CEO age and tax planning

• DOI: http://doi.org/10.1002/rfe.1072
• Published date: 01 April 2020
• Author: Hui Liang James
• Date: 01 April 2020

Rev Financ Econ. 2020;38:275–299. wileyonlinelibrary.com/journal/rfe

|

275

© 2019 University of New Orleans

1

|

INTRODUCTION

This study examines the association between CEO age and corporate tax planning. Tax planning refers to activities of restruc-

turing transactions to increase after‐tax income and/or after‐tax cash flows. As such activities involve significant uncertainty,

tax planning is often viewed as a long‐term investment (Minnick & Noga, 2010). When carried out effectively, tax planning

leads to reduced taxes and increased shareholder’s wealth (Hanlon & Heitzman, 2010; Robinson, Sikes, & Weaver, 2010).

However, aggressive tax planning may subject firms to additional taxes, penalties, and unfavorable public scrutiny (Hanlon &

Slemrod, 2009).

CEOs play an important role in shaping firms’ tax policies by setting general directions of firms’ tax strategies (Dyreng,

Hanlon, & Maydew, 2010). CEOs have incentives to minimize tax payments when their compensations are directly linked

to accounting‐based performance (Phillips, 2003) and/or when they are offered with stock‐based compensation contracts

(Armstrong, Blouin, Jagolinzer, & Larcker, 2015; Rego & Wilson, 2012). Campbell, Guan, and Li (2018) document a positive

relation between CEO severance pay and corporate tax planning, suggesting CEO severance pay induces risk‐averse managers

to engage in more tax planning by providing contractual protection against their career concerns.1

Existing studies provide evidence showing that CEO age affects various corporate policies. Younger CEOs have greater

incentives to favorably influence the labor market perception of their quality in order to obtain higher compensation and job

security (Gibbons & Murphy, 1992; Holmstrom, 1999; Prendergast & Stole, 1996). As such, younger CEOs are more likely to

adopt risky investment and financial policies (Croci, Giudice, & Jankensgård, 2017; Li, Low, & Makhija, 2017; Serfling, 2014).

Conversely, older executives enjoy their “quite life” and prefer less risky investments (Bertrand & Schoar, 2003). Furthermore,

older CEOs are less efficient in complex cognition, new information processing, and problem‐solving (Haug & Eggers, 1991;

Mata, Schooler, & Rieskamp, 2007; Raz et al., 2004; Resnick, Pham, Kraut, Zonderman, & Davatzikos, 2003; Schaie, 1996).

Hence, the ability to engage in actions to effectively lower tax burden is expected to be impaired by age. Lastly, older individuals

Received: 27 February 2019

|

Revised: 15 June 2019

|

Accepted: 19 June 2019

DOI: 10.1002/rfe.1072

ORIGINAL ARTICLE

CEO age and tax planning

Hui LiangJames

Accounting, Finance & Business Law,

Soules College of Business,University of

Texas at Tyler, Tyler, Texas

Correspondence

Hui Liang James, Accounting, Finance &

Business Law, Soules College of Business,

University of Texas at Tyler, SCOB 350.06,

3900 University Blvd, Tyler, TX 75799,

USA.

Email: hjames@uttyler.edu

Abstract

This study investigates the association between CEO age and corporate tax planning.

Using a sample of 11,537 firm‐year observations from the fiscal year 1997–2013,

I find CEO age exerts an economically significant influence on firms’ tax policies,

incremental to economic determinants identified in prior research. Specifically, CEO

age is positively related to cash and GAAP effectivetax rates, and negatively related

to permanent book‐tax difference, suggesting that older CEOs are less likely to take

actions to lower tax burden. The results hold across different model specifications

and robustness tests to address potential bias arising from endogeneity, sample selec-

tion issue, and the confounding effect of CEO tenure.

KEYWORDS

CEO age, ethics, risk‐taking incentives, tax planning

JEL CLASSIFICATION

G30; H26; G40

276

|

JAMES

are exposed to traditional culture for a prolonged time and more likely to draw inferences consistent with ethical behavior

(Dawson, 1997; Deshpande, 1997; Mudrack, 1989; Peterson, Rhoads, & Vaught, 2001; Terpstra, Rozell, & Robinson, 1993;

Twenge & Campbell, 2008). As paying a fair amount of taxes is generally viewed as socially responsible (Freedman, 2003;

Landolf, 2006; Williams, 2007), older CEOs should be less likely to engage in tax planning.

This study investigates the influence of CEO age on corporate tax planning using a sample of 11,537 firm‐year observations

from the fiscal year 1997–2013. GAAP and Cash effective tax rates are employed to capture the cross‐sectional variation in

firms’ total tax planning (Armstrong et al., 2015; Campbell et al., 2018; Dyreng, Hanlon, & Maydew, 2008; Robinson et al.,

2010). To overcome the measurement issue in annual effective tax rates (ETRs) as indicated in Dyreng et al. (2008), I use the

3‐year average GAAP ETR (GAAP_ETR) and 3‐year average Cash ETR (CashETR). Following Armstrong et al. (2015) and

Balakrishnan, Blouin, and Guay (2012), I benchmark a given firm’s ETRs relative to similar firms with the difference in the av-

erage GAAP_ETR/CashETR between the firm’s size and industry peers and the given firm (D_GAAP_ETR/D_CashETR). The

last proxy for tax planning is permanent book‐tax differences (PBTD) (Frank, Lynch, & Rego, 2009; Goh, Lee, Lim, & Shevlin,

2016). Firms engaging in higher levels of tax planning are expected to be associated with lower tax rates, higher differences in

tax rates between industry peers and the sample firms, and higher permanent book‐tax differences.

The results show that CEO age is positively related to tax rates and negatively related to PBTD, suggesting that younger CEOs

are more likely to undertake tax policies to minimize taxes. Other things equal, firms with CEOs under 50years of age are associated

with 0.96 percentage point reduction in CashETR, 1.14 percentage point reduction in GAAP_ETR, 1.07 percentage point increase in

D_CashETR, 1 percentage point increase in D_GAAP_ETR, and 0.57 percentage point increase in PBTD, compared to firms with

CEOs over 60years of age. Moreover, firms headed by CEOs with age between 50 and 59 are associated with 0.71 percentage point

reduction in CashETR, 1.03 percentage point reduction in GAAP_ETR, 0.69 percentage point increase in D_CashETR, 0.94 percent-

age point increase in D_GAAP_ETR, and 0.42 percentage point increase in PBTD, compared to firms with CEOs over 60years of age.

Dyreng et al. (2010) find that the age of executives does not explain much of the variation in tax avoidance. However, their

sample selection and estimation method are different from this study. Following Bertrand and Schoar (2003), Dyreng et al.

(2010) estimate executives’ fixed effect on tax policies using a sample of 908 executives who were employed by more than

one firm in order to isolate the executive effect from the firm effect. As shown in Graham, Li, and Qiu (2012), the sample cre-

ated through this method is “necessarily small due to the relatively small number of executive job changes in most samples.”

Moreover, Graham et al. (2012) argue that managers who never change firms may be significantly different from those who

have moved. Hence, this sample selection process might result in sample selection bias, which could mean that the results are

not indicative of the true relation. Second, the different results may be attributable to different sample periods. The sample in

Dyreng et al. (2010) is from 1992 to 2006, but the sample in this study is from 1997 to 2013. In summary, Dyreng et al. (2010)

identifies, for managers who change jobs, a manager fixed effect associated with corporate tax planning, but the study does not

find that age is significantly associated with this manager fixed effect. In contrast, I find that for a relatively broad sample of

CEOs (not limited to managers who change firms during my study), age is consistently associated with corporate tax planning.

I conduct a variety of tests to mitigate various econometric concerns. First, the observed association between CEO age and tax

planning could arise from the systematic difference between firms led by older CEOs and firms led by younger CEOs. Second, CEO

age and tax planning may be endogenously determined by variables omitted from the model. For example, firms may hire younger

CEOs because they favor lower taxable income; thus, the relation between CEO age and tax planning is due to omitted firm fixed

effect. In both scenarios, the endogeneity between CEO age and tax planning violates the conditions of OLS regressions, leading

to inconsistent and biased coefficient estimates. I employ three identification strategies to alleviate the concern of spurious relation.

First, I apply propensity score matching to identify control firms with older CEOs but are otherwise indistinguishable from firms with

younger CEOs, and repeat the analysis with this matched sample. This method isolates a control sample of firms headed by older

CEOs from a treated sample of firms headed by younger CEOs in a way that the control sample exhibits no observable difference

from the treated sample. Second, I use an instrumental variable to create a source of exogenous variation in CEO age. Following

Serfling (2014), Cline and Yore (2016), and Croci et al. (2017), I instrument CEO age with the level of the Consumer Price Index

(CPI) in a CEO’s birth year. The rationale is that older CEOs are likely to be born in early years with lower CPI. Thus, CEO age and

CPI in the birth year should be significantly and negatively correlated, satisfying relevance restriction. Little evidence, if any, implies

CPI in CEO birth year is related to current tax planning practice adopted by the firm headed by this CEO, satisfying exclusion re-

striction. Lastly, I adopt the firm fixed effect regression to mitigate the bias arising from the time‐invariant fixed effect. Potential time

effect on the tax planning incentives is controlled by year indicator variables. The results consistently indicate that firms with younger

CEOs are associated with significantly lower tax rates and higher permanent book‐tax differences, confirming the baseline results.

Younger CEOs are more likely to be CEOs with shorter service. It is arguably possible that the finding of younger CEOs

are associated with high levels of tax planning is due to the new CEO effect: newly hired CEOs have greater incentives engage

in tax planning to increase compensation and job security. In addition, CEOs near retirement age may be more conservative in