One of the most common justifications for hefty C.E.O. compensation packages is that if the leaders of industry are not paid well, the so-called best and brightest will no longer flock to fill the corporate ranks, and will instead go elsewhere. High salaries (and bonuses, etc) are said to both motivate and retain these brilliant minds.

While this sounds somewhat plausible, as it turns out, a new study shows that it’s just not true. One driver of executive pay, called the peer-group benchmark, compares the salaries of executives among ostensibly similar companies as a way of keeping salaries competitive and within reasonable market limits. The problem is, this measure assumes that a C.E.O. at one company could pick up and leave for greener pastures at another, which, as it turns out, is a false presumption.

The study, conducted by Charles M. Elson and Craig K. Ferrere, shows that many of the skills C.E.O.s possess are specific to the company in which they are acquired, and are not readily transferable to other companies. Their analysis shows that almost every attempted transplant at the top ranks has resulted in failure.

What this means is that all this benchmarking makes the market of C.E.O.s seem like a market with high mobility, allowing for C.E.O.s to move to other companies when in fact a C.E.O. who manages one company well is unlikely to be successful in another.  Therefore, a company looking for a C.E.O. cannot actually consider all C.E.O.s as potential candidates. Benchmarking, then, is little more than a way to inflate executive salaries by comparing jobs in markets that are essentially incomparable.

Ultimately this study shows that determining executive salaries needs to be reevaluated and reconfigured with an eye to empirical data, even if that means reducing C.E.O. pay. After all, we are all shareholders in these companies and they are giving away our money for what turns out to be no good reason.