CEO Power and Firm Performance under Pressure

• Author: Sabatino (Dino) Silveri,Vikram K. Nanda,Seonghee Han
• Date: 01 May 2016
• Published date: 01 May 2016
• DOI: http://doi.org/10.1111/fima.12127

CEO Power and Firm Performance

under Pressure

Seonghee Han, Vikram K. Nanda, and Sabatino (Dino) Silveri∗

Are powerful chief executive officers (CEOs) more effective in responding to pressure from the

economic environment?Concentrating decision-making power may facilitate rapid decision mak-

ing; however, the quality of decision making may be compromised, with severe consequences for

the firm if a powerful CEO is less likely to receive independent advice or to have her decisions

scrutinized. Weempirically investigate the performance of firms with powerful CEOs when indus-

try conditions deteriorate. We focus on industry downturns as these represent an exogenous shock

to a firm’s environment and on settings in which CEO power and access to quality information

is likely more consequential: innovative firms, firms with relatively little related-industry board

expertise, firms operating in competitive industries, and firms operating in industries character-

ized by relatively greater managerialdiscretion. In each of these settings we find powerful CEOs

perform significantly worse than other CEOs, suggesting contexts in which centralized decision

making is potentially of greater concern.

We investigate the effect of CEO power on firm performance. The upper echelons literature

emphasizes the role of top management teams, rather than just CEOs, in determining the strategy

and performance of firms (Hambrick and Mason, 1984). Firms differ, however, in terms of

the balance of power and decision-making authority between the CEO and other executives

and the board. The literature suggests that there may be trade-offs in terms of the costs and

benefits of vesting greater decision-making authority with the CEO. On the benefits side, there

can be efficiency gains from having a powerful CEO. These CEOs can expedite the decision-

making process, resulting in a more timely response to problems or to anticipated changes in

market conditions (Finkelstein and D’Aveni, 1994; Boyd, 1995). The downside of concentrating

decision-making power in the CEO is that CEOs may be more likely to act unilaterally with

less input from the board or other managers (Eisenhardt and Bourgeois, 1988; Haleblian and

Finkelstein, 1993). Theory suggests that when the available information is more uncertain and

noisy, a more decentralized decision-making process tends to produce better outcomes (Sah and

Stiglitz, 1986).

Weempirically investigate whether CEO power is more beneficial in crisis situations, when the

timeliness of the CEO’s response to problems is more likely to be important. We examine decision

making in contexts that differ in terms of the quality of information that might be available. In

Weare grateful to an anonymous referee whosecomments and suggestions greatly improved the paper and to Raghu Rau

(Editor) for additional comments and suggestions. Weare also grateful to Murali Jagannathan, Alminas Zaldokas, and

seminar participants at the following conferences for helpful comments: FMA Annual Meeting (2013), FMA European

Conference (2013), KoreaAmerica Finance Association–Korean Finance Association Joint Annual Conference (2013),

Midwest Finance Association Annual Meeting (2013), and Southwest Finance Association Annual Meeting (2013). We

are solely responsiblefor any errors.

∗Seonghee Han is an Assistant Professor in the College of Business at Frostburg State University in Frostburg, MD.

Vikram K. Nanda is a Full Professorin the Naveen Jindal School of Management at the University of Texas at Dallas,

TX. Sabatino (Dino) Silveri is an Assistant Professorin the Fogelman College of Business & Economics at the University

of Memphis in Memphis, TN.

Financial Management •Summer 2016 •pages 369 – 400

370 Financial Management rSummer 2016

evaluating the performance of CEOs in crisis situations, it is important to examine crises that

are driven by conditions exogenous to the firm. The reason is that if a firm is in trouble because

of poor decisions by a low-ability CEO, this may be associated with a poorer response as well.

Additionally, a firm’s prior performance can affect the level of a CEO’s power if the firm grants

the CEO greater power on account of past performance (Daily and Johnson, 1997; Hermalin and

Weisbach, 1998). Endogeneitycan thus make it diff icult to distinguish the impact of power from

the characteristics of CEOs that enabled them to acquire power in the first place. To circumvent

these concerns we focus on “shocks” to an industry, defined as significant industry downturns:

the notion is that such downturns are outside the control of any one firm or CEO and, hence,

exogenous. This allows us to draw conclusions relating to the impact of CEO power on firm

value, at least under conditions of economic stress.1Tocomplete the analysis, we also investigate

the impact of CEO power on firm value during industry upturns.

In our empirical analysis, we use a sample of firms drawn from the ExecuComp database over

1992-2012 to explore the performance of powerful CEOs across four settings. The first setting is

that of technological innovation.CEOs have fir m-specific knowledge and this maybe par ticularly

important to innovative firms.2However, these are also contexts in which the quality of decision

making is likely to benefit from the advice from board members or from managers within the

firm. Although a powerful CEO may enable swifter decisions, such decisions may be of inferior

quality in the context of innovative firms.

The second setting we consider is that of competitive industries. Although the intensity of

product market competition may benefit from swift action by a powerfulCEO, the risks may also

be larger. A poorly informed decision in this setting mayhave more negative value consequences

if the firm’s rivals can aggressively take advantage of the firm’s missteps.

The third setting we focus on is that of industries with greater managerial discretion (Finkelstein

and Hambrick, 1996). The influence of powerful CEOs on firm outcomes—whether beneficial

or otherwise—will be greater in these settings. To the extent that information quality for decision

making is poorer when the CEO is powerful, weexpect such f irms to haveworse outcomes when

there is greater managerial discretion.

The final setting is that of the board’s related-industry expertise (i.e., whether the board of

directors has expertise in the firm’s upstream or downstream industries). Evidence suggests that

directors from related industries bring valuable knowledge that helps firms overcome informa-

tional gaps. For example, firms with such directors are better at anticipating and coping with

industry-level sales shocks (Dass et al., 2014). When a firm lacks related-industry directors, the

negative consequences are likely to be more severe when the firm has a powerful CEO, if such

firms have a reduction in the gathering and sharing of information.

For our measure of CEO power, we draw upon Finkelstein (1992) and construct a composite

measure of CEO power based on seven variables: CEO Pay Slice, Duality, Triality, Tenure,

Ownership, Dependent Directors, and Founding Family. We examine the impact of industry

shocks on the performance of firms with powerful and nonpowerful CEOs, where an industry

1We note that the shock, in and of itself, does not necessarily solve the problem of distinguishing the impact of CEO

power from the characteristics that enable CEOs to acquire power. We deal with this issue more directly by employing

a propensity score matching technique, an endogenous treatment effects model, and a falsification test, all of which we

discuss in greater detail later.

2Brickley, Coles, and Linck (1999) make a similar argument for information-sensitive firms. Fama and Jensen (1983)

argue that specific information is detailed information that is costly to transfer, and the literature suggests innovative

firms have more specific information unknown to outsiders (Harris and Raviv, 1991; Graham and Harvey, 2001).

Han, Nanda, & Silveri rCEO Power and Firm Performance under Pressure 371

Figure 1. Average Changes in the (Demeaned) Market-to-Book Ratio during

Industry Shock Years

This figure is a plot of the average change in the (demeaned) market-to-book ratio (M/B) during shock

years for the petroleum and natural gas industry and the computer industry (Fama and French,1997; industr y

classifications 30 and 35, respectively). Based on the sample median of the CEO Power Index, firms in

each industry are classified as Powerful CEOs (abovethe sample median) or Other CEOs (below the sample

median). The vertical axis is the average of the demeaned M/Bduring years in which aggregate industry

sales decline by 5% or more (i.e., industry shock years). See Section II or Table I for a full description of

the variables.

-.6 -.4 -.2 0.2

retupmoCsaGlarutaN&muelorteP

Powerful CEOs Other CEOs Powerful CEOs Other CEOs

shock is defined as a 5% or greater decrease in aggregate industry sales.3We identify a total

of 66 industry shock years distributed across 31 industries over our 1992-2012 sample period.

We find, in each of the settings we explore, that CEO power has a negative effect on firm value.

Hence, there is no evidence that the ability of powerful CEOs to act more rapidly is beneficial

for shareholders—even in a crisis situation. In the event of positiveindustr y shocks, we find that

powerful CEOs perform no better than other CEOs. This suggests that our results are not the

outcome of powerful CEOs taking on greater risks.

As an illustration of our analysis, Figure 1 focuses on two industries where one might expect

the impact of CEO power to differ substantially: the computer industry and the petroleum and

natural gas industry (Haleblian and Finkelstein, 1993). We plot the impact on firm value indicated

by the average change in the market-to-book ratio (M/B) during shock years for firms in these

industries. Before computing the average change, we demean the market-to-book ratio (M/B)for

comparison purposes. Figure 1 plots the average change for powerful CEOs versus other CEOs

3Wefollow Mitchell and Mulherin (1996), among others, in using industry sales to define industr y shocks. As wediscuss

later, our results are robust to using a 10% cutoff and to using alternative measures.