CORPORATE LOBBYING AND FIRM PERFORMANCE IN EUROPE. Confirmation of positive firm performance for European firms lobbying in the EU.

(1)

CORPORATE LOBBYING AND FIRM PERFORMANCE IN EUROPE.

Confirmation of positive firm performance for European firms lobbying in the EU.

by

Carl-Cyril J. Dreue

University of Groningen

Faculty of Economics and Business MSc. International Business and Management

First supervisor: dr. A.A.J. van Hoorn Second supervisor: dr. A.N. Kiss

(2)

2

Abstract

This study investigates whether Corporate Political Activity (CPA) through lobbying has an influence on firm performance measured as its Tobin’s Q. The results show that there is a significant positive effect. There is also a strong link between EU funds/grants received and the amount of lobbying expenditure. Ad hoc analysis shows that firms in Politically Sensitive Industries (mostly pharmaceutical firms) actually have a negative relationship between firm performance and CPA. Some avenues for future research are also suggested, such as doing a panel analysis and using several other variables not available at this time.

Keywords:

(3)

3

It’s a widely held view that companies influence politics by their financial contributions. Academic studies, however, have yet to reach a (strong) consensus in how much this benefits firms. Some believe that lobbying activities by Multination Enterprises (MNEs) are not necessarily aimed to gain direct financial support but attempt to gain influence which will ultimately subsidize the firms’ business activities. (Alt and Chrystal, 1983; Boddewyn and Brewer, 1994).

Researching lobbying provides important information for the international business &

management community and until recently many empirical studies have been performed on the incentives to lobby and on Political Action Committee (PAC)1 contributions in the United States. While these studies are interesting, PAC contributions only account for five percent of total expenditure by firms on influencing politics (Chen, Parsley and Yang, 2010). Lux, Crook, & Woehr (2011) performed a meta-analysis of the antecedents and outcomes of Corporate Political Activity and found ‘that firm performance is positively related and is an important determinant to Corporate Political Activity (CPA ). A study done by Chen, Parsley and Yang (2010) found similar results using data from the Lobbying Disclosure Act of 1995. In Europe the need for transparency in lobbying was identified, and resulted in the Transparency Register. Here companies can disclose information on their lobbying activities on a voluntary basis. The European Commission Vice-President Maroš Šefčovič mentions that organizations that have nothing to hide will register and those that do not will be asked why they can’t be transparent.

Another reason why this register has a lot of value pointed out by EurActiv, an independent media portal fully dedicated to EU affairs, is that although ‘Registration in the joint system is

(5)

5

voluntary, all lobbyists who wish to enter the Parliament’s premises will have to register, giving them a strong incentive to do so.2 For Europe, studies on the effectiveness of lobbying have not been undertaken as of this moment. In this paper I will examine the effects of Corporate Political Activity on firm performance for European firms3. I believe that this study will add knowledge on the effects of lobbying on firm performance in the European Union to the academic field. This paper confirms that lobbying does indeed increase firm performance. The results show that firms that lobby do indeed have a higher Tobin’s Q. Additionally, I find that lobbying expenditure is also positively related with firm performance. The effects of lobbying expenditure and the amount of funds/grants (this includes governmental contracts and subsidies) received was also found to be significant, but had a relatively low explanatory value.

There are however some industries in which lobbying does not appear to be lucrative. Lobbying seems to have an inverse effect on firms that are active in Politically Sensitive Industries.

This paper has the following structure. In section 2 theory on the relation between firm performance and lobbying will be discussed. In section 3 the hypotheses and a conceptual model summarizing the relations between the several variables researched are presented. In section 4 data collection and research methodology are described. In section 5 the results will be provided and section 6 concludes with a discussion of the results and limitations and future research directions are suggested.

2_{Source: http://www.euractiv.com/pa/council-ready-join-eu-transparency-register-news-505920} 3

(6)

6 2. Theoretical background

In this section the theoretical background is briefly discussed.

2.1. Theory

Corporate Political Activity (CPA) is defined as ‘corporate attempts to shape government policy by lobbying in ways favorable to the firm’ (adapted from Baysinger, 1984). According to Baysinger (1984) CPA can have three major objectives: 1) to gain special monetary and anticompetitive favors from government-domain management, 2) to manage environmental turbulence created by governmental threats to the legitimacy of organizational goals and purposes-domain defense, and 3) to manage similar threats to the methods by which organizations pursue their goals and purposes-domain maintenance. In this paper the relationship between CPA and firm performance is investigated. In its core most scholars agree that firms engage in CPA to improve performance (Mitchel et al., 1997; Lux et al., 2011; Hillman et al., 2004). CPA affects firm performance when political activities influence the government to take action (or not take action) in a manner that benefits a firm. Obtaining rate increases (Bonardi et

al. 2006), earmarks (An earmark is a legislative (especially congressional) provision that directs

approved funds to be spent on specific projects, or that directs specific exemptions from taxes or mandated fees4) (De Figueiredo & Silverman, 2006), new import tariffs/protectionism (Schuler, 1996), and government sales (Epstein, 1969) are just a few of the many ways policy changes affect firm performance.

Hill et. al. (2011) have found that lobbying firms significantly outperform non-lobbying firms. Furthermore, in a recent paper Lux et. al. (2011) have done a meta-analysis in which they

4

(7)

7

researched 1) what factors and to what extent those factors influence firms to engage in CPA, and 2) if CPA does, in turn, affect firm performance and, if so, to what extent. They found that many factors shape CPA, but very few affect CPA to a large extent.

It is clear that Corporate Political Activity has a large influence on (international) business and is a subject worthwhile investigating. However, these papers focused on lobbying in the United States and have not yet been replicated in Europe. As the governmental structure of the European Union differs immensely from the United States this paper might provide some valuable insights in this area.

Firms lobby for a variety of reasons, mostly to obtain and/or maintain economic returns (e.g., North, 1990; Lux et. al., 2010). Bergström (2000) wrote a paper on whether capital subsidies to firms in Sweden have an effect on the total factor productivity. He finds that subsidization can influence growth, but he finds little evidence that the subsidies affect productivity. Due to the local focus of Bergström his research and the fact that this study is focused on firm performance and not growth or productivity and that firm growth still has an impact on firm performance. It seems reasonable to believe that when firms do receive grants and/or procurements it follows that these funds contribute to firm performance.

(8)

8

average return to one dollar of lobbying of $11-$17 and lobbying universities with representation on the House Appropriations Committee obtain $20-$36 for each dollar spent.

The Transparency Register keeps records of many aspects of firm lobbying, such as: contact information of people involved, fields of interest to the company, goals/remit of the company, and financial data. The Transparency Register defines procurements (or public contracts); 'purchases by a public authority of a service, good, or works' and grants as; 'direct financial

contributions from the EU budget awarded by way of donation to third-party beneficiaries (usually non-profit-making organizations) engaged in activities that serve Union policies5) The variables grants and procurements received are also registered in the Transparency Register. This provides a unique opportunity to research the returns on lobbying.

Furthermore if a firm has an office in Brussels it can be assumed that a firm has easier access to parliamentarians and that their lobbying activities will be more efficient/productive. External parties can be assumed to have greater knowledge on how to lobby more efficiently and an industry association will most likely be helpful to pool knowledge and achieve industry wide benefits to firms in that industry. Although for the last variable freeloading might be an issue. Finally, when a firm receives funding and/or grants one can assume that this also has a positive effect on firm performance.

In the next chapter hypotheses are formulated based on the theories discussed in this chapter.

(9)

9 3. Hypothesis development

In this chapter the hypotheses are formulated and a conceptual model with an overview of all the relations between the variables is presented.

3.1. Hypotheses

As discussed in the theory section, there has been much evidence that CPA has a positive influence on firm performance. The main focus of this paper is to investigate whether lobbying in the European Union has a positive effect on firm performance. Therefore I propose the following hypotheses.

Hypothesis 1a:

Lobbying in the European Union is positively related to firm performance.

And,

Hypothesis 1b:

Higher lobbying expenditures by firms in the European Union are positively related to better firm performance (i.e., firms with higher levels of lobbying expenditures will exhibit better performance).

When it comes to receiving grants and/or procurements the theory suggests that there are positive returns for CPA and presumably also on firm performance. Therefore I propose the following hypothesis.

Hypothesis 2:

(10)

10

The paper by Figueiredo, J.M. & Silverman, B.S. (2002) has shown that lobbying does indeed have a positive return on investment. This has led me to formulate the following hypothesis.

Hypothesis 3:

Higher lobbying expenditures are positively related to the amount of grants and/or procurements received.

Lastly, there are many variables available in the Transparency Register than can provide a better understanding of the factors that affect firm performance. I will investigate whether there are any moderating effects of the variables used in this paper and firm performance. A moderator is a qualitative or quantitative variable that affects the direction and/or strength of the relation between an independent or predictor variable and a dependent or criterion variable. (Baron and Kenny, 1986).

By analyzing these relationships a better understanding of the interaction of certain variables can be created. Which in turn can be used by international businesses to improve their lobbying efforts. The following hypotheses are formulated based on the variables available from the Transparency Register:

Hypothesis 4 a:

(11)

11 Hypothesis 4 b:

Having received funding(grants and/or procurements) from Brussels will have a positive moderating effect on firm performance.

Hypothesis 4 c:

Having a lobbying office in Brussels will have a positive moderating effect on firm performance.

Hypothesis 4 d:

Being a member of a lobbying trade organization will have a positive moderating effect on firm performance.

Hypothesis 4 e:

Having hired an external party for lobbying activities will have a positive moderating effect on firm performance.

(12)

12

3.2 Conceptual model

Figure 1 summarizes all the relations of the variables used and the hypotheses in a conceptual model.

firm lobbying firm lobbying expenditure

Receiving EU grants and/or procurements Firm performance H1a + H1b H3 H2 Moderators - Industry 1 through 4 - Funding received dummy* - Ln funding received* - Office in Brussels dummy

- External party lobbying (trade organizations) - External party lobbying (hired lobbyists)

Figure 1: conceptual model of the relations between variables and the hypotheses analyzed

(13)

13 4. Data collection and research method

The following sections describe the data collection methods and research methodology.

4.1. Data collection

As this study is focused on lobbying in Europe a large sample of European firms is taken.

Because larger firms are more likely to lobby (Figueiredo & Tiller, 2001) the Financial Times list of the 500 largest European firms by firm value is used. A sample of the top 200 firms is taken excluding financial firms due to their offset balance sheets. A full list of the amount of firms from each country can be found in the appendix in table A1.

The European transparency register6 has a large amount of information surrounding lobbying activities of firms. It can be hypothesized that many of the variables have an influence on the lobbying success of firms. The variables that will be used in this paper are: whether a firm has an office in Brussels, if a firm hires an external party to do its lobbying and if a firm is associated with an industry association that lobbies.

For this paper the firm lobbying data is collected from the EU Transparency Register. This register started on the 23rd of June 2011 and as of the 23rd June 2012 has 5,175 organizations registered of which 2,467 in-house lobbyists and trade/professional associations7. The registration in the Transparency Register is voluntary, but to get access to the European Parliament individuals need to register8. This provides an incentive for firms to register and enhances credibility and accuracy of the register. The informationin which industry a firm is active in was collected from ORBIS.

6_{Source: http://europa.eu/transparency-register/} 7

(14)

14

This paper excludes financial firms like banks, insurance firms, etc. due to the large amount of incomplete data of firms in these sectors and rather different balance sheets compared to the other firms in the dataset.

The Financial Times has acquired a reputation as a reliable source and therefore I believe that the information gathered from the Financial Times 500 Europe list is adequate for this paper. The information on the industries where a firm is active in was collected from Orbis which is as well a reliable source that is widely being used by researchers and analysts (OECD, 2010). Finally the lobbying data has some limitations as companies self-report their activities and there is no legal requirement to supply lobbying information unless a lobbyist wants access to the parliament. But, as I explained in the introduction, I believe that the sample has a large enough size and there is enough incentive to provide the lobbying information that it is reliable and valid.

In model 4 a new interaction predictor is calculated. This can lead to the new interaction term to correlate with the two main effects terms used to calculate it. To combat this problem I center the independent variables with the deviation from their means. (Kreft & de Leeuw, 1998)

4.2. Measures

Independent variable

(15)

15

The estimated costs of the organization that are directly related to representing interests in EU institutions of that year were collected as well. This information was provided in €50,000 increments, but as organizational expenditure varied between €0 and €9,000,000 this was not deemed to impact the results.

Dependent variable

Following many studies like Lang & Stulz (1994) and Gande et. al. (2009) I use Tobin’s Q as the measure of firm performance. Tobin's Q is measured as the book value of assets divided by the end-of-year market value of common stock (Lang & Stulz, 1994). The book value of assets and the end-of-year market value of a common stock were obtained from the Financial Times top 500 European companies list.

Other independent variables

Industry effects are combated by using industry dummies. The most used variable when

analyzing large samples of very different firms is an industry classification according to Fama and French (1994). Because the authors use 49 different industry groups it seemed that this would cause problems when analyzing. This is why I chose to use Standard Industrial Classification (SIC) codes as an industry differentiator. SIC codes are established by governmental agencies to classify industrial areas a firm is active in9. The core SIC code consists of two, three, or four digits. In this research I used the first digit as an industry classifier. The SIC codes for all the firms were gathered and five broad industries were identified based on the first digit of the three

(16)

16

digit core industry SIC code. These five industries are: 1) oil & gas producers, oil equipment & services, and mining, 2) pharmaceuticals & biotechnology, food producers, chemicals, tobacco, and beverages, 3) construction & materials, healthcare equipment, industrial engineering, automobile & parts, and aerospace & defense, 4) electricity, telecommunications, gas, and water & multi-utilities, and finally 5) other.

Firm size has been classified as an indicator for the likelihood of CPA (Hillman et. al, 2004), but

also influences profitability (Hall & Weiss, 1967). Furthermore, its used as a control variable to eliminate any effects solely due to firm size (Gande et. al., 2009). This variable is measured as the natural logarithm of a firms total assets provided by the Financial Times 500 Europe list. Finally Tobin’s Qt-1 is a variable used as to offset past firm performance.

4.3 Models

In the following section the models used in this paper will be described.

Model 1a:

Yt = α + β1 Lobbying dummyt + β2 Tobin’s Qt-1 + β3 Ln assetst

+ β4 industry dummies + u

Where Yt is firm performance (measured as its Tobin’s Q) at year t, α is the constant, β1 is a

dummy variable where a value of 1 will be given to firms that lobby in year t and 0 to firms that do not at time t, β2 is a control variable measuring a firm’s past Tobin’s Q at t-1, β3 is another

control variable measured as the Ln (natural logarithm) of a firm’s assets, β4 is the final control

(17)

17

Model 1b:

Yt = α + β1 Ln Lobbying expenditurest + β2 Tobin’s Qt-1 + β3

Ln assetst +β4 industry dummies + u

Where Yt is firm performance (measured as its Tobin’s Q) at year t, α is the constant, β1 is the Ln

(natural logarithm) of the actual amount of lobbying expenditures at time t, β2 is a control

variable measuring a firm’s past Tobin’s Q at t-1, β3 is another control variable measured as the

Ln (natural logarithm) of a firms assets, β4 is the final control variable with industry dummy’s

based on SIC codes, and u is the error term.

Model 2:

Yt = α + β1 Funding received dummyt + β2 Tobin’s Qt-1 + β3 Ln

assetst + β4 industry dummies + u

Where Yt is firm performance (measured as its Tobin’s Q) at year t, α is the constant, β1 is a

dummy variable where a value of 1 will be given to firms that received European funding at time

t and 0 if not, β2 is a control variable measuring a firm’s past Tobin’s Q at t-1, β3 is another

control variables measured as the Ln (natural logarithm) of a firms assets, β4 is the final control

(18)

18

Model 3a:

Yt grants = α + β1 Ln Lobbying expenditurest + u

Where Yt grants is the Ln (natural logarithm) of grants received at time t, α is the constant, β1 is the

Ln (natural logarithm) of funds spent on lobbying, and u is the error term.

Model 3b:

Yt procurements = α + β1 Ln Lobbying expenditurest + u

Where Yt procurements is the Ln (natural logarithm) of funds received at time t, α is the constant, β1 is

the Ln (natural logarithm) of funds spent on lobbying, and u is the error term.

Model 3c:

Yt total procurements and grants = α + β1 Ln Lobbying expenditurest + u

Where Yt total procurements and grants is the Ln (natural logarithm) of the total amount of funds and

grants received at time t, α is the constant, β1 is the Ln (natural logarithm) of funds spent on

(19)

19

4.4. Moderation

In this paper the model by Baron and Kenny (1986) depicted in figure 2 will be followed. To investigate the moderating effects by the variables mentioned in table 1.

Figure 2. Moderator model

The model diagrammed in Figure 2 has three causal paths that feed into the outcome variable: whether a firm lobbies as a predictor (Path a), several variables like whether a firm has an office in Brussels, is a member of industry association (for more see table 1) (Path b), and the interaction or product of these two (Path c). The moderator hypothesis is supported if the interaction (Path c) is significant. There may also be significant main effects for the predictor and the moderator (Paths a and b), but these are not directly relevant conceptually to testing the moderator hypothesis. (Baron and Kenny, 1986)

(20)

20

To combat any multicollinearity the variables used will be centered minus grand mean from independent variable scores (Kreft & de Leeuw, 1998)

The analysis of the moderating relationships is performed according to the following models:

Model 4a:

Yt = α + β1 Lobbying dummyt + β2 Tobin’s Qt-1 + β3 Ln assetst

+β4 industry dummies + β5 moderating variable + β6 (moderating

variable * Ln Lobbying expenditurest) + u

Where Yt is firm performance (measured as its Tobin’s Q) at year t, α is the constant, β1 is the

lobbying dummy, β2 is a control variable measuring a firm’s past Tobin’s Q at t-1, β3 is another

control variables measured as the Ln (natural logarithm) of a firms assets, β4 is a control variable

with industry dummy’s based on SIC codes, β5 is the moderating variables that are dummies (a

full list of variables used can be found in Table 1), β6 is the moderating variable multiplied by the

lobbying dummy, and u is the error term.

Model 4b:

Yt = α + β1 Ln Lobbying expenditurest + β2 Tobin’s Qt-1 + β3

Ln assetst + β4 industry dummies + β5 moderating variable + β6

(moderating variable * Ln Lobbying expenditurest) + u

Where Yt is firm performance (measured as its Tobin’s Q) at year t, α is the constant, β1 lobbying

(21)

21

variables measured as the Ln (natural logarithm) of a firms assets, β4 is a control variable with

industry dummy’s based on SIC codes, β5 is the moderating variable (a full list of variables used

can be found in table 1), β6 is the moderating variable that is a dollar amount multiplied by the Ln

(natural logarithm) of the actual amount of lobbying expenditures at time t, and u is the error term.

Table 1. list of moderating variables used Moderating variables used

- Industry 1 through 4 - Funding received dummy* - Ln funding received* - Office in Brussels dummy

- External party lobbying (trade organizations) - External party lobbying (hired lobbyists)

Note: * in these cases funding received, grants received and both combined were analyzed.

4.5 To summarize

(22)

22 5. Results

In this section first the descriptive statistics are given. Second the main results are presented. Third the hypothesized effects are covered. Subsequently some additional analysis are provided. Finally results from a post-hoc analysis are discussed.

5.1. Descriptive statistics

(23)

23

Table 2. Descriptive statistics

Variables Mean Variables Total amount

Tobin’s Q 1.27 Firms 148

Value of assets €59 billion Firms lobbying 97

Ln of assets 12.1 Firms in industry 1 28

Lobbying expenditure €796,629 Firms in industry 2 37

Grants received* €10 million Firms in industry 3 38

Procurements received* €40 million Firms in industry 4 37 Firms in industry 5 8 Office in Brussels 58 Member of industry

association

43 Hired an external party for lobbying 53 Political Sensitive Industry 19 Grants receive 22 Procurements received 9

NOTE: * The mean for these variables are only for firms that have received a grant and/or procurement. The mean is this high due to two firms receiving procurements of over €100 million.

5.2. Results

(24)

24 Table 3. Correlation 1 2 3 4 1 Tobin’s Qt 2 Tobin’s Qt-1 0.934** 3 Ln of assetst -0.669** 0.634**

4 Does the firm lobby dummy -0.267** -0.346** 0.315**

Mean 1.26 1.08 10.41 0.63

Std. deviation 1.27 1.05 1.34 0.48

NOTE: ** Correlation is significant at the 0.01 level (2-tailed), * Correlation is significant at the 0.05 level (2-tailed).

Table 4 shows the results from the regression analyses. Model 0 provides the baseline for my analyses. Models 1 through 4 present the results for hypotheses 1 through 4.

The main model (Model 0) has an R squared of 0.884 and thus represents 88% of the variance in the Tobin’s Q of a company. The control variables were tested and found to be significant at the 1% level.

5.3. Main hypothesized effects

Hypothesis 1a states that lobbying will have a positive effect on firm performance. The results show that firms that lobby do indeed have a higher Tobin’s Q with a relatively high coefficient of 0.256 at a significance level of 1%.

Hypothesis 1b states that higher lobbying expenditures will lead to a higher Tobin’s Q. The data shows that this is indeed the case.

(25)

25

Table 4. Regression models H1 through H3 Regression of firm performance on lobbying

Covariates Model 0i Model 0 Model 1a Model 1b Model 2

Lobby dummy 0.267*** (0.087) Ln of Lobbying expenditures 0.022*** (0.007) Funding received dummy 0.129 (0.114) Tobin’s Qt-1 1.045*** (0.043) 1.028*** (0.048) 1.035*** (0.047) 1.029*** (0.046) 1.027*** (0.049) Ln assetst -0.214*** (0.046) -0.168*** -0.223*** (0.049) -0.240*** (0.051) -0.189*** (0.050) Industry dummies included?

no yes yes yes yes

Constant 2.216*** (0.494) 1.786*** (0.532) 2.249*** (0.538) 2.421*** (0.554) 1.931*** (0.559) Observations 148 148 148 148 142 R squared 0.891 0.884 0.892 0.892 0.886 Adjusted R Squared 0.889 0.880 0.886 0.887 0.880 F stat 394.970 181.189 165.947 166.494 149.879 Prob F stat 0.000 0.000 0.000 0.000 0.000 Mean VIF 1.620 2.634 2.534 2.583 2.527

(26)

26

Table 5. Regression model H3a through H3c Regression of firm performance on lobbying

Covariates Model 3a Model 3b Model 3c

Ln of Lobbying expenditures 0.298*** (0.063) 0.096** (0.042) 0.302*** (0.065) Constant -0.259 (0.636) -0.034 (0.424) -0.241 (0.651) Observations 153 153 153 R squared 0.128 0.033 0.126 Adjusted R squared 0.123 0.027 0.120 F stat 22.406 5.264 21.595 Prob F stat 0.000 0.023 0.000

Note: standard error in parentheses: * significant at 10%, ** significant at 5%, *** significant at 1%.

Table 5 shows the results for Model 3a through 3c. These models try to find whether there is a connection between lobbying expenditure and the amount of funding/procurements gained. The results show that the amount of grants a firm receives is statistically significantly related to a firms lobbying expenditure with a coefficient of 0.298 at 1% significance. A relationship between procurements and lobbying expenditure was also found with a coefficient of 0.096 at the 5% level. Both grants and procurements result in a statistically significant relationship at the 1% level with a coefficient of 0.302. The adjusted R square ranges from 2.7% to 12.3% and thus has a relatively low explanatory value.

5.4. Lobbying moderators hypotheses

(27)

27

meaningful results. Of these four, mining, construction, and utilities did not seem to have a significant effect. Chemicals however did show a significant effect with a coefficient of

-0.403 at the 5% level. This means that firms that are active in these industries and lobby have lower firm performance than firms that do not. For a full list of all results for the moderating variables see Table A2 and A3 in the appendix.

Having an office in Brussels for lobbying did not reveal any useful results as the lobbying variable was excluded by SPSS. Hiring an external party to do a firms lobbying and/or being a member of an industry association has a positive effect of 0.720 at the 5% level. However when I analyzed the individual variables (hiring a firm to lobby and being a member of an industry association that lobbies for a firm) there were no significant results. Further research needs to be done to investigate these results.

Furthermore, whether a firm received a grant or procurement did not moderate the results, nor for the dummy variables nor when Ln of actual amounts were used. For a full list of the results for moderation of procurements and grants received see Table A3 in the appendix.

5.5 Post hoc analyses

(28)

28

to prevent pharmaceutical patents to expire. I tried to narrow down what could be the cause of this and used the FDA listing of significantly regulated industries.10 The firms in this list did not show an interaction effect. Other factors that might be of influence are the amount of patents a firm owns. As I did not have access to any database with this information it could be an avenue for future research.

(29)

29 6. Discussion

6.1. Key findings and implications

This paper confirms the findings of Lux et al. (2011) that lobbying does indeed increase firm performance. The results show that firms that lobby do indeed have a higher Tobin’s Q. Additionally, I find that firm expenditure is also positively related with firm performance. The effects of lobbying expenditure and the amount of funds/grants received was also found to be significant, but had a relatively low explanatory value.

My results, however, show that lobbying does not appear to be lucrative for all firms. Lobbying seems to have an inverse effect on firms that are active in Politically Sensitive Industries. This might be due to the fact that pharmaceutical firms unsuccessfully lobby to prevent patents from expiring. Moreover, because of strict rules and highly regulated markets these firms have had trouble maintaining their (high) prices as governments are cutting spending and demanding cheaper products due to the financial crisis (Whalen & Stovall, 2012)

6.2. Limitations and avenues for future research

This paper has confirmed that there is a significant and positive relationship between firm performance and lobbying expenditures as well as with lobbying expenditures and the amount of grants received.

(30)

30

2009) and be able to do a panel analysis like Chen et al. (2010). This could further confirm that lobbying in Europe does indeed increase firm performance.

Furthermore, due to their availability and/or lack of significance I was unable to include all variables desired. Future research may be able to obtain information such as: the percentage of foreign sales, EBIT / sales, capital expenditure / sales, R&D / sales, advertisement / sales, Liabilities / assets (Gande et. al., 2009) and industry concentration (Huselid et. al., 1997).

(31)

31 References

Baysinger, B.D., (1984), ‘Domain Maintenance as an Objective of Business Political Activity: An Expanded Typology’, Academy of Management Review, p.248-258.

Bergström, F., (2000), ‘Capital Subsidies and the Performance of Firms’, Small Business

Economics, p183-193.

Bonardi, J.P., Holburn, G.L.F. & Vanden Bergh, R.G., (2006), ‘Nonmarket strategy performance: Evidence from the U.S. Electric Utilities’, Academy of Management Journal, p.1209-1228 Chen H., D. Parsley & Y. Yang, (2010) , ‘Corporate Lobbying and Financial Performance’, working paper.

Cohen, J. P. et. al., (2003), ‘Applied multiple regression/correlation analysis for the behavioral sciences’ (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.

De Figueiredo, J.M. & Tiller, E.H., (2001), ‘The Structure and Conduct of Corporate Lobbying: How Firms Lobby the Federal Communications Commission’, Journal of Economics &

Management, p.91-122.

De Figueiredo, J.M. & Silverman B.S., (2006), ‘Academic Earmarks and the Returns to Lobbying’, Journal of Law and Economics, p.597-625.

Epstein, E., (1969), The corporation in American politics., Englewood Cliffs, NJ: Prentice/Hall. Fama, E.F. & French, K.R., (1997), ‘Industry cost of equity’, Journal of Financial Economics, p.153-193.

Gande, A., Schenzler, C. & Senbet, L.W., (2009), ‘Valuation Effects of Global Diversification’,

Journal of International Business Studies, p.1515-1532.

Hall, M. & Weiss, L., (1967), ‘Firm Size and Profitability’, The Review of Economics and

Statistics, p.319-331.

Hill, M.D. et.al., (2011), ‘Determinants and Effects of Corporate Lobbying’, SSRN abstract, Vol. 1420224.

Hillman, A.J., Keim, G.D. & Schuler, D., (2004), ‘Corporate Political Activity: A review and research agenda.’, Journal of Management, 30: p.837-858.

Huselid, M.A., Jackson, S.E. & Schuler, R.S., (1997), ‘Technical and Strategic Human Resource Management Effectiveness as Determinants of Firm Performance’, Academy of Management

(32)

32

Kutner, Nachtsheim & Neter, (2004), Applied Linear Regression Models, 4th edition, McGraw-Hill Irwin.

Kreft, I & de Leeuw, J, (1998), Introducing multilevel modeling, London: Sage Publications Ltd. p.106-114.

Lang, L.H.P. & Stulz, R.M., (1994), ‘Tobin’s Q, Corporate Diversification, and Firm Performance’, Journal of Political Economy, p.1248-1280.

Lux, S., Crook, T.R. & Woehr, D.J., (2011), ‘Mixing Business With Politics: A Meta-Analysis of the Antecedents and Outcomes of Corporate Political Activity’, Journal of Management, p.223-247.

MacKinnon, D.P., (2008), Introduction to Statistical Mediation Analysis, New York: Erdbaum. Mitchel, N.J., Hansen, W.L. & Jepsen, E.M., (1997), ‘The Determinants of Domestic and Foreign Corporate Political Activity’, The Journal of Politics, p.1096-1113.

North, D.C., (1990), Institutions, Institutional Change, and Economic Performance. Oxford, UK: Cambridge University Press.

Organisation for Economic Co-operation and Development (OECD), (2010), ‘The OECD Orbis database: responding to the need for firm-level micro-data in the OECD’, working paper.

Schuler, D.A., (1996), ‘Corporate political strategy and foreign competition: The case of the steel industry.’, Academy of Management Journal, p.720-737.

Whalen, J. & Stovall, S., (2012), ‘Drug Firms Urge Price Action: Steep Discounts in Five

(33)

33 Appendix

Table A1: company country of origin (total amount per country)

FT Europe 500

Country Amount of firms

(34)

34 Table A2

Regression of firm performance on lobbying. Moderation

Covariates Main model Industry 1 (mining) Industry 2 (chemicals) Industry 3 (construction) Industry 4 (utilities) Office in Brussels External party lobbying total Lobbying by industry association Lobbying by external party Politically sensitive industries Variable 0.147 (0.155) 0.102 (0.144) 0.048 (0.147) 0.091 (0.148) 0.213** (0.097) -0.192 (0.134) X 0.081 (0.177) 0.624*** (0.238) Interaction effects 0.191 (0.192) -0.403** (0.175) 0.174 (0.174) -0.075 (0.185) -0.850*** (0.264) 0.720** (0.330) 0.089 (0.251) -0.119 (0.453) -0.719*** (0.251) Lobbying dummy 0.267*** (0.087) 0.271*** (0.087) 0.273*** (0.085) 0.265*** (0.087) 0.266*** (0.087) X 0.510*** (0.159) 0.279*** (0.094) 0.211** (0.166) 0.334*** (0.089) Tobin’s Qt-1 1.035*** (0.047) 1.037*** (0.047) 1.037*** (0.046) 1.038*** (0.047) 1.038*** (0.047) 1.040*** (0.047) 1.037*** (0.046) 1.032*** (0.047) 1.035*** (0.047) 1.047*** (0.046) Ln assetst -0.223*** (0.049) -0.229*** (0.049) -0.227*** (0.048) -0.221*** (0.049) -0.222*** (0.049) -0.213*** (0.050) -0.212*** (0.048) -0.227*** (0.050) -0.224*** (0.049) -0.251*** (0.048) Industry dummies included?

Yes Yes Yes Yes Yes Yes Yes Yes Yes yes

Constant 1.286*** (0.035) 1.291 (0.036) 1.293*** (0.035) 1.284*** (0.036) 1.287*** (0.036) 1.404*** (0.052) 1.170*** (0.064) 1.277*** (0.043) 1.300*** (0.065) 1.285*** (0.035) Observations 148 148 148 148 148 148 148 148 148 148 R squared 0.892 0.893 0.896 0.893 0.892 0.893 0.895 0.892 0.892 0.898 Adjusted R2 0.886 0.886 0.890 0.886 0.886 0.887 0.889 0.886 0.885 0.891 F stat 165.947 145.315 150.287 145.332 144.365 145.534 132.139 144.320 127.513 135.959 Prob F stat 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Mean VIF 2.540 2.391 2.362 2.365 2.372 2.476 3.078 2.434 3.553 3.182

(35)

35 Table A3

Regression of firm performance on lobbying. Moderation

Covariates Main model Funding received dummy total Procurements received dummy Grants received dummy

Main model Ln Funding received total Ln Procurements received Ln Grants received Variable X X X -0.008 (0.034) -0.075 (0.110) -0.011 (0.032) Interaction effects 0.149 (0.290) 0.025 (0.443) 0.134 (0.299) 0.002 (0.006) 0.014 (0.020) 0.002 (0.006) Lobbying dummy 0.267*** (0.087) 0.280*** (0.090) 0.267*** (0.087) 0.277*** (0.090) Ln lobbying expenditure 0.022*** (0.007) 0.025** (0.014) 0.032* (0.017) 0.026* (0.014) Tobin’s Qt-1 1.035*** (0.047) 1.033*** (0.047) 1.035*** (0.047) 1.034*** (0.047) 1.029*** (0.046) 1.027*** (0.047) 1.026*** (0.047) 1.025*** (0.047) Ln assetst -0.223*** (0.049) -0.228*** (0.050) -0.224*** (0.049) -0.227*** (0.049) -0.240*** (0.051) -0.244*** (0.052) -0.245*** (0.051) -0246*** (0.052) Industry dummies included?

Yes YES YES YES YES YES YES YES

Constant 1.286*** (0.035) 1.278*** (0.039) 1.285*** (0.037) 1.279*** (0.039) 1.285*** (0.035) 1.265*** (0.078) 1.232*** (0.086) 1.256*** (0.074) Observations 148 148 148 148 148 148 148 148 R squared 0.892 0.892 0.892 0.892 0.892 0.892 0.892 0.892 Adjusted R2 0.886 0.886 0.886 0.886 0.887 0.885 0.885 0.885 F stat 165.947 144.477 144.178 144.406 166.494 127.788 128.134 127.988 Prob F stat 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Mean VIF 2.540 2.402 2.406 2.396 2.583 8.241 27.285 7.563

CORPORATE LOBBYING AND FIRM PERFORMANCE IN EUROPE. Confirmation of positive firm performance for European firms lobbying in the EU.

CORPORATE LOBBYING AND FIRM PERFORMANCE IN EUROPE.

Abstract

Table of contents