Entrenchment or efficiency? CEO‐to‐employee pay ratio and the cost of debt

• Published date: 01 August 2021
• Author: Katsiaryna Bardos,Steven E. Kozlowski,Michael R. Puleo
• Date: 01 August 2021
• DOI: http://doi.org/10.1111/fire.12256

DOI: 10.1111/fire.12256

PAPER SUBMITTED FOR

ACCELERATED REVIEW

Entrenchment or efficiency? CEO-to-employee

pay ratio and the cost of debt

Katsiaryna Bardos Steven E. Kozlowski Michael R. Puleo

Dolan School of Business | Department of

Finance, Fairfield University, Fairfield,

Connecticut

Correspondence

MichaelR. Puleo, Dolan School of Business |

Departmentof Finance, Fairfield University,

1073N Benson Rd, Fairfield, CT 06824.

Email:mpuleo@fairfield.edu

Abstract

Using new data on S&P 1500 firms’ chief executive offi-

cer (CEO)-to-employee pay ratios disclosed by mandate of

Section 953(b) of the Dodd–Frank Act, we examine the

effect of within-firm pay inequality on bond yield spreads.

We find a significant negative relation between industry-

adjusted CEO-to-employee payratio and yield spreads while

controlling for covariates and endogeneity. This result is

strongest in financially constrained, labor-intensive, and

small-to-medium-sized firms. The evidence supports the

incentive-provision explanation of CEO-to-employee pay

disparity,reflecting efficient CEO compensation rather than

rent extraction. We also document selection bias in self-

reported pay ratios, highlighting the efficacy of the Dodd–

Frank provisions.

KEYWORDS

cost of debt, Dodd–Frank Act, executive compensation, pay

disparity

JEL CLASSIFICATION

G28, G30, G31

1INTRODUCTION

Reduction of inequality has been named as one of the goals for sustainable development by the United Nations.1

Inequality has also been front and center of the political platforms of many political leaders around the world.2The

rising gap between chief executive officer’s (CEO)pay and the pay of an average employee has been at the center of

1Seehttps://www.un.org/development/desa/disabilities/envision2030.html

2Seehttps://www.axios.com/2020-democrats-economic-inequality-8e4aafc1-a4a1-4921-bede-e9788dccc8a5.html

Financial Review. 2021;511–533. wileyonlinelibrary.com/journal/fire ©2021 The Eastern Finance Association 511

512 BARDOS ET AL.

attention. MarketWatch reports that among the U.S. top 350 businesses, CEO-to-employeepay ratio has increased

to 278:1 by 2018, up from 20:1 in 1965.3However, academic research on the effects of payinequality within firms

has been limited by data availability.Prior to the Dodd–Frank Act increasing mandatory disclosure of compensation

information enacted in 2015, companies did not haveto disclose median employee compensation. Consequently, prior

research on pay inequality faced severelimitations stemming from small sample size and selection bias.

After the passage of the Dodd–Frank Act in 2015, the Securities and Exchange Commission (SEC) adopted a rule

requiring public companies to disclose median employee pay and its ratio to CEO pay beginning in fiscal year 2017.

This paper is the first to use median worker compensation data reported by all S&P 1500 companies following the

mandate of the Dodd–Frank Act to examinewhether CEO-to-employee pay ratio affects firm cost of debt. The inves-

tigation furthers ongoing research on the implications of CEO pay increases over time (Gabaix & Landier,2008;Vo&

Canil, 2019) and extendsthe literature on executive compensation and corporate borrowing costs (Brockman, Martin,

& Unlu, 2010; Bryan, Nash, & Patel, 2006; Liu & Jiraporn, 2010). Our study is also one of the first to use this newly

available comprehensive data set within the broader academic literature. Related studies making use of this data set

include Bardos, Ertugrul, and Kozlowski (2020), who examine the determinants of CEO-to-employee pay ratio and

its effect on productivity and firm performance, and Alan, Bardos, and Shelkova (2020), who examine if CEO gender

explainsCEO-to-employee pay ratio. Additionally, Jung, Kim, Ryu, and Shin (2018) examinewhich firms are more likely

to disclose a supplementary pay ratio and the incentivesfor doing so.

Two concurrent working papers examine the association of CEO-to-employeepay ratio and the cost of debt and

find contradictory results. Lei (2017) finds that increased CEO-to-employee pay disparityis associated with a higher

probability of credit rating upgrades and reductions in the cost of debt. Conversely,Huang, Huang, and Yu (2018)find

a positive relation between CEO-to-employee payratio and bond yield spreads for seasoned corporate bonds, partic-

ularly among financially constrained firms. However,both studies use the same pre Dodd–Frank measure of CEO-to-

employee pay ratioover similar time periods. This pay ratio proxy equals total self-reported labor costs (in COMPUS-

TAT)less the total compensation of the top five highest paid executives (from ExecuComp), divided by the number of

employees reported in COMPUSTAT.However, the self-reported data comprise less than 10% of public firms for the

1992–2014 sample period and thus suffers from potentially severe sample selection and omitted variable biases. For

instance,Lei (2017) finds that larger firms with higher leverage and physical capital intensity but lower market-to-book

ratio and sales per employeeare more likely to report labor expenses and the number of employees.

Wereexamine the relation between CEO-to-employee pay ratio and the cost of debt using newly available data fol-

lowing added disclosure requirements for executiveand employee compensation. The mandate applies to all publicly

traded firms beginning in 2017 as stipulated in Section 953(b)of the Dodd–Frank Act. Several key features distinguish

our measure of CEO-to-employeepay ratio from that used in existing studies examining the cost of debt. First, manda-

tory reporting requirements ensure data availability for the full set of S&P 1500 constituents. Second, our measure

uses median employeecompensation to compute the CEO-to-employee pay ratio, whereas Lei (2017)andHuangetal.

(2018) approximate mean employee pay bysubtracting the compensation of the top five executives from total labor

costs and dividing by the number of employees. This may distort typical employeecompensation if the firm has many

highly paid executivesbeyond the top five or more generally if the company’s pay distribution is highly skewed. Finally,

our data provide industry-median employeecompensation available at the firm level to compute an accurate measure

of industry-adjusted pay ratio.4

Existing literature offers several predictions regarding the relation between CEO-to-employee pay disparity and

the cost of debt. Mueller,Ouimet, and Simintzi (2017) discuss the incentive-provision and talent assignment hypothe-

ses to explain within-firm pay inequality. Incentive provisionpredicts that high CEO-to-employee pay disparity will

3“CEOsAre Paid 278 Times More than the Average U.S. Worker.”MarketWatch, August 31, 2019, https://www.marketwatch.com/story/ceos-are-paid-278-

times-more-than-the-average-us-worker-2019-08-15.

4Lei (2017) computes industry-adjusted CEO-to-worker pay ratio using industry mean wage rates providedby the Bureau of Labor Statistics (BLS) at the

four-digitNorth American Industry Classification System (NAICS) level.