CEO's Religious Affiliation and Managerial Conservatism

• Author: Mufaddal Baxamusa,Abu Jalal
• Published date: 01 March 2016
• DOI: http://doi.org/10.1111/fima.12080
• Date: 01 March 2016

CEO’s Religious Affiliation

and Managerial Conservatism

Mufaddal Baxamusa and Abu Jalal∗

Weinvestigate whether managers’religious affiliations affect corporate decisions. Wehand collect

data on the religiousaff iliations of chiefexecutive off icers (CEOs) and find that firms with Catholic

CEOs have less leverage, issue debt less often, increasebusiness and geographic diversification,

and invest less than firms with Protestant CEOs. We also find that the decisions of Catholic CEOs

are associated with lower firm value. These corporate actions are also reflected in the CEOs’

personal decisions, such as owning fewer company stocksand playing less risky sports.

Prior literature has argued that CEO-specific styles of management play an important role in

shaping corporate decisions (Bertrand and Schoar, 2003; Malmendier and Tate, 2005). CEO-

specific styles stem from values, norms, habits, and tastes that are likely to be fundamentally

shaped by religion. Religion promotes a set of beliefs and actions that guide important aspects of

the lives of its followers.Religious instruction typically begins in early childhood and is regularly

reinforced throughout an individual’s life. Despite its importance, the impact of a CEO’sreligion

on corporate decisions has not been properly investigated due to a lack of data regarding CEO’s

religious affiliations. In this paper, we fill this gap in the literature.

Early economists, such as Adam Smith and Karl Marx, have discussed the importance of

religion in shaping economic behavior. Stulz and Williamson (2003) argue that cultural and

religious forces affect individual beliefs and preferences. As such, we would expect to find a

significant difference in the risk profiles of corporate decisions made by CEOs from different

religious backgrounds. In particular, Stulz and Williamson (2003), Weber (1930), and others argue

that Protestantism, when compared to Catholicism, encourages entrepreneurship and greater risk

tolerance. These fundamental differences between Protestants and Catholics are not just recent

empirical observations, but are supported by the historical developments of these religions and are

ingrained in the cultural and ethical principles of their religious identities (Egnal, 1996; Anderson,

Drakopoulou-Dodd, and Scott, 2000; Rietveld and Van Burg, 2013; Nunziata and Rocco, 2014).

In this study, we follow and build on these arguments in the literature.

Specifically, as Protestants are more entrepreneurial and Kihlstrom and Laffont (1979) show

that lower risk aversion engenders entrepreneurship, the argument suggests that Protestants are

likely to be less risk averse. For example, Weber (1930), through an analysis of the forces

behind the historical developments of modern European economies, suggests that the Protestants

were greater risk-takers than Catholics. Hirshleifer and Thakor (1992) define a preference for

safe decisions by managers as “managerial conservatism.” Therefore, we test the hypothesis

Weare grateful to the Editor of this journal, Raghavendra Rau, and an anonymous referee. Both were extremelygenerous

with their time and helped to immensely improve the study. We thank Rajesh Aggarwal, Anand Jha, John Thornton,

and seminar participants at the University of St Thomas, Suffolk University, Financial Management Association 2012

meetings, and Eastern Finance Association 2012 meetings for their helpful comments. All errorsare our own.

∗Mufaddal Baxamusa is an Associate Professor in the Department of Finance in the Opus College of Business at the

University of St. Thomas in St. Paul,MN. Abu Jalal is an Associate Professor in the Department of Finance at the Sawyer

Business School at Suffolk University in Boston, MA.

Financial Management •Spring 2016 •pages 67 – 104

68 Financial Management rSpring 2016

that Catholic CEOs are more conservative in their corporate decisions than Protestant CEOs.

Furthermore, as corporate policies based on personal biases may reduce the profitability of a

firm, we explore the implications of these decisions on firm value.

Broadly speaking, there are three sets of decisions that corporations make: 1) capital struc-

ture, 2) diversification, and 3) investments. If Catholic CEOs are more conservative in their

corporate actions, then they should decrease leverage, increase diversification, and decrease

investments (e.g., research and development (R&D)). Furthermore, the behavioral consistency

theory (Cronqvist, Makhija, and Yonker, 2012) states that individuals tend to behave consistently

across different situations. As a result, these differences in managerial conservatism should also

be visible in the personal actions of the CEOs, such as CEO ownership of the shares of the firms.

To test these hypotheses, we hand collect CEOs’ self-reported religious affiliations from

different volumes of Marquis Who’s Who. We identify the religious affiliations for 457 CEOs

from 1992 to 2010. There does not appear to be an obvious selection bias as there are no

statistically significant differences on a number of dimensions between our sample and that from

ExecuComp. Specifically, the age and tenure of CEOs in our sample are similar to those of

the average ExecuComp CEOs. Moreover, the average firm in our sample is similar to that in

ExecuComp in terms of size, leverage, business segments, capital expenditures, market-to-book

ratios, and stock return volatility. Our empirical estimates indicate that firms with Catholic CEOs

have lower leverage, issue debt (equity) less (more) frequently,increase diversif ication measured

through business segments and geographic segments, invest less in both capital expenditures and

R&D, and personally ownfewer shares in their fir ms. Taken together, these results are consistent

with the idea that Catholic CEOs tend to be more conservative in their corporate decisions than

Protestant CEOs. Additionally,we f ind that these corporate decisions are significantly negatively

associated with firm valuation. After Catholic CEOs undertake these decisions, their fir ms exhibit

lower Tobin’s Q and reduced profitability.

There are studies that question the empirical methodologies employed by earlier work investi-

gating managerial style, including those of Bertrand and Schoar (2003). Fee, Hadlock, and Pierce

(2013) find that the standard F-test on a joint set of manager-specific dummy variables often

employed in prior studies to identify managerial style effects is inappropriate and suffers from

endogeneity concerns. In those studies, the CEO dummy variable used when the CEO changes

firms may be picking up similarities between the old firm and the new fir m and not managerial

style. In our study, wedo not employ this methodology. Instead, we match the observations using

propensity scores to compare very similar firms, except that they differ in the religious affiliation

of their CEOs. This method has been found to be generally empirically reliable (Smith and Todd,

2005; Roberts and Whited, 2012).

We undertake a number of additional steps to correct for possible endogeneity concerns.

Following Guiso, Sapienza, and Zingales (2003), we use the Second Vatican Council as an

instrument. The Second VaticanCouncil (or VaticanII) has had a substantial effect on Catholicism.

For example, the use of local languages in Mass was allowed and laymen were encouraged to

participate in certain rituals. These changes resulted in Catholics having a clearer understanding

of the Church’s teachings and traditions. Guiso et al. (2003) suggest that Catholics born after the

creation of Vatican II received a different education and style of preaching than those born prior

to it. In contrast, there should be no such effect on Protestant CEOs. Our findings remain robust

to this methodology.

Furthermore, to account for unobserved factors that are time invariant, we present robustness

tests that include firm f ixedeffects. Our results continue to hold, suggesting that the estimates are

not driven by a nonrandom assignment of CEO religiosity. Additionally, we conduct difference-

in-difference tests to identify the effect of CEO change on corporate policy. First, we compare

Baxamusa & Jalal rCEO’s Religious Affiliation and Managerial Conservatism 69

the changes in characteristics between firms that experience a change in CEO (from Protestant to

Catholic) and the firms where the new CEO belongs to the same religion as the old CEO. Then,

we compare the changes in characteristics between the firms that experience a change in CEO

(from Catholic to Protestant) and the firms where the new CEO belongs to the same religion as

the old CEO. Our contention that Catholic CEOs are more conservative in their decisions and

policies continues to hold.

We perform out-of-sample tests using Hofstede’s (2001) country-level cultural measures to

evaluate whether these findings regarding the differences in risk-aversionof Catholic and Protes-

tant CEOs are robust. We find that the uncertainty avoidance index is significantly higher for

Catholic countries, as compared to Protestant countries. This indicates that the behavior of the

CEOs in our sample is representative of the preferences of the general Catholic and Protestant

populations. Finally, we examine the preferences of CEOs for sports that may lead to personal

injury. We find that Protestant CEOs are more likely to engage in riskier sporting activities.

This study provides evidence that the religious affiliations of CEOs affect corporate decisions

through conservative corporate policies. The primary implication of our findings is that a CEOs’

religion can help to partially address some of the empirical puzzles regarding why firms in

the same industry that are similar in size, opportunities, and governance display significant

differences in capital structure, diversification, investments, and CEO stock ownership.

There are a number of recent studies that empirically investigate the relationship between the

religiosity in the location of the firm and corporate decision making concer ning investments

(Hilary and Hui, 2009), executive compensation (Kuhnen and Niessen, 2012), option grants to

corporate executives (Kumar, Page, and Spalt, 2011), earnings management (Dyreng, Mayew,

and Williams, 2012), and tax avoidance (McGuire, Omer, and Sharp, 2010). These studies use

county-level measures of religiosity,including differentiating between Catholicism-majority and

Protestant-majority counties, and draw implications for firm behavior. However, these studies

propose that the religious beliefs and viewpoints of the majority of the inhabitants in a county

will be reflected in corporate decisions without demonstrating a channel through which the

county’s culture affects firm decisions. For example, if a Catholic CEO in a Protestant majority

county adopts policies that do not conform to the Protestant culture, it is not clear how the

Protestant majority of that county will enforce its culture on the CEO. In fact, our results indicate

that the religious beliefs and viewpoints of the majority of the inhabitants in a county are not

always reflected in corporate decisions. In our study, the channel through which culture affects

corporate decisions (i.e., the preferences of the CEOs) is explicit and well documented in the

corporate finance literature.

Another major drawback of the literature is that it uses a county-level variable to proxy

for a manager’s preference. However, any county-level variable represents an array of county

characteristics that may lead to possible identification biases making these results less reliable.

Our study makes a unique contribution to the literature by hand collecting individual CEO’s

religious affiliations and testing the hypothesized relationships directly.

The rest of the study is structured as follows. Section I reviews the literature and presents the

predictions. Section II describes the data collection and variable construction. Section III tests

the predictions linking corporate decisions and CEO’s religious affiliations. Section IV subjects

the results to robustness tests, while we provide our conclusions in Section V.

I. Literature Review and Empirical Predictions

There is a vast strand of literature, starting with Adam Smith that explores the relationship

betweenreligion and economic outcomes. In his seminal work, Weber (1930) arguesthat economic