What makes a good CEO?

My AFR oped today looks at some new research on what makes a good business leader. Full text over the fold.

Execution Beats Teamwork
Australian Financial Review, 24 March 2009

Since the 1970s, Roy Morgan has surveyed Australians to ask them whether various professional groups are ethical and honest. When it comes to business executives, attitudes seem to follow the economic cycle. In good times, we are more likely to rate CEOs as ethical and honest. As the economy turns down, so does our view of executives.

Unless you think that all recessions are caused by a rapid decline in the quality of business leadership, there appears to be a tendency to give bosses too much blame in recessions (and too much credit in booms). True, executive incompetence probably mattered more in this downturn than in the typical slump. But in 2009 (as in the early-2000s), it remains the case that some executives are lazy while others are energetic; some are talented while others are in the wrong occupation.

One way of viewing executive bonuses is that they reflect an attempt on the part of shareholders to get executives to work harder. But the other part of the equation is that the choice of CEO matters. A company that chooses a great CEO might also persuade that person to put in more effort. But for the firm that picks a dud CEO, no bonus system in the world will save its share price.

How can we know who makes a good leader? The famously faddish management literature often claims that there is a single ‘secret’ to leadership. To believe airport business books, this could be anything from creativity to time management, strength to sensitivity. Yet while plenty of people have theories about business leadership, few of these theories have any strong statistical evidence to back them up.

Using an intriguing new dataset, US economists Steven Kaplan, Mark Klebanov and Morten Sorensen shed new light on an old problem. In 2000-2006, a group of private equity investors insisted on all their CEO candidates undergoing extensive personality testing. Using data from 316 personality testing interviews – each taking around four hours – Kaplan and co-authors are able to empirically test whether the assessments matter.

Comparing the subsequent performance of the firm (as measured by investors’ assessments and public information such as bankruptcies or IPOs), it turns out that personality testing is a robust predictor of CEO performance. While this may come as a surprise to those who have rolled their eyes as they and their co-workers were classified into the Myers-Briggs quadrants, this result probably reflects the fact that the CEO assessments were much more extensive than the typical multiple-choice psychological survey.

The researchers then test whether successful CEOs are more likely to be people who excel in execution-related capabilities (“aggressive”, “fast mover”, “persistent” and “proactive”) or team-related abilities (“teamwork”, “listening skills”, “open to criticism” and “treats people with respect”). As archetypes of these two sets of skills, they cite General Electric CEOs Jack Welch and his successor Jeffrey Immelt. While Welch was often referred to as “Neutron Jack”, Immelt was known for holding “dreaming sessions” with clients and building “imagination breakthrough” teams.

In theory, either set of skills could be important. Yet when Kaplan and co-authors turned to the data, they found clear evidence that execution-related capabilities mattered more than teamwork. Leaders who were conscientious and driven performed significantly better than good listeners who exhibited modesty.

The US finding that execution-related skills matter more than teamwork accords with results from Australian National University researchers Deborah Cobb-Clark and Michelle Tan, who look at the relationship between a five-factor personality metric and labour market outcomes. In the Australian workplace, they find that those who become managers are less likely to be agreeable and more likely to be conscientious. And if they focus just on managers, it seems that those who are less agreeable tend to earn higher wages. In other words, not only are bosses as a group less agreeable than people in other occupations; successful bosses are less agreeable than unsuccessful bosses.

(One caveat: the measure of success used in these studies is corporate value and managerial earnings. However, these narrow measures don’t capture the impact that a congenial manager can have on the happiness of those around them. Indeed, it is quite possible that ‘successful’ managers in these studies actually made their workers’ lives more miserable than ‘unsuccessful’ managers.)

To labour economists, hiring the right candidate and rewarding good performance are closely related. When it comes to executives, both are important. But if bonuses are curtailed, then the hiring decision will become even more critical. In such an environment, we can expect firms to spend more resources working out what kinds of people make the best CEOs.

Andrew Leigh is an economist in the Research School of Social Sciences at the Australian National University.

This entry was posted in Labour Economics. Bookmark the permalink.

3 Responses to What makes a good CEO?

  1. Robert Wiblin says:

    Isn’t the real question whether or not those bonuses really are going to executives who perform well – or whether they’re just rigged for the executive to pay out almost regardless. This seems possible if executives have a lot of control over shareholders (who may have only diffuse concern in the outcome) in the decision making process.

  2. TerjeP (say tay-a) says:

    Indeed, it is quite possible that ‘successful’ managers in these studies actually made their workers’ lives more miserable than ‘unsuccessful’ managers.

    If unsuccessful means letting the company go down the plug hole then I doubt that successful managers make workers more miserable.

    A CEO in a large company that spends too much time listening is going to have difficulties. A high degree of information filtering is necessary simply to avoid cognitive overload. The flips side is that when something really bad is brewing the warning signs may get missed. There will always be great batsman who get bowled out. Likewise we are always going to have a percentage of good CEOs who stuff up.

    My understanding is that in the USA caps on fixed salary renumeration is part of what has driven the move to bonus based income incentives. The problem is that a lot of these incentives focus on short term performance. And the other problem is that this form of renumeration is addictive. Fixed salary renumeration should be uncapped.

  3. TerjeP (say tay-a) says:

    p.s. Corporate governancy based on shareholder appointed company boards are a form of representative democracy. The imperfections are rather plain to see.

Comments are closed.