Sunday, 13 September 2009

CEOs seek external advice – if you pay them for it…

There is ongoing debate whether performance related pay for top managers – in the form of stock ownership, options, or other types of financial incentives – actually works. We know it alters their behavior but does it improve it?

I’ve quoted some of the research in this area before but, in a way, whether or not it does, it remains a bit strange that top managers would need performance related pay. As I have said before, do you really want someone at the helm of your company if he or she only works hard and smart if they are directly rewarded for it? On the other hand, I have to admit, no matter how rhetorical this question is intended, I do guess it is only human…

It is only human that our behavior is altered due to performance related pay; and you and I are probably no exception. The trick then, of course, is to get the right measurement system, and perhaps to no overdo it; too much performance related pay may alter the behavior of top executives in ways you had not quite in mind when putting the measures in place! We’ve seen ample examples of that over recent years…

So, how might it bias top executives behavior in useful ways? Professors Michael McDonald from the University of Central Florida, Poonam Khanna from Arizona State University, and Jim Westphal from the University of Michigan examined an intriguing aspect of CEO behavior, and that is their inclination to seek advice from others.

CEOs often seek advice on strategic issues from executives of other firms. However, we also know from research that – just like humans – they are often inclined to solicit that “advice” from friends and other people who are just like them. In such cases, it is not really genuine advice-seeking, but it serves more in a self-confirmatory fashion; people seek confirmation that what they are doing is right, and what better way to get that by asking the opinion of your friends and look-a-likes.

To examine which CEOs engage in this pseudo-advice seeking and which ones truly turn to people who might actually disagree with them, McDonald and his colleagues surveyed 225 large American industrial and service firms. They managed to obtain information on how often their CEOs sought the input of other top managers outside their own firm and how well acquainted they were to them. Subsequently, they statistically correlated that to the extent to which these top managers received performance-contingent compensation packages, and found a very clear result.

Those CEOs who had a very small performance-related pay component in their compensation package sought very little true external advice. They relied on asking their friends – and perhaps their wife, uncles, and mother – whether they too thought that what they were doing was great, splendid, and spot-on. I guess it helps people feel more confident and self-assured…

In contrast, CEOs with a relatively large performance-contingent component in their remuneration package much more often sought advice from other executives who were not their friends and who had different backgrounds than themselves. These people may be slightly scary (they may actually tell you that what you’re saying is nonsense!) but perhaps also more useful. Moreover, McDonald and colleagues showed that this true advice-seeking significantly helped the financial performance of the CEOs’ companies, in the form of an increase in the company’s market-to-book and return on assets. Thus, the scary stuff actually led to hard cash!

The pay-for-performance construction paid off; it stimulated executives to repress their “it’s-only-human” inclination to avoid asking people’s opinion who might actually disagree with you. It is much safer and more pleasant to make sure to solicit advice from people who will say that you’re splendid, but it is much more useful – and lucrative – to really put yourself to the test. And if you reward them for it, and only if you reward them for it, CEOs – just like humans – will actually be brave enough to take this test.

No comments: