Local religious beliefs and municipal bond market outcomes

• DOI: http://doi.org/10.1111/fima.12271
• Date: 01 June 2020
• Published date: 01 June 2020
• Author: Alex Annan Abakah

DOI: 10.1111/fima.12271

ORIGINAL ARTICLE

Local religious beliefs and municipal bond market

outcomes

Alex Annan Abakah

Department of Accounting and Finance, School

of Business, St. John Fisher College, Rochester,

New York

Correspondence

AlexAnnan Abakah, Department of Accounting

andFinance, School of Business, St. John Fisher

College,3690 East Avenue, Rochester, NY 14618.

Email:aabakah@sjfc.edu

Abstract

This study investigates whether religion-induced risk aversion

affects municipal bond market outcomes from 1990 to 2017. The

results indicate that local government bonds issued from U.S. coun-

ties with a high Catholic-to-Protestant population ratio have lower

credit risk ratings and lower yield spreads, and are less likelyto have

credit enhancement. The results stand up to additional tests. I con-

trol for issuer’s county political party affiliation and state term lim-

its, and continue to find significant effects. The effects are not driven

by the issuer’s county fiscal policies. Furthermore, the effects per-

sist when I use an alternate specification that controls for omitted

factors that are time invariant. Overall, my evidence suggests that a

bond issuer’s religion-induced risk aversion playsa significant role in

the pricing of local government bonds.

1INTRODUCTION

State and local government bonds provide funding for critical infrastructure that affects daily lives. These municipal

bonds finance airports, roads and highways, bridges, schools, water and sewer facilities, and other infrastructure that

supports the economy. These projects are important to the survival of existing enterprises and the creation of new

enterprises. The municipal bond market plays an important role in the financial market that allocates money from

investors to state and local governmentsto enable them to undertake infrastructure projects. Unlike corporate bonds,

municipal bonds are directly linked to residents in the local government area.1If residents are less tolerant of risk

taking (i.e., more risk averse), they are likely to elect local governmentofficials who exhibit less risk-taking behavior.

Therefore, it is reasonable to assume that, all things being equal, these elected decision makers are more likelyto make

financial and economic decisions that reflect the preferences of the residents they serve.

The literature has extensively examinedthe impact of religion on personal behavior and the extent to which social

norms tied to religious identities influence economic outcomes (see, e.g., Becker & Woessmann, 2009; Lehrer,2004).

c

2019 Financial Management Association International

1Thefocus of this study is on bonds issued by local governments. Later in the study,I examine the relation between state-level religious beliefs and state-issued

bonds.

Financial Management. 2020;49:447–471. wileyonlinelibrary.com/journal/fima 447

448 ABAKAH

Studies link religiosity to higher ethical behavior and lower risk taking by corporate executives and employees(e.g.,

Hilary & Hui, 2009; Longenecker, McKinney, & Moore, 2004). However, a growing literature shows that risk-taking

behavior differs across religious beliefs. For instance, recent empirical studies provide mixedresults about the effect

of Catholic and Protestant beliefs on risk-taking behavior (see, e.g., Baxamusa & Jalal, 2016; Gao, Wang,& Zhao, 2017;

Kumar, Page, & Spalt, 2011; Shu, Sulaeman, & Yeung, 2012). This paper extends the literature by investigating the

impact of local religious beliefs on municipal bond market outcomes. Tothe extent that the effect of local religious

beliefs on corporate executivesand employees is reflected in a firm’s corporate culture and its behavior, I expect local

religious beliefs to havea direct impact on credit risk via the risk-taking behavior of bond issuers, who are local govern-

ment officials elected from the local resident population.2Specifically,I examine the effect of religious beliefs on credit

risk ratings, credit enhancement choice, and yield spreads.

To shed light on the influence of local religious beliefs on municipal bond marketoutcomes, I examine its impact

on credit risk ratings and credit enhancement choice using new local government bonds issued from 1990 to 2017. I

measure local religious beliefs (CPratio) as the ratio of the number of Catholic adherents to the number of Protestant

adherents in a county.3Icontrol for bond characteristics that include, among others, proceeds, maturity, general obli-

gation (GO) bonds, and negotiated bonds. Furthermore,I control for county-level economic and demographic factors,

such as per capita income, unemployment rate, education level,gender, age, race, population size, and ethnic diversity.

I also control for corruption, as bonds issued from corrupt states are associated with higher credit risk ratings and are

likely to have credit enhancement (Butler, Fauver, & Mortal, 2009). Consistent with recent empirical evidence (see,

e.g., Baxamusa & Jalal, 2016; Gao et al., 2017), the results show that CPratio has negative effect on credit risk ratings,

indicating that bonds issued from counties with a high Catholic-to-Protestant population ratio have lower credit risk

ratings. Credit enhancement improves bond issuer’s credit risk (i.e., reduces default risk of municipal debt), thereby

lowering the credit risk rating and costs of borrowing. Whereas bond issuers with lower credit risk are less likely to

purchase credit enhancement when they issue bonds, those with higher credit risk are more likely to purchase credit

enhancement to improve credit quality,ceteris paribus. If high CPratio leads to lower credit risk, I expect bonds issued

from counties with high CPratio to be less likely to have credit enhancement. The hypothesis is strongly supported. I

find that a one-standard-deviation increase in Catholic-to-Protestant population ratio decreases the odds of purchas-

ing credit enhancement by 12.16%. Also, the effect remains even after controlling for credit risk rating.

Next,I investigate whether CPratio has an effect on yield spreads. In addition to the previous controls, I include term

slope and the 1-year Treasury rateas controls (see, e.g., Chen, Lesmond, & Wei, 2007; Collin-Dufresne, Goldstein, &

Martin, 2001). I find that high CPratio has a negative impact onyield spreads. Moving from the first quartile to the third

quartile of CPratio decreases yield spreads by 2 basis points. Furthermore, I probe whether CPratio has an additional

effect on yield spreads beyond credit enhancement and credit risk ratings. I observe that the CPratio effect persists

after controlling for these two variables.

The results are robust to several additional tests. One potential concern is that the fiscal condition arising from the

bond issuers’ political structure and conservative beliefs might be driving the main results. States adopt gubernatorial

and legislative term limits to reduce wasteful government spending and control governmentalagency costs. However,

the literaturedocuments that both term limits are positively associated with government spending and that these term

limits lead to greater volatility in state fiscal activity (see, e.g., Besley& Case, 1995; Crain & Tollison, 1993). Abakah and

Kedia(2018) hypothesize that this greater fiscal volatility increases the risk of municipal debt and should be associated

with higher yields. They empirically examine this hypothesis and find that municipal bonds issued from states with

gubernatorial and state legislative term limits have significantly higher yields, whereas bonds issued from states with

neither term limits have lower yields. As the political structure influences municipal bond yields, I control for this and

continue to find significant results for the effect of local religious beliefs on local government bond outcomes. There is

2Althoughit is common for corporate executives and employees to live outside the county where the firm’s headquarters is located, local governmentofficials

liveamong the residents in the county.

3Thismeasure of local religious beliefs is standard in the literature (see Gao, Wang, & Zhao, 2017; Shu, Sulaeman, & Yeung, 2012).