‘The effect of the gender quota in Germany’
Abstract
As of 1 January 2016 Germany requires publicly listed and co-determined firms to allocate at least 30 percent women on their board. This paper examines the effect of this women’s quota on firm performance using a sample of 86 listed German firms during the time period of January 2013 till December 2016. This paper finds that gender diversity has no effect on firm performance and does therefore not reject the first hypothesis. Furthermore, it finds that the firms that have achieved the 30 percent quota perform better than those that have not which rejects the second hypothesis. This research lacks external validity due to different legislations in different countries. Moreover, future research should expand the time frame of the research and the control variables and take into account reverse causality.
Author:
Laura Van Landeghem (10988599)
University of Amsterdam
Coordinator:
Mr P.R Stastra
University of Amsterdam
Bachelor thesis – Economics and Business (specialization: finance and organisation) Faculty of Economics and Business at the University of Amsterdam
Statement of originality
This document is written by Laura Van Landeghem 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.
Table of contents
1. Introduction ...4
2. Literature review ...6
2.1 Studies that found a positive effect on firm performance ...6
2.2 Studies that found a negative or no effect on firm performance ...7
2.3 Studies regarding gender equality quotas ...8
2.4 Hypotheses ...9 3. Methodology... 10 3.1 Data sample ... 10 3.2 Methodology ... 10 4. Results ... 14 4.1 Descriptive statistics... 14 4.2 Results ... 16
5. Conclusion and limitations ... 21
1. Introduction
Gender diversity in boardrooms gained considerable interest in the twenty-first century. Generally, women are under-represented at the top of the labour market often referred to as the glass-ceiling. In Europe, the national governments and the European Parliament are taking measures to shatter this glass-ceiling for women. Several European countries have implemented a gender equality quota. Norway was the first country to impose such a quota. It required public limited firms to have at least 40 percent women on their board. This was quota mandatory, since not
reaching the required percentage of women led to dissolvement of the firm. Since this quota was successfully implemented as the percentage women in boardrooms
increased, other European countries followed Norway’s example. In the years following different European countries implemented a mandatory gender equality quota such as Italy. There are countries that impose a non-binding gender quota ranging from 25% to 40% required women on a board such as the Netherlands.
The quota in Germany is a combination of the ones mentioned before. The quota they adopted is mandatory and enforces companies that have not reached 30 percent women on their board by the first of January 2016 to leave board seats empty. This thesis will focus on the gender equality quota in Germany. Specifically, it will investigate the following research question:
What is the effect of the women’s quota in Germany on firm performance?
The previous literature is ambiguous about the topic of how gender diversity (i.e. female representation) affects firm performance. For instance, Carter et al. (2003) and Campbell and Vera (2008) found that higher female representation leads to significantly higher firm performance. Additionally, Adams and Ferreira (2009) show that increased gender diversity has a negative effect on firm performance. This ambiguous results can be due to the different estimation methods and the different data samples used (Campbell & Vera, 2008). Lastly, the results from the mandated quota in Norway show that a forced gender quota has a negative effect on firm performance, even though the number of female in boards increased (Ahern & Dittmar, 2012), (Dale-Olsen et al., 2013).
In line with the expectations, the evidence shows that increased gender
diversity has no effect on firm performance. Therefore, the first hypothesis will not be rejected. The other hypothesis stating that firms that have reached the quota do not
perform better than those that have not is rejected. This means that firms that have achieved the mandated quota do perform better compared to those that have not reached it. Therefore, this paper concludes that the women’s quota in Germany has positive effect on firm performance.
This paper is organized as follows. Section 2 provides an overview of the existing literature regarding gender diversity and gender equality quotas. The hypotheses based on the relevant literature will be discussed in section 2.4.
Afterwards, section 3 describes the data sample and the methodology used to test the hypotheses. Section 4 displays the results and provides an answer to the
research question. Finally, section 5 concludes by summarizing the main results and discusses the limitations of this paper.
2. Literature review
In this section the previous literature will be presented. The relationship between firm performance and gender diversity has been studied widely. In this section studies which resulted in a positive effect or in a negative effect will be discussed, also
studies on gender quotas will be presented. This will be followed by the formulation of the hypotheses based on the existing literature.
2.1 Studies that found a positive effect on firm performance
Carter et al. (2003) test whether there is a significant link between gender diversity and firm performance using a sample of S&P 500 firms. They measure gender
diversity as the percentage women in the board and firm performance is measured in Tobin’s Q. By using a 2-SLS measure they solved for reverse causality. They
controlled for size, industry and other corporate governance measures. The result of their research was a positive relationship between female representation in boards and firm performance. This led to the conclusion that diversity on a board results in higher firm value.
Nguyen and Faff (2007) did a similar research studying the relationship of gender diversity and firm performance. They followed to same methodology as Carter et al. only the data sample differs since Nguyen and Faff are using Australian firms over a multiple year time period. They used Tobin’s Q as a performance measure and percentage women for gender diversity. Their main conclusion was that presence of women on a board has a positive effect on firm performance.
Campbell and Vera (2008) used the fixed effects regression model to examine the relationship between firm performance and gender diversity. They use two
different measures for gender diversity in their research of Spanish firms, whereas most researchers only use one independent variable. They focus on either a gender dummy or the proportion female board members. The use of the dummy variable resulted in no effect of presence of women on firm performance measured as Tobin’s Q. However, the regressions being done with the ratio measure found a positive effect of female board members. They conclude that since presence has no influence but diversity does, the focus should be on the balance between men and women instead of just the presence of women in a board.
2.2 Studies that found a negative or no effect on firm performance
There are also academic papers that found either no effect or a negative effect on firm performance. Adams and Ferreira (2009) did quite a similar research as Carter et al. (2003) did. They followed the same methodology and used the same variables namely Tobin’s Q and percentage women. However, the main difference is that Carter et al. only examined one year of data and Adams and Ferreira used multiple years of S&P 500 data to reach their conclusion. Using multiple years of data is profitable since the quality of the analysis will improve due to dynamic factors (Marinova et al., 2015). The results of the research of Adams and Ferreira are the opposite of what Carter et al. found as they found a negative relationship between the percentage women in the board and firm performance.
Rose (2007) researched the topic of gender diversity looking at Danish firms over multiple years. He did a cross-sectional analysis and discovered that there is no significant link between firm performance measured in Tobin’s Q and female board representation. He also provided a reason for this insignificant relationship saying that board members who are not originating from the traditional ‘old boys club’ may have adopted the behaviour and norms of the conventional board members, which means that any benefits from having female board members are thus never realized or reflected in firm performance.
Bøhern and Strøm (2007) used repeated observations of the same firms in Norway over time and used the fixed effects technique to control for unobserved firm heterogeneity. Using a generalized least squares method they found that low gender diversity is more profitable for firms.
The ambiguous empirical evidence can be explained in multiple ways (Campbell & Vera, 2008). The researchers before all used different estimation methods. Some researchers do control for certain factors while others do not (e.g. leverage). Also, there may be unobserved factors which affect firm performance, and it therefore is more useful to use panel data than just cross-sectional studies
according to Campbell and Vera. Another reason for the differences in the results could be due to the data sample. The previous discussed researchers examine different countries and different time periods. The effect of gender diversity may be dependent on timing and legal context.
2.3 Studies regarding gender equality quotas
Previous papers did examine gender equality quotas. The main published articles are about the Norwegian quota, this is because it is the precursor so there is frequent data and since the quota is a natural experiment the findings are more likely to have causal relations (Borjas, 2015)
Ahern and Dittmar (2012) used the Norwegian quota to examine the influence of female board representation on firm performance measured in Tobin’s Q. Using an IV-regression they discovered that the Norwegian quota has a large negative impact on firm value. Moreover, they stated that a massive reorganization of corporate boards is imposed by the quota. This quota thus imposes extra costs for firms since it is forcing certain regulations which they have not met before (Ferreira 2015).
Dale-Olsen et al. (2013) examined the same quota. They found that the effect of the women’s equality quota is negligible, although it did increase female
representation. They used Return on Assets as a performance measure for this quota since they compared the listed firms for which the quota is implemented and non-listed firms as the control group. For non-listed firms it is difficult to measure Tobin’s Q since the market value is not known.
Matsa and Miller (2013) use the same approach as Dale-Olsen et al. when studying the Norwegian quota. The method they use is the difference in difference approach. They first look at firm performance before and after the quota was
implemented. They do this for both the natural and the control group. Subsequently, they take the difference for each group between the firm performance before and after the quota. Thereafter, the differences of both groups are compared and they took the difference by subtracting those differences from each other. The final difference is then tested to see whether this is significant. Matsa and Miller (2013) found that in three years the percentage women on the board increased from 18% to 40%, however this rise had a negative effect on firm profitability.
Since the implementation of the quota can come with costs, some firms decided to avoid the law by becoming a firm outside of Norway (Ahern & Dittmar, 2013). Since firms are now obligated to hire a certain number of female board members, there is a risk that the new board members are not as qualified as the incumbent, which may be a reason of the negative results of the Norwegian quota (Ferreira, 2015).
2.4 Hypotheses
As previously stated, this paper aims at broadening the existing literature about the relationship between gender diversity and firm performance using quotas in
Germany. Based on the previous literature two hypotheses are formulated.
Hypothesis 1: Gender diversity has no influence on firm performance.
The first hypothesis examines whether gender diversity has an effect on firm performance. Since the results of previous literature is ambiguous, this hypothesis expects that an increased gender diversity in a board does not have any influence on firm performance. This hypothesis is relevant to investigate the effects of gender diversity on firm performance.
Hypothesis 2: The quota implementation in Germany has not led to better firm performance for firms that have achieved the quota compared to those that have not.
This hypothesis is based on previous literature by Dale-Olsen et al. (2013) and Matsa and Miller (2013) who both found that the Norwegian quota had negative effects on firm performance. Therefore, this thesis expects that the German quota will also have a negative effect on firm performance.
3. Methodology
This section discusses how the data sample used for this research is created.
Additionally, it presents the methodology of the conducted research and displays the descriptive statistics.
3.1 Data sample
The research sample used in this thesis originally consisted of 106 listed firms in Germany. The FidAR Women-on-board-index 100 was used as the leading list of companies. All these firms have co-determined supervisory boards and are listed on the Germany exchange, which are the two requirements for a firm to be obligated to the law. The gender quota law began to be discussed frequently in 2013. The law passed in March 2015, and was officially enforced in January 2016. Therefore the data collected for this research will be over the time period January 2013 till December 2016.
The data sample contains annual data for a three-year window before and one-year after the law implementation on 1 January 2016. Data regarding the board of directors was collected from the ASSET4 ESG Database from Datastream. The financial data for this research was obtained through the Worldscope Database from Datastream. Additional data is possessed from the annual report of each individual firm if it was unavailable in Datastream. The original sample consisted of 106 firms, however the final sample was reduced to 86. This reduction was due to data
unavailability. Eight companies were removed from the sample since there was no data available of those firms in Datastream. Subsequently, twelve firms were
removed from the data sample since they went public during the years of the sample used, which entails that there is missing data for these companies. This leaves a total sample consisting of 86 companies and 344 observations used in this research. 3.2 Methodology
To examine the effect of the women’s quota implementation in Germany on firm performance two models are constructed. The dataset in this research has
information on 86 different entities observed at four different time periods, i.e. panel data. This panel data set has some missing data points for the variable
independence which is solved by using multiple imputation in Stata.
When using panel data, the two most commonly estimated models are the random effects regression model (further abbreviated RE) and the fixed effects model
(further abbreviated FE). To decide whether to use RE or FE in with this data sample, a Hausman (1978) test is executed. The null hypothesis of the Hausman test is that the preferred model is RE, the alternate hypothesis says that FE is preferred. The FE regression is a method for controlling for omitted variables in panel data when the omitted variables vary across entities but do not change over time (Stock & Watson, 2015). When doing a FE regression industry dummies are omitted, since they do not vary over time. According to Stock and Watson (2015) dummies variables can be added for industry and accordingly an OLS regression can be used. The industry
Utilities is excluded due to perfect multicollinearity.
𝐹𝑖𝑟𝑚 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒𝑖𝑡 = 𝛽0+ 𝛽𝑖∗ 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦𝑖𝑡+ 𝛽𝑖 ∗ 𝐶𝑜𝑛𝑡𝑟𝑜𝑙 𝑉𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠𝑖𝑡+ 𝛽𝑖 ∗
𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠𝑖𝑡+ 𝜀𝑖𝑡 (1)
Firm performance is be measured in two ways: through financial accounting data and market value based data. Following prior research by Adams and Ferreira (2009) and Ahern and Dittmar (2012) both Tobin’s Q and Return on Assets will be used as
measures for firm performance.
Tobin’s Q is a market value-based measure. This measure has been used in similar research (Adams & Ferreira, 2009) (Carter et al, 2003).
𝑇𝑜𝑏𝑖𝑛′𝑠 𝑄 = 𝑡𝑜𝑡𝑎𝑙 𝑚𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑓𝑖𝑟𝑚 𝑡𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑓𝑖𝑟𝑚
The Return on Assets is an accounting based measure. This measure has been used in the research of for instance (Ahern & Dittmar, 2012).
𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠 = 𝑛𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒 𝑡𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠
The independent variable that are used in this research are gender diversity and one binary variable. The binary variable is the quota dummy for firms that have achieved the quota in any given year. Gender diversity is measured as the proportion of women of the total board.
The first control variable used is firm size. According to Lee (2009) firm size has a positive impact on firm performance and it is therefore included. Firm size is measured as the natural logarithm of total assets of a specific firm. The second control variable is firm age. Baker and Kennedy (2002) state that often the better
performing firms survive, so this implies that age has a positive influence on firm performance. Moreover, firms that exist longer have a better understanding of their abilities and know their strengths and weaknesses, which increases firm
performance. Since the data sample used consists only of listed firms, the age of a firm is 2018 minus the year at which a firm went public. The third control variable is
independence. This is the independence of the board members. The agency theory
says that there are differences in the goals of managers and the shareholders. Increasing the independency of the board directors will lead to a decrease in these differences which will results in a better working firm. Thus independency can have a positive impact on a firm’s performance. This will be measured as the fraction of independent board members of the total board. The missing data points for this variable have been replaced by using multiple imputation in Stata. This means that the missing values for independence will be replaced by the estimated coefficient resulting from an estimation made with the observed data points.
The last control variable are eight industry dummies. Datastream classifies all firms into one of nine different industries. The nine industries included are industrials,
financials, consumer services, telecommunications, consumer goods, health care, basic materials technology and utilities. The last industry will be excluded from the
formula to prevent multicollinearity.
The regression equations used will be as following.
𝑇𝑜𝑏𝑖𝑛′𝑠 𝑄𝑖𝑡 = 𝛽0 + 𝛽1∗ 𝑊𝑜𝑚𝑒𝑛 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒𝑖𝑡+ 𝛽2∗ 𝐹𝑖𝑟𝑚 𝑠𝑖𝑧𝑒𝑖𝑡+ 𝛽3∗ 𝐹𝑖𝑟𝑚 𝑎𝑔𝑒𝑖𝑡+ 𝛽4∗ 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖𝑡+ 𝛽5∗ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙𝑠 + 𝛽6∗ 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 + 𝛽7∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑠𝑒𝑟𝑣𝑖𝑐𝑒𝑠 + 𝛽8∗ 𝑇𝑒𝑙𝑒𝑐𝑜𝑚𝑚𝑢𝑛𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠 + 𝛽9∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑔𝑜𝑜𝑑𝑠 + 𝛽10 ∗ 𝐻𝑒𝑎𝑙𝑡ℎ𝑐𝑎𝑟𝑒 + 𝛽11∗ 𝐵𝑎𝑠𝑖𝑐𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙𝑠 + 𝛽12 ∗ 𝑇𝑒𝑐ℎ𝑛𝑜𝑙𝑜𝑔𝑦 + 𝜀𝑖𝑡 (2) 𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠𝑖𝑡 = 𝛽0+ 𝛽1∗ 𝑊𝑜𝑚𝑒𝑛 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒𝑖𝑡+ 𝛽2∗ 𝐹𝑖𝑟𝑚 𝑠𝑖𝑧𝑒𝑖𝑡+ 𝛽3∗ 𝐹𝑖𝑟𝑚 𝑎𝑔𝑒𝑖𝑡+ 𝛽4∗ 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖𝑡+ 𝛽5∗ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙𝑠 + 𝛽6∗ 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 + 𝛽7 ∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑠𝑒𝑟𝑣𝑖𝑐𝑒𝑠 + 𝛽8∗ 𝑇𝑒𝑙𝑒𝑐𝑜𝑚𝑚𝑢𝑛𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠 + 𝛽9∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑔𝑜𝑜𝑑𝑠 + 𝛽10∗ 𝐻𝑒𝑎𝑙𝑡ℎ𝑐𝑎𝑟𝑒 + 𝛽11 ∗ 𝐵𝑎𝑠𝑖𝑐𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙𝑠 + 𝛽12 ∗ 𝑇𝑒𝑐ℎ𝑛𝑜𝑙𝑜𝑔𝑦 + 𝜀𝑖𝑡 (3)
𝑇𝑜𝑏𝑖𝑛′𝑠 𝑄𝑖𝑡 = 𝛽0 + 𝛽1∗ 𝑄𝑢𝑜𝑡𝑎 𝑑𝑢𝑚𝑚𝑦 + 𝛽2 ∗ 𝐹𝑖𝑟𝑚 𝑠𝑖𝑧𝑒𝑖𝑡+ 𝛽3 ∗ 𝐹𝑖𝑟𝑚 𝑎𝑔𝑒𝑖𝑡+ 𝛽4∗ 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖𝑡+ 𝛽5∗ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙𝑠 + 𝛽6∗ 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 + 𝛽7∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑠𝑒𝑟𝑣𝑖𝑐𝑒𝑠 + 𝛽8∗ 𝑇𝑒𝑙𝑒𝑐𝑜𝑚𝑚𝑢𝑛𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠 + 𝛽9∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑔𝑜𝑜𝑑𝑠 + 𝛽10 ∗ 𝐻𝑒𝑎𝑙𝑡ℎ𝑐𝑎𝑟𝑒 + 𝛽11 ∗ 𝐵𝑎𝑠𝑖𝑐𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙𝑠 + 𝛽12 ∗ 𝑇𝑒𝑐ℎ𝑛𝑜𝑙𝑜𝑔𝑦 + 𝜀𝑖𝑡 (4) 𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠𝑖𝑡 = 𝛽0+ 𝛽1∗ 𝑄𝑢𝑜𝑡𝑎 𝑑𝑢𝑚𝑚𝑦 + 𝛽2∗ 𝐹𝑖𝑟𝑚 𝑠𝑖𝑧𝑒𝑖𝑡+ 𝛽3∗ 𝐹𝑖𝑟𝑚 𝑎𝑔𝑒𝑖𝑡+ 𝛽4∗ 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖𝑡+ 𝛽5∗ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙𝑠 + 𝛽6∗ 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 + 𝛽7∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑠𝑒𝑟𝑣𝑖𝑐𝑒𝑠 + 𝛽8∗ 𝑇𝑒𝑙𝑒𝑐𝑜𝑚𝑚𝑢𝑛𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠 + 𝛽9∗ 𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟𝑔𝑜𝑜𝑑𝑠 + 𝛽10 ∗ 𝐻𝑒𝑎𝑙𝑡ℎ𝑐𝑎𝑟𝑒 + 𝛽11∗ 𝐵𝑎𝑠𝑖𝑐𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙𝑠 + 𝛽12 ∗ 𝑇𝑒𝑐ℎ𝑛𝑜𝑙𝑜𝑔𝑦 + 𝜀𝑖𝑡 (5)
4. Results
In this section the results of this research will be presented. First, the descriptive statistics will be displayed. Furthermore, the hypotheses will be tested using the regression results.
4.1 Descriptive statistics
The data sample used to describe the influence of the gender quota in Germany on firm performance is a strongly balanced panel consisting of 86 companies during the years 2013-2016. Table 4.1 presents the mean values of all variables over those years.
As Table 4.1 shows the composition of the boards did change over the years. The average board gender diversity i.e. the percentage women on the board,
increased from 22% in 2013 to 28% in 2016. Additionally, the number of firms that successfully implemented the gender quota has increased from 28% in 2013 to 45% in 2016. Therefore, one could assume that the increased gender diversity could be linked to the gender quota. Since Germany has not implemented any sanctions for companies which will not make the required percentage women, there are still firms which have not implemented the imposed 30% women on their board.
The dependent variables Tobin’s Q and Return on Assets increased from 2013 to 2014 by 12% and 27% respectively. In 2015 both firm performance measures decreased, however they were still higher than in 2013. Tobin’s Q increased in 2016 while the Return on Assets decreased compared with 2015. Overall the firm performance measures increased from 2013 to 2016 by 20% and 4%.
The means of the control variables in table 4.1 show that the average firm size remains relatively constant over the years. Firm age increases over the years. The control variable independence is relatively high in 2013 compared to the following years. A possible explanation for this is the unavailability of data for this control
variable. The nine industries are presented last in table 4.1. It provides information on the share of firms representing a certain industry.
Table 4.2 illustrates all the averages, standard deviations, the minimum and
maximum values. The observations for Tobin’s Q range from 0.007 to 6.333, with a standard deviation of 0.896. The high standard deviation value may be due to the differences in firm size. The other dependent variable Return on Assets has a minimum value of -16.4 and a maximum value of 53.58. The percentage of women ranges from zero to 41.67%. This means no firms has a majority of female board members. On average firms have 25% women in their board, which is almost the required amount by the gender quota.
4.2 Results
The Hausman test is conducted to know whether to use the RE regression model or the FE regression model. Since, the null hypothesis of the Hausman test is rejected this paper will use the FE regression model. Both the OLS and the FE regression results are tabulated. However, the conclusion for the hypothesis will be made based on the FE regression results, since this model takes more time-constant variables in consideration than only industries.
Table 4.3 reports the regression results for equations (2) and (3). Specification (1) to (4) are for the dependent variable Tobin’s Q and specification (5) to (8) use Return on Assets as dependent variable. The first two columns of each dependent variable are estimates done with an OLS regression and the last two columns of each dependent variable present the results using the fixed regression model.
The OLS regression results (1) show that when not controlling for anything percentage women has a positive effect on Tobin’s Q (t=1.97; p<0.05). This effect
increases when controlling for certain variables (t=2.84; p<0.01). This shows that when gender diversity increases firm performance will also increase. Specifications (3) and (4) are the results based on the FE model. This shows that there is no effect of gender diversity on firm performance measured in Tobin’s Q. The same
regressions are performed using Return on Assets as dependent variable. Using an OLS regression with Return on Assets results in an insignificant effect, as does using a fixed regression. These results after including the control variables are in line with those of Rose (2007), namely gender diversity has no significant effect on firm performance.
The control variable firm size shows different results than expected. Firm size has a negative effect on firm performance measured in both Tobin Q and Return on Assets using a fixed regression (p<0.05). This contradicts Lee (2009) who stated that firm size had a positive influence on firm performance. The industries consumer
goods and technology both have a positive effect on both firm performance
measures (p<0.01). When looking purely at the specification (6) it shows that all industries are significant positive except for industrial and financials.
The conclusion for hypothesis (1) will be based on the fixed regression results as mentioned. The first hypothesis stated that gender diversity has no effect on firm performance. The results indeed show that percentage women (i.e. gender diversity) has no influence on firm performance measured in both Tobin’s Q and Return on Assets. This is in line with the research of Rose (2007) who found that gender diversity has no effect on firm performance. Therefore, hypothesis (1) will not be rejected.
The results of the regressions used for the second hypothesis analysed with equation (4) and (5) are reported in table 4.4. In these equations a quota dummy is used as the independent variable. The OLS regression results for Tobin’s Q are presented in specification (1) and (2). When including the control variables the quota dummy has a positive effect on Tobin’s Q (t=1.94; p<0.1). This means that firms that have achieved the 30% quota have a higher firm performance than firms that have not. The fixed regression results are reported in column (3) and (4). When just looking at the quota dummy without any control variables is has a positive effect on firm performance (t=2.57; p<0.05). After including the control variables this effect remains positive (t=1.70; p<0.1). The positive effect of the quota is not in line with both Matsa and Miller (2013) and Dale-Olsen et al. (2013) who find that a gender quota had negative effect on firm performance.
The results using Return on Assets as dependent variable are all insignificant. Firm size has a negative effect on firm performance which is against expectations. Firm age has a significant positive effect on firm performance when measured in Tobin’s Q. This result is in line with Baker and Kennedy (2002) who suggest that firm performance is positively influenced by firm age. The industries do again have a positive effect on firm performance.
The hypothesis (2) that the quota implementation in Germany has not led to better firm performance for firms that have achieved the quota compared to those that have not is rejected. The fixed regression results using Tobin’s Q as independent variable show that the quota dummy has a positive effect on firm performance.
Therefore, firms that have reached the 30% gender diversity quota have a higher firm performance than the firms that have not yet reached the mandated quota.
The overall conclusion based on these results is that the first hypothesis will be not rejected since there is no effect of gender diversity on firm performance. The second hypothesis will be rejected since the quota has a positive influence on firm
performance measured in Tobin Q. The effect of the women’s quota in Germany is thus positive.
5. Conclusion and limitations
This paper aims to identify the effect of the women’s quota in Germany on firm performance.
Two hypotheses were tested using a data sample consisting of 86 firms over the years 2013 to 2016. Since the literature on gender diversity is ambiguous the first hypothesis states that gender diversity has no effect on firm performance. In line with the expectations, the results show that gender diversity (i.e. women percentage) has no influence on neither Tobin’s Q nor Return on Assets. Therefore, the first
hypothesis was not rejected. As Rose (2007) states this might be due to women adopting the behaviour and norms of the conventional board members and therefore making no difference compared to the old board composition. The second hypothesis stated that firms that have achieved the mandated percentage of women (i.e. the quota) do not perform better than firms that have not. When exploiting the results it is clear that the firms that have reached the quota do perform better than the firms that have not. Therefore the second hypothesis is rejected.
The conclusion is that gender diversity in general has no
influence on firm performance in Germany. However, the quota has a positive effect on firm performance. Nonetheless, there are some limitations to this research.
First of all, the time frame. Since the quota has been implemented the first of January of 2016 there is only one year of data available for the
after-implementation period. This paper mostly analysed the effect of the quota during the implementation period. Therefore, all the results in this paper are short-term and the ‘after-effect’ cannot be analysed precisely. Furthermore, due to the accessibility of Datastream database there is missing data. Some firms for which the quota was originally implemented were reduced from the data sample due to data availability. Also the variable independence has many missing data points. In this paper this is solved by using imputation, however it is preferred to have all the data to the variable. Subsequently, there are more board characteristics that affect firm performance indirectly such as age, nationality education and previous board experience (Ahern & Dittmar, 2012). Future research should include these characteristics in their research, as to fully understand the effect of changes in boardrooms on firm performance.
Additionally, this paper is limited by causality. The results suggest that the presence of women has a positive effect on firm performance. However, a firm which
is performing better may take higher risks and pursuit changes in their boardroom. It is not evident whether firm performance changes because of adjustments in board compositions or that increased firm performance is the reason for changes in board composition. Therefore, there is a problem of reverse causality, which should be solved in future research. Lastly, this quota has a low external validity. This quota was implemented for German firms which are both listed and have co-determined boards. Firms which do not meet the quota need to leave a board seat open, while other quotas have financials penalties or even dissolvement in the case of Norway. This differences in legislation is something to consider when comparing this research to other quotas.
6. Reference list
Adams, R., & Ferreira, D. (2009). Women in the boardroom and their impact on governance and performance. Journal of financial economics, 94(2), 291-309. Ahern, K. R. & Dittmar, A. K. (2012). The Changing of the Boards: The Impact on
Firm Valuation of Mandated Female Board Representation. The Quarterly
Journal of Economics, 127(1), 137-197.
Baker, G. P., & Kennedy, R. E. (2002). Surviorship and the economic grim reaper.
Journal of Law, Economics, and Organization, 18(2)324-361.
Borjas, G. J. (2015). Labor Economics. New York United States of America: McGraw-Hill Education.
Bøhren, ø., & Strøm, R. ø. (2007). Aligned, informed, and decisive: characteristics of value-creating boards. Ljubbjana Meetings Paper.
Campbell, K. & Minguez-Vera, A. (2008). Gender diversity in the Boardroom and Firm Financial Performance. Journal of Business Ethics, 83(3), 435-451. Carter, D. A., Simkins, B., & Simpson, W. (2003). Corporate Governance, Board
Diversity, and Firm Value. The Financial Review, 38(1), 33-53.
Dale-Olsen, H., Schone, P., & Verner, M. (2013). Diversity among Norwegian boards of directors: Does a quota for woman improve firm performance. Feminist
economics, 19(4), 110-135.
Fereirra, D. (2015). Board diversity: should we trust research to inform policy?
Corporate Goverance: An International Review, 23(2), 108-111.
Hausman, J. A. (1978). Specification Tests in Econometrics. Econometrica, 46(6), 1251-1271.
Lee, J. (2009). Does size matter in firm performance? Evidence from US public firms.
International Journal of the Economics of Business, 16(2), 189-203.
Marinova, J., Plantenga, J., & Remery, C. (2015). Gender diversity and firm
performance: evidence from Dutch and Danish boardrooms. The International
Matsa, D. A., & Miller, A. R. (2013). A female style in corporate leadership? Evidence from binding quotas. American Economic Journal: Applied Economics, 5(3), 136-169.
Nguyen, H., & Faff, R. (2007). Impact of board size and board diversity on firm value: Australian evidence. Corporate ownership and control, 4(2), 24-32.
Rose, C. (2007). Does female board representation influence firm performance? The Danish evidence. Corporate Governance: An International Review, 15(2), 404-413.
Stock, J. H., & Watson, M. W. (2015). Introduction to econometrics. Essex England: pearson education limited.
