Directors ’ Political Connections and Compliance with Board of Directors Regulations : The Case of S & P / Tsx 300 Companies

Further to the various scandals that shook North-American markets in the early 2000s, Canada reacted in 2004 and 2005 by changing its regulations on the governance practices of listed companies. Faccio (2006) and other authors have argued that politically connected companies can have less regulatory oversight than unconnected firms. This affirmation raises an issue that this article attempts to solve, i.e., whether politically connected Canadian companies are less compliant with regulatory requirements on governance than unconnected firms, and with board of directors requirements in particular. Although politically connected and unconnected firms differ significantly in their compliance with regulation, the analyses show that the state of being politically connected tends to have little bearing on regulatory compliance.


Introduction
Corporate governance is assured through a number of structures and mechanisms that reconcile the divergent interests of officers and shareholders.Different governance mechanisms are instituted to maximize the value of the firm (Agrawal & Knoeber, 1996), ensure better returns and limit the transfer of wealth among shareholders and officers so as to reduce the risk of total shareholder loss (Parrat, 2003).
Further to the various scandals that shook North-American markets in the early 2000s, Canada reacted in 2004 and 2005 by changing its regulations on the governance practices of listed companies.(Note 1) The objective of the Canadian regulations was to restore investor confidence in the Canadian financial market and ensure the protection of investments while remaining open to changes in governance trends in the United States and the rest of the world.(Note 2) Given that most of these regulations are voluntary, it is relevant to investigate the factors that influence compliance decisions by Canadian firms.
Quality of corporate governance is understood to depend on the institutional and regulatory environments of countries (Claessens & Yurtoglu, 2012; LaPorta, Lopez-de-Silanes, Shleifer & Vishny 1998;Liu, 2006;Matoussi & Jardak, 2012).However, regulation and its application are factors strongly influenced by political power (Claessens & Yurtoglu, 2012).Faccio (2006) and other authors have maintained that politically connected companies can have less regulatory oversight than unconnected firms.This affirmation raises an issue that this article attempts to solve, i.e., whether politically connected Canadian companies are less compliant than unconnected firms in regard to regulatory requirements on governance, and on the board of directors in particular.

Canadian Institutional Context
The Canadian financial market has undergone dramatic legal, institutional and operational changes since the beginning of 2000 (Carnaghan & Gunz, 2007).Prior to these changes, the Toronto Stock Exchange (TSE) had implemented the recommendations of the Dey Report, based on the conclusions of a committee of inquiry helmed by Peter Dey. (Note 3) The TSE's approach was founded on self-regulation, with no prescriptions to conform to the report's guidelines (Rousseau & Talbot, 2007).
In 2004 and 2005, a set of national policies and regulations were instituted to regulate some governance practices in Canada; most notably National Instrument  Corporate Governance Guidelines.These regulations apply to all listed companies in Canada and contain several recommendations about the nature and operations of the board of directors.
In a departure from tradition, this last reform sent a mixed message of tolerance and lack of constraint on the one hand and absolutism on the other; hence its characterization by Rousseau and Talbot (2007) as a middle-ground measure.Adopters of National Policy 58-110 are free to apply best practices as convenient, as the guidelines it contains are not prescriptive.Firms are responsible only for producing a comparative table outlining how they implemented the policy, or else declare their non-implementation (along with their reasons).
The prescriptive text is National Instrument 52-110 on Audit Committees.Application of this regulation was mandatory beginning on the first annual meeting held after July 2004, with the final deadline being July 2005.The reform's most tolerant aspect is reflected in National Policy 58-201 Corporate Governance Guidelines, which came into force in June 2005.The policy recommends best governance practices such as board and chairperson independence, separate meetings for independent administrators, and the establishment of a written charter delineating the board's role.

Literature Review and Hypotheses
According to several authors, the level and quality of corporate governance and the level of protection of investor interests depend on institutional factors, the country's level of development, and ownership structure (Claessens & Yurtoglu, 2012;LaPorta et al., 1998;Liu, 2006;Matoussi & Jardak, 2012).The basic premise behind these studies is that a legal system that properly fulfills its functions protects external investors, which in turn helps improve the firm's ability to increase its external financing and make the most of its opportunities for growth.Indeed, as underlined by Hail and Leuz (2006), the strong protection of outside investors limits their expropriation by top officers.
However, politics is one of the key determinants of a country's institutional landscape.According to Faccio (2006), many firms around the world are politically connected.Studies have demonstrated that being politically connected can result in numerous privileges for firms and lead to their improved financial performance (Dinc, 2005;Charumilind, Kali, & Wiwattanakantang, 2006;Faccio, 2006;Claessens, Underhill, & Zang, 2008;Goldman, Rocholl, & So, 2009;Dicko & Breton, 2013;Dicko & El Ibrami, 2013).Thus, Dinc (2005), in a study of emerging countries, showed that banks in which the government had at least a 20% stake increased their loans by 11% at election time, with the goal of serving the interests of politicians and politically connected firms.Charumilind et al. (2006) argued that politically connected companies have greater access than unconnected firms to long-term debt in Thailand.Claessens et al. (2008) studied a sample of Brazilian companies and found that firms that contributed financially to the coffers of elected officials during the 1998 and 2002 electoral races experienced an increase in market returns compared with those that did not make such contributions.Goldman, Rocholl, & So (2009) examined the effects of the political connections of S&P 500 companies and concluded that firms that appointed a politically connected person to their board experienced market returns at the time of the announcement of the nomination.In addition, companies connected to the Republic Party at the time of the 2000 US elections increased in value, while those connected to the Democratic Party had the opposite experience.
From an OSISRIS data base sample of top 100 Canadian firms, Dicko & Breton (2013) showed that directors' political connections have a positive and significant link with financial performance (measured by return on assets).Furthermore, Dicko & El Ibrami (2013) concluded that directors' political connections influence positively and significantly the long term debt increase, on a Compustat randomly selected sample of 300 Canadian companies.
According to Faccio (2006), politically connected firms can have lighter taxation, preferential treatment in competition for government contracts or relaxed regulatory oversight than unconnected firms.So far few studies have examined the link between political connections and regulatory compliance, especially in the Canadian context.The present study aims to fill this gap, particularly as regards regulations that have both mandatory and elective components.
In view of the foregoing, and given the influence of political connections and the elective aspect of the board of directors' regulations under study, it is expected that: H 1 : Politically connected firms are less likely than unconnected firms to comply with board of directors regulations.More specifically, H 1a : Politically connected firms are less likely to comply with regulations on specific characteristics of the board of directors.H 1b : Politically connected firms are less likely to comply with regulations on the structure and operation of board of directors committees.

Sample
This study was conducted on TSX 300 firms in 2010, i.e., five years after the regulations came into force (2005).By then companies had had the time to make adjustments and weather the 2008 crisis.For the investigation, financial data were pulled from COMPUSTAT.Board of directors data were collected using information from directors' circulars and annual notices available on Sedar.com.Data on board of directors' political connections were collected from company websites and the SEDAR database.The sample data are presented in Table 1.Missing data 9 Final sample 199

Model and Analyses
The main goal of this study was to compare and contrast compliance with regulations on the board of directors by politically connected and unconnected firms.Variance analysis (ANOVA) was used for this purpose.
Based on the research hypotheses whereby such compliance is influenced by a firm's political connections via its directors, the following linear regression model was used:

Compliance = Political Connections + Indebtedness + Firm Size + Sector + US Market Value
Where: -Compliance is the board's level of conformance to Canadian regulations on the board of directors (30-points index); -Political connections take the value of 1 if at least one of the firm's directors is or has been a member of parliament, works for the government, belongs to a political party or has made a financial contribution to a political party; -Size is the logarithm of the firm's assets in 2010; -US market value takes the value of 1 if the firm is subject to US stock exchange regulations, and 0 otherwise; -Sector is the firm's industry group as defined in SEDAR.Nine sectors were identified and coded 1 to 9; -Indebtedness is measured by the long-term debt to total assets ratio.

Dependent Variable: Compliance with Canadian Regulations
Levels of conformance to Canadian regulations on the board of directors was measured through a 30-points index, consistent with other studies that used similar indices (Gombers, Ishii & Metrick, 2003;Defond, Hann & Hu, 2005;Bhagat & Bolton, 2008).The index was validated previously by financial experts and used in other studies (Khemakhem & Naciri, 2011;Khemakhem, Baillargeon & Gélinas, 2013).
It provides the advantage of a better global measurement and comparability among firms (Bhagat, Bolton & Romano, 2008).For this study, an objective frame of reference tailored to the Canadian context was selected: National Policy 58-201 and National Instrument 52-110, adopted by financial market authorities in Canada.
National Policy 58-201 provides examples of best governance practices such as board and chairman independence, separate meetings for independent directors and the existence of a written charter (Note 4) and a code of ethics.This regulation provides a source of compliance points in regard to board of directors and committee characteristics (see Appendix containing the measurement grid).
The index was further enriched by National Instrument 52-110 Audit Committees, which stipulates the use of independent external auditors and an independent audit committee.This regulation provides a source of compliance points in regard to the audit committee and the independence of board members and committees.

Measurement of the "Political Connections" Independent Variable
The literature defines the firm's political connections in various ways.According to Faccio (2006) and other authors (Boubakri, ElGoul & Saffar, 2012), a firm is politically connected if its majority shareholder or one of its officers or directors is a member of parliament, a minister or government leader, or is politically connected (through family ties or friendships, for example).
In addition, a firm is politically connected if it contributes financially to a political party, according to Dinc (2005), Claessens et al., (2008) and Goldman et al., (2009).
The present study looks at the firm's political connections by way of its directors, using criteria found in the literature (particularly Faccio, 2006;Goldman et al., 2009) and applying them to a single director.Thus, the variable takes the value of 1 if at least one director is or has been a member of the government or parliament, has contributed financially to a party or is a member of a political party.

Model Control Variable
The firm's characteristics impact on the level and quality of governance as well as the institutional context.For example, Aggarwal, Erel, Stulz, & Williamson (2009) demonstrated that the following variables characterizing the firm impact on its governance: size (measured by the logarithm of its assets), long-term debt divided by total assets, and industry sector.
Similar to Aggarwal et al. (2009), the current investigation used the following control variables: firm size (measured by the logarithm of sales), relative indebtedness (measured by long-term debt divided by total assets), and industry sector (measured by a dummy variable coded 1 to 9).The firm's US market value was also taken into consideration.Given that a number of Canadian firms were subject to US regulations as a result of being listed on the US stock exchange, and that US regulations are considered to be considerably more stringent, the firm's listing on the US market was included as a control variable.

Results and Discussion
Variance analysis (ANOVA) was used to study the difference between politically connected and unconnected firms in terms of their compliance with regulations on the board of directors.
Thereafter, linear regressions helped determine whether the fact of being politically connected influences the firm's compliance with regulations on the board of directors.

Compliance and Political Connection Indices
For this study, several indices were developed to measure compliance from the aspect of:


The board of directors (board of directors score);  Board committees other than audit (committee score); (Note 5)


Total compliance (Total score).
The mean results for the variance analysis are presented in Table 2. Table 2 shows a general and significant difference between politically connected and unconnected firms in their conformance to regulations on the board of directors (significant total score).Taken individually, only the difference in conformance relative to other committees appears to be non-significant among politically connected and unconnected firms (other committees score).This is simply because regulations about other committees are recommendations rather than prescriptions.However, this study's hypotheses are still not confirmed because the manner in which political connections influence conformance to regulations is unknown.
To answer that question, regression analyses were performed.

Analysis of Linear Regressions of Compliance and Political Connection Scores
An analysis was performed of the three aforementioned scores: board of directors, other committees (other than the audit committee) and total score (board and all its committees).

Relationship between Board of Directors Score and Political Connections
Contrary to predictions, political connections positively influence compliance with regulations on the board of directors (Table 3).This could be explained by the fact that given their probable political exposure and visibility, firms with political connections have an interest in using the very best of practices in regard to regulations so they can avoid being accused of conflicts of interest and favouritism.The results for this category are similar to those obtained for the relationship between political connections and compliance with regulation.

Sensitivity Analysis
Given that a number of Canadian firms are subject to mandatory US regulations, it was decided that this variable should be eliminated from the model when performing the analyses.The results obtained were roughly the same, and were not tabulated.

Conclusion
The Canadian financial market has undergone numerous changes since 2000 (Carnaghan & Gunz, 2007).The early 2000s were marked by a series of financial scandals that shook the confidence of investors and required lawmakers and regulatory bodies to intervene and shore up several aspects of governance among listed firms.These events prompted changes in the regulations on the corporate governance practices of Canadian firms, beginning with the board of directors and its committees.(Note 6) The goal of this research was to study the relationship between compliance with Canadian regulations on characteristics of the board of directors and its committees, and the firm's political connections.
Although a significant difference was noted in the compliance levels of politically connected and unconnected firms, the results of the analyses did not overwhelmingly demonstrate the impact of political connections.In terms of compliance with regulations on the board of directors, firm size and listing on the US market are the two foremost determining factors.Political connections may lead to the adoption of best practices as a way of avoiding accusations of favouritism or potential conflicts of interest.In cases of non-compliance, firms risk being penalized on the financial markets (Gompers et al., 2003), and as a result may see an increase in financing costs (Khemakhem & Naciri, 2011).
This study has some limitations.The most widespread variables in the literature were used to explain compliance with governance regulations despite the fact that the Canadian context has a strong concentration of ownership and controlling shareholders (Bozec & Laurin, 2007).Some studies (Klein, Shapira & Young CGIR, 2005;Bozec & Bozec, 2007) showed that ownership structure could be a determining factor in the quality of corporate governance mechanisms.Future research may benefit from focussing on ownership structure along with political connections.

Table 1 .
Sample selection

Table 3 .
Linear regression results dependent variable: board score

Table 4 .
Linear regression results dependent variable: total score Relationship between the Scores for Other Committees and Political Connections

Table 5 .
Linear regression results dependent variable: score total other committees