The Market for CEOs

Peter Cziraki is Assistant Professor of Economics at the University of Toronto and Dirk Jenter is Associate Professor of Finance at the London School of Economics & Political Science. This post is based on their recent paper.

CEOs have first-order effects on firms, which makes an efficient CEO labor market important. Several influential studies argue that the market for CEOs is well described by models with perfect competition and no frictions (Tervio 2008; Gabaix and Landier 2008; Edmans, Gabaix, and Landier 2009). Other influential studies argue that firms’ demand for managerial skills has shifted from firm-specific to general (and therefore transferrable) skills (Murphy and Zabojnik 2004, 2007; Frydman 2019).

In this paper, we document actual CEO hiring practices and compare them to the predictions of these (and other) theories. For all new CEOs in the S&P 500 from 1993 to 2012, we document their prior connections to the hiring firm, whether new CEOs were raided from other firms, and how hiring choices differ across firms. We focus on the largest publicly-traded companies as they face the fewest frictions in the managerial labor market and, because of the range of their activities, are likely to require CEOs with general skills.

Our results show that S&P500 firms hire from a surprisingly small pool of candidates, and that the vast majority of new CEOs have close prior links to the hiring firm. 72% of new CEOs are promoted internally, and 8.4% are former executives or current or former board members. Thus, 80.4% of CEO hires are insiders, and only 19.6% are new to the firm. There are slightly fewer outsiders in later years, so the previously observed trend to more outsider hiring has ended.

We next show that most of the 19.6% outsiders have prior connections to the hiring firm’s board. Fifty-four percent have worked with at least one of the hiring firm’s directors, compared to only 3% for a matched sample of alternative candidates. Thus, more than 90% of new CEOs are either insiders or co-workers of its directors. This evidence is hard to reconcile with the view that CEOs are chosen for their general managerial skills and move freely across firms.

Our second set of results reveals where firms find the 19.6% outsider hires. The most striking result is the rarity of CEO raids: only 3.2% of new CEOs are poached from the CEO position at another firm. When firms poach CEOs, it is typically from firms that are three to four times smaller. This suggests severe frictions in the reallocation of CEO talent across firms. Instead of raided CEOs, most outsider hires are below-CEO executives at other (typically much larger) firms (55%) or unattached, i.e., individuals not currently in an executive position (31%).

We next analyze how hiring choices differ across firms. Most notably, larger firms are even more likely to promote internally than smaller ones. A top-quintile S&P 500 firm by market value has a 91% probability of hiring an insider, compared to 75% for a bottom-quintile firm. Consistent with prior studies, firms with low stock returns and low sales growth are more likely to hire outsiders, but even for them insiders remain the most frequent choice. Finally, former executives and current or former board members are most often chosen by firms with low stock returns, low operating performance, and low sales growth. Hence, many badly performing firms turn to former employees and directors rather than to outsiders for help.

Our final set of results explores whether differences in CEO pay might explain firms’ hiring choices. Outsiders are indeed more expensive than internal promotions. However, the differences are small compared to the scale of S&P 500 firms: in the first full year of employment, outsiders receive on average $1.5 million more than internal promotions, and the differences between raided CEOs, raided other executives, and unattached managers are even smaller. If, as prior evidence suggests, CEOs have large effects on firm value, these pay differences appear too small to explain the dominance of internal promotions or the reluctance to raid other firms’ CEOs.

The patterns we document have implications for our understanding of the CEO labor market. The most striking result is that firms hire CEOs they are already familiar with—insiders or managers the firm’s directors have worked with—more than 90% of the time. This is hard to reconcile with a competitive labor market in which all firms select CEOs from the same talent pool. Instead, CEO hiring appears to be determined by firm-specific human capital and personal connections.

Our findings also affect our understanding of CEO compensation. The outside options of both firms and CEOs appear severely limited, as firms’ effective candidate pool is small and as incumbent CEOs rarely move to other firms. This suggests an imperfectly competitive market in which pay is determined, at least in part, by CEOs and shareholders bargaining over the surplus created by each CEO-firm match.

The complete paper is available for download here.

Both comments and trackbacks are currently closed.