment and avoid losses and potentially increase their profits (Hillman & Hitt, 1999). Government dependent firms depend on government purchases which could directly reduce their sales if the government decide to lower certain purchases. Consequently, the government policies could be considered a bigger threat for government dependent firms. Therefore, political activities will be more beneficial for government dependent firms since it also helps them to keep up the direct revenues, created through sales to the government (Hillman & Hitt, 1999). This is in line with the RDT, since it states that the higher the firms government dependency the higher their interest will be to interact with the government (Hillman, et al, 2009).

In short, according to the RDT firms that are government dependent have a higher tendency to participate within political activities. This is because government dependent firms are more vulnerable to policy changes because these could directly affect the firms sales. As a result, for government dependent firms the returns of political activities will be higher compared to firms that don’t have sales directly to the government. Moreover, since government dependent firms have a higher tendency to participate in political activities, they build up better partnerships, resulting in a higher return on CPA investments. Consequently, this study’s final hypothesis states:

H4: The relationship between corporate political activities and financial performance will be stronger within firms with higher government dependency

(17)

17

Figure 1. Conceptual model

H1 H2 , H3 H4 Financial performance Environmental uncertainty -Munificence - Dynamism Government dependence

(18)

18

3. METHOD

3.1 Sampling strategy

Within this study the sample is based on firms listed within the Standard & Poor’s stock market index (S&P 500). The S&P 500 is a stock market index based on firms within the United States. To date some researchers consider it the best representation of the US stock market and because of that the index is widely used within academic studies. The initial sample in this study was based on 635 firms that were listed within the Fortune 500 in 2010. I started to check the initial sample to ensure that all errors have been corrected or deleted from the file. Reasons for error could for example be bankruptcy, a merger between firms, name change, acquisition or anything else. As a result, the final data set contained 596 firms that could be used for this study. The reason this research is only based on large firms located within the US is because of limited available data. Consequently, this research is focused on large US firms and thus the results won’t be relevant for other countries with different political systems and small firms.

3.2 Variables

3.21 Dependent variable

Financial performance will be the dependent variable within this study. According to Orlirzky et al (2003), there are market-based measures and accounting based measures for financial performance. Market-based measures could be measured through calculating Tobin’s q or the price per share. On the other hand, accounting based measures are based on return on sales or the return on assets (ROA). The most popular measure for financial performance and the measure that will be used within this study is the accounting-based measure (Orlirzky et al, 2003). According to Gentry & Shen (2010), researchers generally use market measures to look at the long term or future of financial performance and the accounting measures are more suitable for the past or short term financial performance. This study will look at the average ROA (2013-2016) and because this study doesn’t take longitudinal effects into account it seems more suitable for our study to look at the ROA. I take

(19)

19 the average of four years to reduce the impact of abnormalities within specific years. According to Carpenter & Sanders (2002), ROA shows the efficiency of the firm’s assets in relation to their returns. The ROA is obtained from the COMPUSTAT database and has been calculated through dividing the firms net income by the total assets. As a consequence, the higher the number the higher the firm’s financial performance.

3.22 Independent variable

Corporate political activities will be the independent variable within this study. This independent variable, CPA, will be measured through looking at two different variables; the lobbying expenditures and the political action committee (PAC) contributions. This study collected this data through the center of center for responsible politics. The center of responsible politics is a non-profit organization that mainly focus on tracking the money within the U.S. politics. They try to determine what the effect is of money on elections and the public policies. Their vision is to provide clear and unbiased information on the role of money within the politics and policies. Because the information is complete and completely public available through their Opensecrets.org website, it could be considered a reliable source.

The lobbying expenditure variable is measured through taking the average of each firms four year lobbying expenditures between 2013 and 2016. The reason for this is to reduce the impact of

abnormalities within specific years.

The PAC purpose is to raise and spend money to elect and defeat certain candidates. PAC’s can donate 5000$ to the committee of a candidate each election. Furthermore, they are allowed to give a maximum of 15000$ each year to a national party committee, and 5000$ to another PAC. PAC expenditures is measured through taking the average of 2014 and 2016. The reason for this is to reduce the impact of abnormalities within specific years.

3.23 Moderating variables

Environmental uncertainty will be one of the moderating variables within this study. According to Dess and Beard (1984), environmental dynamism and environmental munificence could be considered indicators of environmental uncertainty. These indicators have also been used by Sun & Price (2016) to measure the environmental uncertainty in their study. The data has been collected using the COMPUSTAT database with the fundaments and industry specific annual data. Each firms industry specific annual sales were collected for the years 2013-2016 through using a four digit

(20)

20 standard industrial classification (SIC) code combined with a ticker symbol.

Environmental dynamism is calculated through aggregating the years 2013-2016. First, this study calculates the standard deviation for a four year period for each firm. Secondly, the standard deviation is divided by the annually mean industry sales for the four year period. This results in the change in industry sales within this four year period. (Sun & Price, 2016).

Environmental munificence is measured through calculating a regression coefficient for the four years of each firm’s industry. Finally, the regression coefficient is divided by the mean of the four years industry sales to capture the growth trend of the industry (Boyd, 1995).

Government dependence within this study has been turned into a dummy variable to see the difference between firm’s that are government dependent and firms that are not. When the firm’s didn’t dependent on the government they were coded as 0, and when they did they were coded as 1. This study measured government dependence through looking at the defense contracts with the us government and has been collected through examining the government’s fiscal year reports. This is a common way to measure government dependence and has been applied by multiple scholars (

Salancik, 1979;

Schuler, Rehbein & Cramer, 2002

).

3.24 Control variables

Firm size is considered to be a relevant control variable since previous research states that CPA depends on firms their ability to invest in non-market strategies (Mellahi et al, 2016). Moreover, firm’s ability to invest in non-market strategies is related to their slack resources and the size of the firm. Generally, larger firms will have more available (financial) resources and interest in CPA activities (Hillman, 2013; Coates & John, 2012; Lenway & Rehbein,1991). Within this study the firm size is measured as the logarithm of the number of employees. The data has been derived from COMPUSTAT and the average of four years have been taken to reduce the impact of abnormalities within specific years.

Leverage has been used as a control variable. It is the ratio of the total liabilities divided by the firm’s total equity. I use it as a control variable within this study because leverage has a direct link to CPA and financial performance. First of all, firm need (monetary) resources to fund political activities and high amounts of leverage could become problematic. Consequently, firms have less resources to pay for lobbyist and thus CPA is lower. Moreover, high leverage could result in less growth investments because priority lies with the higher debt expenses and thus it could affect a firms financial

(21)

21 performance as well. The data has been derived from COMPUSTAT and the average of four years has been taken to reduce the impact of abnormalities within specific years.

Slack resources could be considered a relevant control variable since various of studies propose that the resource availability affect the ability for a firm to participate in politics (Schuler, 1996; Bourgeois, 1981). According to Keim & Baysinger (1998), maintaining and securing resources is very important for a firm to implement a political strategy successfully. Within this study slack resources are measured through dividing the current assets with the current liabilities. The data has been derived from COMPUSTAT and the average of four years has been taken to reduce the impact of abnormalities within specific years.

3.3 Data analysis

Before conducting the hierarchical multiple regression and moderating analysis, this study tested several relevant assumptions concerning this statistical analysis. The independent variables; firm size, leverage, slack, lobbying expenditures and PAC contributions were not a combination of other independent variables and thus I met the assumption of singularity. Moreover, I examined the correlations (table 1) between the variables and concluded that no variables were relatively highly correlated, except for lobbying expenditures and PAC contributions (r = .72). However, both variables are part of the independent variable corporate political activities and thus I expected a correlation. Moreover, when I look at the control variables within this study I see that, as expected firm size and slack have a significant relationship with ROA. Firm size shows the highest, but still weak correlation with ROA (r=0.24) and slack the second highest (r=0.15).

Furthermore, after I checked for multi-collinearity and heteroscedasticity through examining the variation inflation factor (VIF) there were no reasons to believe that there are any problems of multicollinearity, since all VIFs within the models are below 10 (Hardin, 1996). Furthermore I checked our variables for their kurtosis’s and skewness and after a visual examination I concluded that there is skewness within our variables. However, since I am working with economic data, I could assume that some outliers are relevant and thus they haven’t been removed.

(22)

22 Table 1. Correlations Variables 1 2 3 4 5 6 7 8 9 1. Lobbying expenditures - 2. PAC 0.72** - 3.ROA 0.88** 0.06 - 4.Munificence 0.41 0.02 0.28** - 5.Dynamism -0.53 -0.11* -0.29** -0.05 - 6.Government dependence 0.29** 0.36** 0.033 -0.03 -0.01 - 7. Firm Size 0.35** 0.34** 0.24** 0.14** -0.28** 0.15** - 8. Leverage 0.08 0.11** -0.48 0.01 -0.12** 0.03 0.00 - 9. Slack 0.03 -0.07 0.15** 0.12* 0.12* -0.05 -0.10* -0.08 -

Note: PAC, political action committee; ROA, Return on assets *Correlation is significant at p<0.05

**Correlation is significant at p<0.001

Table 2 below shows the descriptive statistics concerning the N, mean and standard deviations of our variables.

Table 2. Frequencies (N), Mean & Standard deviation for the variables

N Mean Std. Deviation Lobbying expenditures 589 1669146.22 2963539.09 PAC 590 201345.02 375251.60 Return on Assets 573 0.04 0.07 Munificence 483 -0.00 0.14 Dynamism 526 0.14 0.14 Government dependence 589 0.11 0.31 Firm Size 567 1.38 0.54 Leverage 567 3.12 8.08 Slack 489 1.59 0.82

(23)

23 The direct effect of CPA on financial performance

A two stage hierarchical multiple regression has been conducted with the dependent variable return on assets (ROA). First, I entered the control variables firm size, leverage and slack. Thereafter, I entered both lobbying expenditures and PAC contributions. The reasons I added lobbying expenditures and PAC contributions at the same stage is because they are both part of the independent variable CPA. The regression statistics are reported in Table 3. The hierarchical multiple regression showed that at step one, firm size, leverage and slack all three together significantly contributed to the regression model, (F(3,474) = 16.43, p < .001), with an total explained variance R² of 0.09. Consequently, 9% of the variation in the return of assets is caused by the firm size, leverage and slack. The hierarchical multiple regression showed that at step two, adding lobbying expenditures and PAC to the model was not significant. In the model only the variables firm size (β = 0.27, p < .001), and slack resources (β = 0.17, p < .001) showed to be significant predictors in the relationship with ROA. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 2.07). So no causes for concerns of independent error (Field, 2009).

Table 3. The hierarchical regression analysis for the variables predicting ROA.

Variable

β

t

SE

R

2

∆R

2 Step 1 0.09 0.09** Firm size 0.27** 6.18 0.06 Leverage 0.01 0.14 0.00 Slack 0.17** 3.90 0.04 Step 2 0.10 0.09 0.00 Firm Size 0.27** 5.80 0.07 Leverage 0.01 0.18 0.00 Slack 0.17** 3.76 0.04 Lobbying expenditures 0.04 0.56 0.00 PAC -0.04 -0.57 0.00

Note: PAC, Political action committees. *Correlation is significant at p<0.05 **Correlation is significant at p<0.001

(24)

24 The moderating effect of environmental munificence on the CPA – FP relationship

Table 4 and 5 present the results of the moderation analysis of environmental uncertainty. As mentioned before environmental uncertainty contains environmental munificence and environmental dynamism. In this study we will test the moderating effect of both the environmental uncertainty variables on the corporate political activities (CPA) – financial performance (FP) relationship. Moreover, CPA will be evaluated through looking at the lobbying expenditures and PAC contributions and thus I will test each moderator on both the CPA variables.

Table 4. The moderating effect of munificence on the lobbying expenditures – ROA relationship

ROA Adjusted

R

2

R

2 change Model 1 0.15**

β

SE Model 2 0.16** 0.01*

β

SE ZLobbying expenditures (X) 0.02 0.00 .05 0.000 ZMunificence (M) 0.24** 0.02 0.23** 0.02 ZFirm Size 0.24** 0.01 0.24** 0.01 ZLeverage 0.00 0.00 0.00 0.00 ZSlack 0.13* 0.00 0.16** 0.00 Moderator (X*M) -0.11* 0.00

Note: ROA, Return on assets *Correlation is significant at p<0.05 **Correlation is significant at p<0.001

Table 4 includes the hierarchical regression including two separate models. First, model 1 contains all the variables except the lobbying expenditures - munificence interaction term (moderation). This model shows a significant result, F(5,420) = 16.08, p<.01. Moreover, Model 1 has an adjusted R2_of .15 which indicates that the 15% of the variance in the return of assets could be explained by the model. Individually munificence (β = 0.24, p < .001), firm size (β = 0.24, p < .001) and slack (β = 0.13, p < .001) show a significant relationship with ROA. Meaning that all have a positive

(25)

25 contribution with regard to ROA. Lobbying expenditures and leverage did not show a significant relationship with ROA. Secondly, I introduced the moderation into the analysis (model 2) to test the moderating effect of environmental munificence. This model was significant result, F(6,419) = 14,35, p<.05. The adjusted R2_{of model 2 is .16 which indicates that 16% of the variance in the return of} assets could be explained by this model. Clearly model 2 did not add a fair amount of total explained variance in comparison with model 1, the without the interaction coefficient, R2_{change accounts} only for 1%. The significant predictors in model 2 were: munificence (β = 0.23, p < .001), firm size (β = 0.24, p < .001), slack (β = 0.16, p < .001) and the interaction term (β = −0.11, p < .05). The interaction had a negative impact on the dependent variable This indicates that there is a small negative moderating effect of munificence on the relationship between lobbying expenditures and ROA. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 1.94). So no causes for concerns of independent error (Field, 2009).

Table 5 examines the moderating effect of munificence and includes the standardized value of munificence, PAC, all the other variables and the interaction term between PAC and munificence.

Table 5. The moderating effect of munificence on the PAC contributions – ROA relationship

ROA Adjusted R2 R2_change Model 1 0.15**

β

SE Model 2 0.15 0.00

β

SE ZPAC contributions(X) 0.00 0.00 .01 0.00 ZMunificence (M) 0.24** 0.02 0.22** 0.02 ZFirm Size 0.25** 0.01 0.25** 0.01 ZLeverage 0.00 0.00 0.00 0.00 ZSlack 0.13** 0.00 0.14** 0.00 Moderator (X*M) -0.07 0.04

Note: PAC, Political action committees; ROA, Return on assets *Correlation is significant at p<0.05

(26)

26 Table 5 includes the hierarchical regression including two separate models. First of all, model 1 contains all the variables except the moderation. This model shows a significant result, F(5,420) = 16,02, p<.01. Moreover, Model 1 has an adjusted R2_{of .15 which indicates that the 15% of the} variance in the return of assets could be explained by the model. Individually munificence (β = 0.24, p < .001) ,firm size (β = 0.25, p < .001) and slack (β = 13, p < .001) show a significant positive relationship with ROA. However, PAC contributions and leverage don’t show a significant relationship. Model 2 contains all the variables including the interaction term of PAC and munificence to test the moderating effect of environmental munificence. As a result, model 2 shows a significant result, F(6,419) = 13.76, p<.01. Individually, munificence (β = 0.22, p < .001) , firm size (β = 0.25, p < .001) and slack (β = 0.14, p < .001) showed a significant beta. Moreover, R2 _change, the interaction term between PAC and munificence, PAC contributions and leverage were not significant meaning that there was no significant moderation effect. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 1.96). So no causes for concerns of independent error (Field, 2009).

The moderating effect of environmental dynamism on CPA – FP

Table 6 examines the moderating effect of dynamism and it includes the standardized value of dynamism, lobbying expenditures, all the control variables and the interaction term between lobbying expenditures and dynamism. On the other hand, Table 7 examines the moderating effect of munificence and includes the standardized value of dynamism, PAC, all the other variables and the interaction term between PAC and dynamism.

(27)

27

Table 6. The moderating effect of dynamism on the lobbying expenditures – ROA relationship

ROA Adjusted R2 R2 _change Model 1 0.18** Beta SE Model 2 0.18 0.00 Beta SE ZLobbying expenditures (X) 0.04 0.00 .05 0.00 ZDynamism (M) -0.31** 0.02 -0.30** 0.02 ZFirm Size 0.20** 0.01 0.20** 0.01 ZLeverage -0.03 0.00 -0.02 0.00 ZSlack 0.19** 0.00 0.19** 0.00 Moderator (X*M) -0.04 0.00

Note: ROA, Return on assets *Correlation is significant at p<0.05 **Correlation is significant at p<0.001

Table 6 includes the hierarchical regression including two separate models. First of all, model 1 contains all the variables except the Lobbying expenditures - dynamism interaction term. This models shows a significant result, F(5,428) = 19.98 p<.01. Moreover, Model 1 has an adjusted R2_of .18 which indicates that the 18% of the variance in the return of assets could be explained by the model. Individually dynamism (β = 0.31, p < .001), firm size (β = 0.20, p < .001) and slack (β = 0.19, p < .001) show a significant relationship with ROA, all are positive predictors. Furthermore, lobbying and leverage don’t show a significant relationship. Model 2 contains all the variables and test the moderating effect of environmental dynamism. As a result, model 2 shows a significant result, F(6,419) = 13.74, p<.01. Individually, dynamism (β = −0.30, p < .001) , firm size (β = 0.20, p < .0001) and slack (β = 0.19, p < .001) showed a significant beta. With dynamism decreasing ROA and the other significant predictors increasing ROA. Moreover, the R2_{change, interaction term between} lobbying expenditures and dynamism, lobbying expenditures and leverage didn’t show a significant beta, so according to this regression no significant moderation effect occurred. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 1.93). So no causes for concerns of independent error (Field, 2009).

(28)

28

Table 7. The moderating effect of dynamism on the PAC contributions – ROA relationship

ROA Adjusted R2 R2_change Model 1 0.18** Beta SE Model 2 0.18 0.00 Beta SE ZPAC contributions (X) -0.01 0.00 -0.01 0.00 ZDynamism (M) -0.30** 0.02 -0.31** 0.02 ZFirm Size 0.21** 0.01 0.21** 0.01 ZLeverage -0.02 0.00 -0.02 0.00 ZSlack 0.20** 0.00 0.20** 0.00 Moderator (X*M) -0.01 0.01

Note: ROA, return on assets; PAC, political action committees *Correlation is significant at p<0.05

Table 7 includes the hierarchical regression including two separate models. First of all, model 1 contains all the variables except the PAC contributions - dynamism interaction term. This models shows a significant result, F(5,428) = 19,82 p<.01. Moreover, Model 1 has an adjusted R2_{of .18 which} indicates that the 18% of the variance in the return of assets could be explained by the model. Individually dynamism (β = −0.30, p < .001), firm size (β = 0.21, p < .001) and slack (β = 0.20, p < .001) show a significant relationship with ROA, all are positive predictors. Furthermore, PAC and leverage don’t show a significant relationship. Model 2 contains all the variables and test the moderating effect of environmental dynamism. As a result, model 2 shows a significant result, F(6,427) = 16.48, p<.01. Individually, dynamism (β = −0.31, p < .001) , firm size (β = 0.21, p < .000) and slack (β = 0.20, p < .000) showed a significant beta. With dynamism decreasing ROA and the other significant predictors increasing ROA. Moreover, the R2_{change, interaction term between PAC} and dynamism, PAC contributions and leverage didn’t show a significant beta, so according to this regression no significant moderation effect occurred. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 1.94). So no causes for concerns of independent error (Field, 2009).

(29)

29 The moderating effect of government dependence on the CPA – FP relationship

Table 8 examines the moderating effect of government dependence and it includes the standardized value of government dependence, lobbying expenditures, all the control variables and the interaction term between lobbying expenditures and government dependence.

Table 8. The moderating effect of government dependence on the lobbying expenditures – ROA relationship ROA Adjusted R2 R2_change Model 1 0.09** Beta SE Model 2 0.08 0.00 Beta SE ZLobbying expenditures(X) 0.01 0.00 0.02 0.00 ZGovernment dependence (M) 0.00 0.01 0.01 0.01 ZFirm Size 0.27** 0.01 0.27** 0.01 ZLeverage 0.01 0.00 0.01 0.00 ZSlack 0.20** 0.00 0.17** 0.00 Moderator (X*M) -0.03 0.00

Note: ROA, return on assets

*Correlation is significant at p<0.05 **Correlation is significant at p<0.001

Table 8 includes the hierarchical regression including two separate models. First of all, model 1 contains all the variables except the lobbying expenditures – government dependence interaction term. This models shows a significant result, F(5,472) = 9.927 p<.01. Moreover, Model 1 has an adjusted R2_{of .085 which indicates that the 8.5% of the variance in the return of assets could be} explained by the model. Individually firm size (β = 0.27, p < .001) and slack (β = 0.20, p < .001) show a significant relationship with ROA, all are positive predictors. Furthermore, lobbying expenditures, government dependence and leverage don’t show a significant relationship. Model 2 contains all the

(30)

30 variables and test the moderating effect of government dependence. As a result, model 2 shows a significant result, F(6,471) = 8.211, p<.01. Individually, firm size (β = 0.27, p < .001) and slack (β = 0.17, p < .001) showed a significant beta. Both increasing ROA. Moreover, the R2_change, interaction term between lobbying expenditures and government dependence, lobbying expenditures, government dependence and leverage didn’t show a significant beta, so according to this regression no significant moderation effect occurred. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 2.07). So no causes for concerns of independent error (Field, 2009).

Table 9 examines the moderating effect of government dependence and includes the standardized value of government dependence, PAC, all the other variables and the interaction term between PAC and government dependence.

Table 9. The moderating effect of government dependence on the PAC contributions – ROA relationship ROA Adjusted R2 R2 _change Model 1 0.08** Beta SE Model 2 0.08 0.00 Beta SE ZPAC contributions(X) -0.12 0.00 -0.20 0.00 ZGovernment dependence (M) 0.00 0.01 0.00 0.01 ZFirm Size 0.27** 0.01 0.28** 0.01 ZLeverage 0.01 0.00 0.01 0.00 ZSlack 0.17** 0.00 0.17** 0.00 Moderator (X*M) 0.01 0.00

Note: ROA, return on assets; PAC, political action committees *Correlation is significant at p<0.05

Table 9 includes the hierarchical regression including two separate models. First of all, model 1 contains all the variables except the PAC contributions – government dependence interaction term.

(31)

31 This models shows a significant result, F(5,473) = 9.81 p<.01. Moreover, Model 1 has an adjusted R2 of .08 which indicates that the 8% of the variance in the return of assets could be explained by the model. Individually firm size (β = 0.27, p < .000) and slack (β = 0.17, p < .000) show a significant relationship with ROA, all are positive predictors. Furthermore, PAC contributions, government dependence and leverage don’t show a significant relationship. Model 2 contains all the variables and test the moderating effect of government dependence. As a result, model 2 shows a significant result, F(6,472) = 8.17, p<.01. Individually, firm size (β = 0.28, p <.000) and slack (β = 0.17, p < .000) showed a significant beta. Both increasing ROA. Moreover, the R2_{change, interaction term between} PAC contributions and government dependence, PAC contributions, government dependence and leverage didn’t show a significant beta, so according to this regression no significant moderation effect occurred. Finally, the scatter plot seem to show a normal distribution and the assumption of independent error is accepted (Dublin-Watson = 2.07). So no causes for concerns of independent error (Field, 2009).

4. Discussion

4.1 Findings

Previous literature has attempted to clarify the relationship between CPA and financial performance for multiple years. To analyze this relationship studies have taken into account multiple theories such as the resource-based view, agency theory, institutional theory, stakeholder theory and the resource dependence theory (Mellahi et al, 2016). All these theories examined CPA from a different theoretical point of view, and as a result it is hard to draw an overall conclusion at this moment.

According to Melahi, et al (2016), the RDT has been used sixteen times within their study on non-market activities from 2000 to 2014. As a result, nine studies found a significant positive relationship. This research also examines CPA and financial performance based on the RDT assumptions. The RDT suggests that organizations will use political activities to position themselves in the environment which is in their personal interests (Pfeffer & Salancik, 1978). According to the RDT, firms that use political activities are driven by the tendency to reduce the external uncertainties that they face (Pfeffer & Salancik, 1978). Building on these assumptions, I studied the roles of government dependency and environmental uncertainty in relation to the CPA – FP relationship. Our first hypothesis indicated that CPA has a positive effect on financial performance. However, our results are contradictory to most of the previous studies and didn’t show a significant relationship. According to Mellahi, et al (2016), research between 2000 and 2014 suggests that only

(32)

32 two of the forty non-market studies concerning CPA concluded that there was an insignificant relationship and thus our results are contradictory to past studies. One possible reason for this positive effect can be the difference in period of time. Since this research is conducted within a four year time frame, the longitudinal effects of CPA can have an effect on the results (Rehbein & Schuler, 1999). Multiple other studies did their study within a timespan of ten years or more (Aggarwal, Meschke & Wang, 2012; Bonardi, Holburn & Vanden Bergh, 2006; Rehbein & Schuler, 1999). Furthermore, the moment in time could arguably have resulted in the contradictory results. More recent studies within non-market strategies (2010-2014) have displayed more cases of mixed, insignificant and negative relationships compared to the more early studies (Mellahi, et al, 2016). This could imply that nowadays political activities are less effective as they were in the past. Next to that, the agency theory could explain the insignificant relationship. This theory states that the only reason for managers to participate in CPA is because of their personal interests and thus they don’t act in the firms best interest (Eisenhardt, 1989). For instance, managers participate in CPA for personal reasons such as job security (Eisenhardt, 1989). Another theory that might explain the insignificant results is the marketplace theory perspective. This theory assumes that there are multiple firms competing for a limited amount of legislators. Consequently, the participation of most firms in CPA is rather expensive then beneficial (Francione, 1987)

Several studies indicated that the external environment has a significant impact on the managers decisions concerning the firm’s behavior and the competitive situation in the future (Milliken, 1987). As a result, environments could have different effects on the effectiveness of political activities and thus on the firm performance. In this study I divided environmental uncertainty into two environmental dimensions; munificence and dynamism (Dess & Beard, 1984).

First of all, based on the RDT perspective our hypothesis indicated that environmental munificence has a positive moderating effect on the CPA – FP relationship. Consequently, I found that there was a small significant relationship between environmental munificence and the lobbying expenditures – FP relationship. Moreover, I also found an insignificant relationship for the munificence and the PAC – FP relationship. However, contradictory to our hypothesis the moderation between munificence and the lobbying expenditures – FP relationship was negative. One possible reason for the negative moderation could be found in the marketplace theory perspective. The marketplace theory perspective indicates that there are multiple firms competing for just a limited amount of policy makers (Hadani & schuler, 2013). Therefore, firms within high munificent markets have additional (monetary) resources which they use to invest in lobbying activities. Consequently, as more firms invest in lobbying, the individual returns on CPA for each firm decreases as there are less policy makers available for each firm (Sun & Price, 2016).

(33)

33 Secondly, I hypothesized that environmental dynamism would have a positive moderating effect on the CPA – FP relationship. Results show an insignificant relationship for dynamism on lobbying expenditures and also an insignificant relationship on PAC and financial performance. A reason for this insignificant relationship could be that within dynamic environments the predictability of external events become more difficult compared to stable environments (Aldrich, 1979). Therefore, firms don’t have an explicit idea on how to shape their external environment since it is difficult to predict where the environment is moving toward (Aldrich, 1979). As a result, managers within dynamic markets have a difficult time to fully utilize the firm’s political activities.

According to Hillman and Hitt (1999), government dependent firms partly rely on purchases done by the government that have a direct effect on their sales. Consequently, political activities for government dependent firms can not only positively affect policies that indirectly affect the firm, but also policies that directly affect sales. Government dependent firms expect to be more influenced by policies compared to firms that have a lower government dependency. Therefore our final hypothesis stated that government dependency would have a positive moderating effect on the relationship between CPA and financial performance. However, our results reveal that there was no significant relationship between either government dependency and lobbying expenditures nor PAC contributions. These results are contradictory to most previous studies that found a significant relationship between government dependency and political activities (Meznar & Nigh, 1995; Birnbaum, 1985). One possible explanation for the insignificant results is that there are other specific firm factors that might affect the CPA – FP relationship (Marquis & Qian, 2013). All firms have different specific characteristics such as their position, reputation and mission and as a consequence some firms are more in need of policy regulations compared to others.

4.2 Limitations and future research

Within this study no relationship between CPA and financial performance was found. Consequently, studies that will be conducted in the future should consider other possible factors that might affect this relationship. As stated before, according to Hillman (2009), multiple external conditions and study limitations affect the CPA – FP relationship results. These external conditions could explain the mixed evidence that has been found within the CPA – FP studies in previous literature (Mellahi, 2016). This study also has several limitations that might affect the outcomes. Due to the limited time and resources of this research some limitation must be mentioned Furthermore, within this study

The influence of corporate political activities on financial performance : a few new insights