Institutional background

After its independence in 1947, India followed a mixed economy model, allowing both private and public sectors to contribute to economic growth. It emphasized the establishment of heavy industries and, in practice, accorded a prominent role for the public sector. The government regulated the private sector through licensing policy and provided protection from competition by preventing international imports. These license-permit raj policies made the entrepreneurial activities (both private as well as public) less productive and failed to yield desired economic growth. With the below-par success of this mixed economy model, since the late 1980s, India started adopting the probusiness policies, shifting the focus from redistribution of wealth to economic growth. The government started opening sectors that were formerly earmarked for the public sector to the private sector. These promarket reforms were implemented in two stages.Footnote ¹⁵ First, internal liberalization policies were introduced in 1980s. These policies include easing the restrictions of the Monopolies and Restrictive Trade Practices Act, allowing broad-banding and delicensing of twenty-five industries (e.g., paper, printing, pharmaceuticals, rubber, ceramics). Second, the reforms of 1991, reconfigured both external and internal economic policies including the elimination of industrial licensing system and reducing the government control over the private sector. These reforms turned the focus to the “private-sector investment” as the main engine of economic growth. The virtual elimination of the licensing system increased capital mobility and removed the controlling role of government in the location of private-sector investments. Further, multiparty coalition governments in India have advanced these reforms in fits and starts. This economic liberalization in India resulted in a greatly increased role for the private sector.

In India, political decision making is decentralized, with both central and state governments having constitutionally defined jurisdictions. Firms are governed by the guidelines and policies framed by both these central government and state governments. Central government makes laws pertaining to natural resources, financial markets, taxation, and international trade, among others. The state governments control policies on land, labor, police, local infrastructure, utilities (including electricity and water), and so on. With the delicensing of most of the industries since the 1980s, the role of state governments has increased in facilitating the private-sector investments and creating probusiness environment. Further, policy makers’ political fortunes at both central and state governments, mostly tied to economic performance and job creation; hence they need to attract private capital to spur economic growth.

An examination of Indian business composition, three decades after proliberalization policies, reveals that old conglomerates and large business houses control a large share of its economy.Footnote ¹⁶ This can be attributed to the “emerging” status of the Indian business environment that still suffers from several institutional voids. Though India is the most densely populated and democratically governed, its bureaucratic and political system is largely unaccountable and corrupt.Footnote ¹⁷ Transparency InternationalFootnote ¹⁸ places India as one of the highly corrupt countries with an index value of 41 on a scale of 0 to 100 (average score of all 180 countries assessed has been 43, and 67% of the countries score lower than 50), where 100 is very clean and 0 is highly corrupt, for the year 2019. Legal systems in India are overburdened and characterized by prolonged procedural delays in contract enforcement.Footnote ¹⁹ Bureaucratic red tape often causes inordinate delays, higher costs, and reduced flexibility for businesses to embark on their enterprising activities. However, India has witnessed rapid growth on a few elements of democracy that help entrepreneurship. They include a vibrant press, vocal media, and vigilant civil society. They keep a careful watch on politicians, bureaucracy, and companies.Footnote ²⁰ Nevertheless, institutional mechanisms continue to provide avenues for government officials to apply their discretions.Footnote ²¹ In such an institutional environment, firms seek political connections to navigate through institutional and market imperfections and to obtain favorable consideration from the local and central governments.Footnote ²² In this context, we attempt to investigate the influence of political connections on various financial outcomes of Indian listed firms.

Literature and hypotheses development

Positive effects of political connections

Positive effects of firms’ political connections can be explained using the resource-based theory of firm.Footnote ²³ The theory suggests that firms possessing “strategic resources” have competitive advantage over their rivals, leading to sustained profits and higher market value. “Political connections” can be one of such firms’ strategic resources as they are valuable, difficult to imitate, and nonsubstitutable, especially in emerging economies such as India. India, though adopted market-centric economic policies way back in 1980s, governments still control key resources such as business operating licenses in various industries,Footnote ²⁴ infrastructure, electricity, land, and natural resources (crucial raw material for various industries). Owing to the underdeveloped nature of its market institutions, businesses in India face numerous obstacles.Footnote ²⁵ Further, legal systems in India are too weak to enforce contracts and protect property rights. As argued by Li et al. (Reference Li, Meng, Wang and Zhou2008), in such an environment, close connection with the policy makers and/or bureaucrats helps businesses to successfully navigate through these ill-functioning institutions. In a way, political connections provide value to firms in India in terms of beneficial treatment by government-controlled enterprises (e.g., raw material producers for energy-generating firms), preference in government contracts, relaxed regulatory oversight, and access to the communication on expected changes in the relevant policies.Footnote ²⁶

Further, political connections, as strategic resources, help firms to obtain favorable policy environment,Footnote ²⁷ access to institutional debt financing,Footnote ²⁸ and support in times of their economic turbulence,Footnote ²⁹ thereby leading to their enhanced market value.Footnote ³⁰ Fisman (Reference Fisman2001), examining a sample of Indonesian firms during 1995 and 1997 when there were rumors about President Suharto's health conditions, reported that firms’ political connections positively influence their value. Johnson and Mitton (Reference Johnson and Mitton2003) noted the increased market value of politically connected Malaysian firms after the imposition of capital controls in September 1998. Faccio (Reference Faccio2006), from her multicountry analysis, reports that political connections are valuable in the countries with a greater incidence of corruption. Bao et al. (Reference Bao, Johan and Kutsuna2016) report that political connections help firms in accessing the financial markets by reducing the costs of initial public offering (IPO), including IPO rejection risk. Further, Goldman et al. (Reference Goldman, Rocholl and So2009) note that political connections are strategic resources even in the context of well-developed legal and institutional mechanisms. In India, with its higher incidence of corruption and interventional government policies, we expect that firms prefer to have political connections to manage their external environment. Apparently, barring a few studiesFootnote ³¹ that examined Indian firms as part of their multicountry analysis, there is no focused study that examined the value of corporate political connections in the context of Indian firms. Hence, we present our prediction in the form of the following hypothesis.

Negative effects of political connections

While resource-based theory of firm postulates positive association between firms’ political connections and their financial outcomes, many scholars advocate negative effects of corporate political connections. In one of the seminal papers, Shilefer and Vishney (Reference Shleifer and Vishny1994) show that firms benefit from political/bureaucratic connections if associated marginal benefits outbalance the marginal costs. The costs of political connections can be comprehended using several implications of the agency theory of firm. First (rent-seeking activities of politically connected board members), politically connected board members may pressurize the firms to undertake the activities that serve their political interests. These activities include providing the excessive employment, undertaking investment activities in the areas of their political interest, deferring the plant winding up activities to the nonelection years, aiding/acquiring poorly performing but sensitive companies for their political interest, and so on. These rent-seeking activities of politically connected board members can be expensive and negatively influence the value of politically connected firms.Footnote ³² Second (information asymmetry/opaqueness), corporate political connections are more often the manifestation of existing executives’ personal preference for ideology rather than to further economic benefits of the firm.Footnote ³³ Aggarwal et al. (Reference Aggarwal, Meschke and Wang2012), examining the US firms, report that firms’ political donations negatively influence their long-term stock return performance. Further, politically connected firms are more opaque with their managed earnings and higher related party transactions than similar nonconnected firms.Footnote ³⁴ Gul (Reference Gul2006), examining the Malaysian politically connected firms, reported that they are prone to financial misstatements leading to increased audit efforts and related fees. Such an opaque environment provides a ground to indulge in expropriation by controlling shareholders. These agency problems between politically connected board members and minority shareholders contribute to the lower valuation of politically connected firms. Third (less professionalism), firms appoint politically connected board members for their network connections with the policy formulation and implementation. These politically connected board members tend to have less business acumen and subject expertise leading to inferior corporate decisions.Footnote ³⁵ Further, firms do not prefer to fire politically connected board members even if their performance don't meet the expectations.Footnote ³⁶ In a way, these directors may not be truly independent and may not deliver their fiduciary duties in monitoring the senior management. They will exacerbate the agency problems, especially in emerging markets leading to lower firm valuation.Footnote ³⁷ Fan et al. (Reference Fan, Wong and Zhang2007), examining the post-IPO stock return performance of Chinese firms, reported that politically nonconnected firms outperform connected firms as market participants believe that Chinese firms with political connections have the board of directors with a lower degree of professionalism.

Fourth (uncertainty in contract enforcement), politically connected firms may potentially influence the contract enforcement, leading to an unfavorable ex-ante adjustment of their contracts.Footnote ³⁸ Keefe (Reference Keefe2019) shows that loans to a politically connected borrower have a lower contract enforcement probability, causing an ex-ante upward adjustment of their interest rates. Bliss and Gul (Reference Bliss and Gul2012), examining Malaysian firms, reported that lenders charge higher interest rates on the loans when advanced to politically connected firms. They consider these firms are risky in terms of uncertainty in enforcing contracts. For similar reasons, politically connected firms face difficulties in nurturing or developing associations with technology providers, raw material suppliers, and customers. Luo et al. (Reference Luo, Yang and He2020) show that connected new generation firms have difficulty in making partners as other firms will be concerned about the possible abuse of power to extract more than negotiated in contracts, leading to depriving politically connected firms of advanced technology and other capabilities available with prospective partners. Finally (ex-ante poor performance), politically connected firms are often ex-ante poor-performing ones. That is, poor-performing firms often seek political connections to obtain favorable policy support in terms of subsidies and bailout in the case of their financial distress.Footnote ³⁹ Managers of these politically connected firms then may indulge in excessive risk-taking with the confidence that political patronage would bailout them in the case of financial distress, leading to their lower shareholders’ wealth.Footnote ⁴⁰

We note that our sample firms are not responsible for policy outcomesFootnote ⁴¹ as is the case of Chinese firms but are highly prone to incur the previously described agency costs due to political connections. These agency costs are expected to be prominent in the context of our sample firms, given the underdeveloped nature of legal and institutional mechanisms in India. Hence, we present our prediction in the form of the following hypothesis.

Moderating effect of firm- and industry-specific characteristics

Is the influence of political connections the same across all the firms? We argue that the effect of political connections may be heterogeneous across the connected firms. Though the available literature has largely ignored this heterogeneity, we posit that firm- and industry-specific characteristics, including their size, operational efficiency, financial leverage, and the degree of concentration in their industry, would differently moderate the cost and benefits of corporate political connections.

Firms’ political connections and their leverage

The available empirical evidence indicates that politically connected firms often use greater debt capital than nonconnected firms.Footnote ⁴⁹ Lenders prefer to extend the credit to politically connected firms despite their poor operating performance as they are more likely to get rescued in the instance of their distressing financial position.Footnote ⁵⁰ Further, politically connected firms can use their political network in securing preferential access to debt financing,Footnote ⁵¹ especially from government-controlled institutional lenders.Footnote ⁵² In India, public debt market is not yet well-developed, and firms predominantly depend on the institutional lenders for their debt financing. India's value of corporate bonds to GDP ratio has been a mere 17.16%, whereas the same statistic for the United States, Brazil, and Malaysia are 123.47%, 99.05%, and 44.5%, respectively.Footnote ⁵³ Because Indian firms predominantly depend on banks (of which public-sector banks, where the government is the major shareholder, account for 59.8% of the total value of loansFootnote ⁵⁴ as of November 2020) for their debt financing, we expect that firms’ political connections help them in securing favorable treatment from these lending institutions, leading to the higher leverage in the case of politically connected firms.

However, lenders may charge higher interest rates or shun politically connected firms, given their inherent internal agency problems, associated uncertainty with their contract enforcement, and well-documented opaqueness or prevalence of their accounting misstatements. Lenders understand that politically connected firms may (not deploy the borrowed funds productively) divert the portion of the borrowed funds for unproductive purposes, which decrease the returns/cash flows available to repay the loans.Footnote ⁵⁵ As mentioned earlier, lenders also see uncertainty in ex-post contractual enforcement with politically connected firms, and ex-ante charge higher interest rates while lending to politically connected firms.Footnote ⁵⁶ Further, politically connected firms in emerging markets, with their opaque financial disclosures, are less likely to choose inexpensive global financing as it attracts greater public scrutiny with the attention of foreign analysts and international financial press.Footnote ⁵⁷ With the preceding understanding, we propose to empirically test the following hypotheses in the context of Indian firms.

Sample and data

We consider the study period from 2006 to 2019. We choose 2006 as the beginning year for our study as capital markets regulator, Securities Exchange Board of India (SEBI) implemented revised Clause 49 of the listing agreement to the Indian stock exchange, effective 31 December 2005. It has then become mandatory for the Indian firms to implement the robust corporate governance guidelines along with their disclosures from 2006. We obtained sample firms’ financial and market price information from PROWESS-DX, a database created and maintained by the Center for Monitoring Indian Economy (CMIE). The PROWESS-DX provides the information on financial performance of Indian firms obtained from their annual reports and various disclosures (including statutory disclosures). The PROWESS database was earlier utilized by many studies for examining the Indian firms’ performanceFootnote ⁵⁸ and leverage choices.Footnote ⁵⁹ We accessed information about sample firms’ board of directors from the Indian Boards database maintained by Prime Database Group. The Indian Boards database includes the information on the board of directors of all the companies listed on the National Stock Exchange (NSE).Footnote ⁶⁰ Using the board of directors’ profiles from the database, we hand collected the relevant information, including their professional details. Because the board of directors’ profiles are accessible only for the NSE-listed firms, we confine our empirical investigations to the NSE-listed firms.

We exclude all government-controlled firms from the analysis as their board of directors’ selection process and their operating decisions are often not driven by objective economic and market criteria. We also exclude all financial services firms as they follow varied disclosure requirements. Further, we eliminate firm-year observations with negative book value of equity and those with negative net sales to exclude the firms with financial distress and those with no or negligible operating activities.Footnote ⁶¹ Our final sample is an unbalanced one with 13,753 firm-year observations, consisting 1,649 unique firms with an average of 8.34 firm-year observations each. We note that the percentage of firms with political connections in each year of the study periodFootnote ⁶² averages approximately 25%. Table 1 provides an industry-wide distribution of firm-year observations, including those with political connections across the industries. We identify the sample firms with their three-digit National Industrial Classification (NIC) codeFootnote ⁶³ and further regroup them into thirty broad industries. In classifying sample firms into industry groups, we follow Fama and French (Reference Fama and French1997) industry classificationFootnote ⁶⁴ in spirit, though we made minor changes to their basic criteria to adapt to the differences between NIC and SIC (Standard Industrial Classification) codes.

Table 1:

Sample distribution across industries

Note: Table presents the distribution of all sample firm-year observations along with those politically connected across the industries. We identify the sample firms with their three-digit NIC code and further regroup them into thirty broad industries using Fama and French (Reference Fama and French1997) industry classification.

From table 1, we observe that a greater percentage of sample firms operating in the industry of electricity and gas (i.e., 59.7%) are politically connected, followed by firms operating in the industries such as transportation and warehousing (47.9%) and mining (41.8%). These findings are consistent with those of Agrawal and Knoeber (Reference Agrawal and Knoeber2001), who reported the highest share of politically connected companies in the industries where political connections matter more. In the case of utility industries such as electricity and gas, where government is the predominant customer and also plays a role of supplier or facilitator of crucial raw materials (such as coal) in these industries. The working knowledge with government regulations and licensing are crucial for the firms to successfully navigate in the industries such as transportation and mining. We also note that firms operating in the industries such as education (7.3%), agriculture (11.5%), and computers and electronics (12.2%) are less likely to have political connections given the limited importance of regulatory process in these industries.

Variables

Estimation methods

Effect of political connections on firm value

Firms’ decision to have political connections is their endogenous choice and needs to be empirically accounted while assessing the influence of firms’ political connections on their financial outcomes, such as firm value. Barring a few,Footnote ⁸¹ most of the prior studiesFootnote ⁸² often ignored the same and considered that firm's choice to have political connections as exogeneous, leading to the violation of random assignment assumption, causing serious self-selection problem. We make use of Heckman two-step treatment effects procedureFootnote ⁸³ to address the self-selection problem. Accordingly, we use the following empirical models to examine the influence of firms’ political connections on their market value.

(a)$$PolConn_{it} = {\rm \;}f( {X_{it}, \;{\rm \;\;}\delta_t, \;{\rm \;}\alpha_d} ) {\rm \;} $$

(b)$$FirmValue_{it} = {\rm \;}g( {PolConn_{it}, \;{\rm \;}X_{it}, \;{\rm \;\;}\delta_t, \;{\rm \;}\alpha_d} ) {\rm \;} $$

In the preceding models, PolConn _it is a binary variable taking one if firm i is politically connected in year t and zero otherwise. FirmValue _it is the market value measure of firm. X _it is the vector of control variables including size, financial leverage, sales growth, asset turnover ratio, capital expenditure, industry four-firm concentration ratio, age, business group affiliation, and firm-specific risk. These models also include time (δ _t) and industry (α _d) dummies to account for the time and industry fixed effects. The functional forms of the preceding models can be specified as follows.

(1)$$d_{it} = X_{it}\gamma + \delta _t + \alpha _d + v_{it}$$

In the preceding equation, d _it refers to the firm's decision to have political connection.

(2)$$y_{it} = d_{it}\gamma + X_{it}\beta + \delta _t + \alpha _d + e_{it}$$

where y _it is the financial outcome variable, firm value, of firm i in period t. To account for the self-selection bias arising from the firms’ endogenous choice of having political connections, we estimate selection equation (i.e., equation – 1) using the Probit regression estimator. It allows us to compute inverse Mills ratio (λ).Footnote ⁸⁴ We use it as one of independent variables in the outcome equation (i.e., equation 2). Thus, the extended outcome equation as per the Heckman two-step method is specified as follows.

(3)$$y_{it} = d_{it}\gamma + X_{it}\beta + \lambda _{it}{\rm \rho } + \delta _t + \alpha _d + \varepsilon _{it}$$

This estimation procedure assumes that the error terms of equations 1 and 2 follow a standard bivariate normal distribution and accounts for the potential correlation between them. Further, such a selection model requires valid “exclusion restriction.” That is, a variable(s) that determines a firm's decision to have political connection but does not affect the outcome variable, firm value. We considered variables such as firms’ internationalization,Footnote ⁸⁵ economic policy sensitivity,Footnote ⁸⁶ institutional ownership,Footnote ⁸⁷ and industry political connectednessFootnote ⁸⁸ as the possible exclusion restrictions. But, we find that firms’ economic policy sensitivity, internationalization, and institutional ownership are not the valid exclusion restrictions in our dataset. Hence, because we use industry political connectedness as the exclusion restriction, firms transacting in the industries that are highly regulated are more likely to have a political connection. However, firms’ market value is more influenced by other firm-specific variables than their industries’ political connectedness. The findings of exclusion restriction testFootnote ⁸⁹ indicate the validity and relevance of industry political connectedness as the exclusion restriction in our analysis as it is correlated with firms’ political connections and uncorrelated with firms’ market value.

Further, unobserved firm-specific heterogeneity may affect the findings as firms’ political connections may be correlated with the unobserved firm characteristics. Heckman's two-step procedure, described previously, does not control this potential endogeneity problem. Semykina and Wooldridge (Reference Semykina and Wooldridge2010) showed that, in the presence of unobserved firm-specific effects, mere inclusion of the Mills ratio in equation 2 will not provide the consistent estimates. They offered a correction for selection bias that accounts for potential endogeneity. The proposed correction suggests an addition of time average of all the exogeneous explanatory variables in the equations 1 and 2. Because this procedure corrects for selection bias and potential endogeneity along with the accounting for firm-specific unobserved heterogeneity, researchersFootnote ⁹⁰ exploring the similar issues in the extant literature have preferred the results obtained using Semykina and Wooldridge (Reference Semykina and Wooldridge2010) approach. We also focus on the estimates obtained using Semykina and Wooldridge (Reference Semykina and Wooldridge2010) method in examining the influence of political connections on the sample firms’ market value.

Effect of political connections on leverage

We investigate the effect of firms’ political connections on their leverage using the following functional form.

(4)$$Lev_{it} = {\rm \;}Lev_{it-1}\partial + d_{it}\gamma + X_{it}\beta + \delta _t + \alpha _d + \vartheta _{it}{\rm \;}$$

where Lev _it is the leverage of i^th firm in year t. The X _it in the preceding equation is the vector of control variables, including return on assets, sales growth, nondebt tax shields, tangibility, size, age, business group affiliation, industry concentration (four-firm concentration ratio), firm-specific risk, and industry median leverage, that affect firms’ leverage choice. We estimate the preceding dynamic equation using system generalized methods of moments (system GMM) estimator suggested by Blundell and Bond (Reference Blundell and Bond1998). The GMM estimator corrects the endogeneity of the input variables using internally generated instruments. Further, system GMM estimator addresses the “finite sample bias”Footnote ⁹¹ inherent in the estimation of dynamic panel data model (i.e., equation 4). The system GMM estimator is a system of two equations at levels as well as at differences. We consider lagged values of all the independent variables up to three years as instruments in the case of both level and difference equations. We test the null hypothesis of zero serial correlation in first differenced residuals (AR1 and AR2) and report their p-values. Further, we investigate the validity of instruments using Hansen test of overidentification and difference-in-Hansen test of exogeneity. We report p-values of these two tests with the null hypotheses that the instruments are both valid and exogenous respectively.

Univariate analysis

Table 2 reports the summary statistics along with pairwise correlation coefficients of the variables included in the empirical analysis. We winsorize the ratio variables at 1 and 99 percentiles to do away with the effect of outliers. In our sample, 25.4% of the total firm-year observations have political connections. The reported statistics indicate that political connections are positively correlated with firms’ size, age, market value, operating profit, and industry concentration. It is negatively correlated with firms’ leverage. Sample firms’ leverage reports to have positive correlation with size, tangibility, capital expenditure, and nondebt tax shields. It has negative correlation with sample firms’ performance (Tobin's Q and ROA), efficiency (asset turnover ratio), and sales growth. However, variance inflation factors suggest that there is no multicollinearity problemFootnote ⁹² in our dataset.

Table 2:

Descriptive statistics and pairwise correlation matrix

Note: Table provides the descriptive statistics and pairwise correlation of the variables included in the analysis for all the firm-year observations. The definitions of these variables are provided in appendix A. To control for the outliers, we winsorize all the ratio variables at 1 and 99 percentiles. The correlation coefficient estimates among the included variables have been tested for their statistical significance and reported accordingly.

***, **, and * refer to 1%, 5%, and 10% significant levels, respectively.

Difference between politically connected and nonconnected firms

Table 3 presents univariate comparative statistics of the variables for the samples split based on their political connections. Politically connected firms, both in terms of their mean and median values, are largerFootnote ⁹³ and more matured than nonconnected firms.Footnote ⁹⁴ They spend more on capital expenditure, and report greater market valuation and better operating income (measured using the return on assets; ROA) than nonconnected firms. The reported findings suggest that firms operating in highly concentrated industries are more likely to have political connections. Contrary to the evidence reported by the earlier studies,Footnote ⁹⁵ our politically connected sample firms report on an average lower leverage ratio than nonconnected firms, though the median difference in their leverage is not statistically significant at the acceptable levels. Further, we note that politically connected and nonconnected sample firms have similar historical sales growth rates.

Table 3:

Comparison of politically connected and nonpolitically connected sample firms

Note: Table presents comparative statistics of sample firms partitioned based on their political connections. The definitions of above variables are provided in appendix A. To control for the outliers, we winsorize all the ratio variables at 1 and 99 percentiles. The values of total assets are expressed in real terms (at the constant prices of 2011) of INR millions. The mean differences are tested using the t-test for their statistical significance. The median differences are tested for their significance using Wilcoxon rank sum test. Accordingly, the p-values of t-test and rank sum test with the null hypothesis that there is no difference between politically connected and nonconnected firms are reported.

Multivariate/regression analysis

This section presents findings of our regression analysis, examining the effect of corporate political connections on the sample firms’ financial outcomes, namely market value and financial leverage. Further, we assess the moderating effect of firm- and industry-specific characteristics on the relationship between corporate political connections and firms’ market value. As described in the “Estimation Methods” section, we use Heckman's two-step procedure to control the possible self-selection problem. As per the Heckman procedure, first, we investigate the firm's choice of having political connections (equation 1) using the Probit regression estimator.Footnote ⁹⁶ We then estimate inverse mills ratio and include the same in equation 3 to account for the self-selection problem. In the second stage, we estimate equation 3 using pooled ordinary least squares estimator. Further, to account for the potential endogeneity as well as self-selection problems, we implement the Semykina and Wooldridge (Reference Semykina and Wooldridge2010) procedure. We report the findings of Heckman's second-stage estimation and Semykina and Wooldridge's (Reference Semykina and Wooldridge2010) estimation approach in table 4.

Table 4:

The influence of political connections on firms’ market value.

Note: Table presents the findings of estimating equation 3. It includes (1) second step of Heckman two-step estimation method (i.e., outcome equation corrected for self-selection by including inverse Mills ratio) and (2) endogeneity and selection corrected (i.e., second step of Semykina and Wooldridge [Reference Semykina and Wooldridge2010] approach). Time and Industry refer to the year and industry fixed effects. The numbers in the parentheses are corresponding robust standard errors. Appendix A provides the definitions of all the variables.

The superscripts of ***, **, and * represent statistical significance at the 1%, 5%, and 10% confidence levels, respectively.

The reported findings in table 4 suggest that after controlling for all the potential determinants, political connections negatively influence the sample firms’ market value. The selection and endogeneity corrected findings imply that, after controlling for all the relevant factors, politically connected sample firms on an average report 3.13% lower market value than nonconnected firms. The other determinants of firms’ market value take the expected signs. In accordance with the results of Chen et al. (Reference Chen, Firth and Xu2009) and Wu et al. (Reference Wu, Wu and Rui2010), firm size has a negative influence on the firms’ market value. As expected, firms’ operating efficiency, industry concentration and growth opportunities (measured using sales growth) have a significant positive influence on the market value of firms, whereas leverage, age, and risk influence the sample firms’ market value negatively. The reported negative influence of political connections on the sample firms’ market value supports our hypothesis: H1b. The evidence suggests that financial market participants believe that, on average, the costs of firms’ political connections outweigh the benefits from those connections. As discussed, these costs could be a combination of rent-seeking activities of politically connected board members,Footnote ⁹⁷ higher information asymmetry associated with politically connected firms,Footnote ⁹⁸ the opaqueness of their earnings numbers,Footnote ⁹⁹ lack of business acumen or professionalism in the case of connected board members,Footnote ¹⁰⁰ and contract enforcement uncertainty associated with the politically connected firms.Footnote ¹⁰¹ Accordingly, market participants report to value, on average, politically connected firms lower than the nonconnected firms.

Further, we argue that the influence of political connections is heterogeneous across the connected firms. We estimate the moderating effect of firm- and industry-specific factors such as their size, operational efficiency, financial leverage, and degree of industry concentration on the relationship between firms’ political connections and their market performance. Table 5 reports the findings of an empirical investigation into the interaction effect of firms’ specific factors on the relationship between sample firms’ political connections and their market value. The reported findings suggest that political connections continue to have negative influence on the sample firms’ market value. The significantly positive coefficient of interaction term (refer to model 5) between firm size and political connections supports our hypothesis (H2) that larger firms utilize the benefits of political connections better than the corresponding smaller sample firms. That is, an unit increase in the size of the sample firms reduces the negative effect of political connection on their market value on an average by 6.74%. Such a finding is attributable to the resource advantage of larger firms and their requirement of protection from the nonmarket forces through political connections. The reported finding is also in accordance with Goldman et al. (Reference Goldman, Rocholl and So2009) who reported that the influence of political connections is more evident in the case of larger firms. The results in table 5 suggest that operational efficiency, measured in the form of asset turnover ratio,Footnote ¹⁰² have significant positive influence on the relationship between the firms’ political connections and their market value. It implies, as hypothesized in H3, that firms with greater operating efficiency are perceived by market participants to be in an advantageous position to make use of the positive external environment facilitated by the political connections. The reported results suggest that the negative effect of political connections reduces on an average by 15.7% for an unit increase in the operational efficiency. Accordingly, financial markets value operationally better efficient politically connected firms greater than the respective less efficient firms.

Table 5:

Influence of firm- and industry-specific factors on the relationship between corporate political connections and market value of firms

Note: Table presents the findings of second step of Semykina and Wooldridge (Reference Semykina and Wooldridge2010) estimation method. Models 1–5 consider political connection variable as a binary one identifying the politically connected firm-year observations. Models 6–10 consider political connection variable as continuous variable measuring the number of politically connected board of directors. Time and Industry refer to the year and industry fixed effects. The numbers in the parentheses are corresponding robust standard errors. Appendix A provides the definitions of all the variables.

The superscripts of ***, **, and * represent statistical significance at the 1%, 5%, and 10% confidence levels, respectively.

We empirically estimate the moderating effect of leverage on the relationship between firms’ political connections and their market value. In consistent with our hypothesis (H4), the coefficient of the interaction term between political connections and leverage is negative and statistically significant. That is, an one unit increase in the leverage amplifies the negative effect of political connections on the market value by 25.9%. This evidence supports the argument that politically connected firms with higher leverage are more likely to engage in the suboptimal investment decisions, leading to their lower valuations in the market. Further, we investigate whether the political connections have differential impacts on the market value of firms operating in the industries with varied degree of concentration. The reported results, supporting our hypothesis (H5), suggest that political connections positively influence the value of firms operating in the more concentrated industries. If the industry concentration, measured by four-firm concentration ratio goes up by one unit, it reduces the negative effect of political connections on the market value by on an average 50.3%. This implies that financial markets in India believe that firms operating in more concentrated industries derive greater benefits from political connections than those operating in the less concentrated industries. This finding complements the earlier studies such as Pittman (Reference Pittman1988), Zaleski (Reference Zaleski1992), and Bhuyan (Reference Bhuyan2000). They report that firms operating in the concentrated industries are more inclined to embark on political connections as they gain more from beneficial regulations.

Case of regulated industries

Government is the major source of uncertainty for the firms operating in heavily regulated industries.Footnote ¹⁰³ Through political connections, firms operating in these industries aim to obtain favorable policy formulations and implementation. Accordingly, firms operating in these industries are more presumably appoint current or former politicians/bureaucrats to their board of directors than firms operating in the less-regulated industries. Hillman (Reference Hillman2005), examining a sample of US firms, reported that firms with political connections have better market performance, and this positive association is more prominent among firms operating in heavily regulated industries. Hadani and Schuler (Reference Hadani and Schuler2013), however, report that firms with politically connected board members are negatively associated with market-based performance measures. However, they find that board-level political connections have a positive influence on the value of firms operating in heavily regulated industries. Our preliminary summary statistics presented in table 1 suggest that firms operating in heavily regulated industries (e.g., electricity and gas, transportation, mining, healthcare) are more likely to have political connections. But, are these firms (those operating in heavily regulated industries) truly achieving their financial objectives by having politically connected members on their board? We examine this empirical question in the backdrop of our sample firms. Further, we assess if there is any heterogeneity based on firm characteristics in the impact of political connections on the market value of firms’ operating in industries with varying degrees of regulation.

We consider an industry as regulated if the firms operating in that industry face a significant influence from government policy instruments. Following Pittman (Reference Pittman1977), we consider three aspects of government's influence on the industry. They are government regulation of industry, government purchase of industry output (or government supply of industry's crucial input), and government antitrust investigation of industry.Footnote ¹⁰⁴ We employ this definition to define regulated industries. Such industries include electricity and gas, transportation, mining, petroleum and chemicals, telecommunication, construction of buildings and civil engineering, healthcare, and pharmaceuticals and medical equipment. We estimate equation 3 separately for firms operating in regulated and less regulated industries using the Semykina and Wooldridge procedure. The findings are recorded in table 6.

Table 6:

Subsample analysis: The effect of political connections on firms’ market performance among regulated and unregulated industries.

Table presents findings of second step of Semykina and Wooldridge (Reference Semykina and Wooldridge2010) approach for the subsamples classified based on the intensity of their industry regulation. Time and Industry refer to the year and industry fixed effects. The numbers in the parentheses are corresponding robust standard errors. Appendix A provides the definitions of all the variables.

The superscripts of ***, **, and * represent statistical significance at the 1%, 5%, and 10% confidence levels, respectively.

The evidence in table 6 suggests that corporate political connections have a significant negative influence on the market value of sample firms operating in both regulated and less regulated industries. Similar to the findings reported earlier for the entire sample, larger firms and firms that are operationally more efficient make use of the benefits of political connections better than their corresponding counterparts in the case of both regulated and less regulated industries. Further, we note that unlike the findings of the entire sample, firms operating in concentrated industries that are heavily regulated do not benefit from political connections. But, firms operating in the concentrated industries that are less regulated seem to benefit from the political connections in terms of their market value as we report a significant positive coefficient estimate for the interaction term of industry concentration with political connections. These findings imply that firms operating in the heavily regulated industries that are more concentrated seek political connections for the benefit of overall industry. But the firms operating in the less regulated industries that are more concentrated seek political connections for their “private benefits” such as obtaining government contracts/licenses and advance information about the possible policy changes that their competitors might otherwise capture.

Political connections and financial leverage

We examine the association between sample firms’ political connections and their financial leverage by estimating equation 4. We employ system-GMM estimator to estimate the dynamic leverage equation as it accounts for the endogeneity of input variables using the internally generated instruments. Table 7 reports the findings. The Hansen test of overidentification and difference-in-Hansen test of exogeneity validate the instruments and tests that they are exogenous. We consider both the definitions of political connection variable: political connections as a binary variable (models 1 and 2) and as a continuous variable (models 3 and 4). We also control for the conventional determinants of leverage. The reported evidence suggests that political connections do not influence sample firms’ financial leverage. The results are consistent for both the measures of political connections: binary and continuous variables. In a way, these findings do not support our hypotheses H6(a) and H6(b). They imply that both the positive and negative aspects of political connections on the firms’ financial leverage tend to offset each other in the context of our sample firms. That is, political connections do not seem to influence an average sample firm's financing choices.

Table 7:

Influence of firms’ political connections on their financial leverage

Note: Table records the results of system – GMM estimator, used to estimate the effect of political connections on the sample firms’ leverage (equation 4). The Hansen test of overidentification and the difference-in-Hansen test of exogeneity are based on the null hypothesis that all the instruments are valid and exogenous, respectively. Time and Industry refer to the year and industry fixed effects. The numbers in the parentheses are corresponding robust standard errors. Appendix A provides the definitions of all the variables.

The superscripts of ***, **, and * represent statistical significance at the 1%, 5%, and 10% confidence levels, respectively.

Additional tests

We employ at least four additional analyses to further confirm the findings presented in the previous subsections. First, following earlier studies,Footnote ¹⁰⁵ we also consider log transformation of market capitalization to measure sample firms’ market value. We again estimate equation 3, where log-transformed market capitalization as the dependent variable, using Heckman as well as Semykina and Wooldridge procedures. The results are qualitatively similar to those reported in the previous sections where we use Tobin's Q to measure the market value. Political connections continue to have negative influence on the market value of sample firms. Further, the evidence also suggests that sample firms with larger size, with greater operating efficiency, and those operating in the more concentrated industries benefit more from the political connections than their corresponding counterparts.

Second, we have measured industry concentration using four-firm concentration ratioFootnote ¹⁰⁶ and presented the analyses in the previous sections. Alternatively, researchers also use Herfindahl–Hirschman index (HHI) to quantify the industry concentration. HHI is computed by squaring the market share (in terms of sales) of each firm operating in the industry and then adding the resulting numbers. To test the robustness of our findings, we also consider HH index as a measure of sample firms’ industry concentration and reestimate the proposed econometric models. We find that the results are consistent and similar whether we employ HHI or four-firm concentration ratio to measure the sample firms’ industry concentration.

Third, we report that, in the context of India, political connections do not significantly influence the firms’ financial leverage. This finding is contradictory to the observations of earlier studies that reported a significant positive influence of the political connections on the corporate leverage.Footnote ¹⁰⁷ The anecdotal evidence in the context of India, particularly during the early 2014 blamed the corporate political connections for the overwhelming nonperforming assets of institutional lenders, suggesting the wide presence of crony capitalism in India.Footnote ¹⁰⁸ This coincided with the Indian central bank releasing the stringent guidelines for revitalizing distressed assets in January 2014.Footnote ¹⁰⁹ These guidelines outline corrective action plans for institutional lenders that necessitated early identification of distressed assets and timely restructuring of those assets. These guidelines prompted the lenders to be more vigilant while advancing loans under the political/bureaucratic pressures as such loans also attract greater attention from the media and public scrutiny. We are concerned that our findings are probably influenced by such lending preferences of institutional lenders following the release of central bank guidelines on the identification of potential distressed assets in early 2014. To examine such a possibility, we partition the sample observations at the year 2014 into two subsamples and estimate equation 4 using system-GMM estimator separately for both the subsamples. We note that the results of the subsample analysis are similar to those reported for the entire sample (table 7). Further, we also considered alternative definitions of leverage including the ratio of total borrowings to market value of assets (i.e., sum of the book value of debt and market value of equity) and reestimated the equation 4. This additional analyses confirm that political connections do not significantly influence the sample firms’ financing choices.

Finally, as we noted earlier that firms establish political connections to manage their external environment either by contributing to their election process or by employing the politically connected individuals on their board as the directors. The analyses presented in the previous sections focus on the later in identifying the politically connected sample firms. It is possible that there are a few sample firms that are accruing the benefits of political connections by donating to the political parties. Unfortunately, such donations to political parties by firms or individuals are still a taboo phenomenon in India and are not well reported/documented. However, we make use of the donations records that are made available on data depository platform,Footnote ¹¹⁰ managed by the Association of Democratic Reforms (ADR), an Indian nongovernmental organization that works in the area of electoral and political reforms. A little more than 95% of the donation records available on this platform are of the individual donors. We employ an intensive process of company name search and matching with that of our sample firms. A sample firm is then identified as politically connected in a year if it has contributed to any political party in the corresponding year or in the past. In this way, we could identify 949 firm-year observations with political connections, leading to an increase in the total percentage of firm-year observations with political connection to 29.9%. We then reexamined the effect of political connections on firms’ market value including the potential influence of firm- and industry-specific factors on this relationship. The findings are reported in the appendix D. The results of this additional analysis remain qualitatively similar to those reported in the earlier sections.