Finance chiefs have any number of reasons to be cautious when joining a company whose CEO is also a founder, not the least of which is how much control they’ll be afforded over the finances. Now a new academic study offers one more item to put on the list: a higher likelihood of being fired when there’s a restatement.

Andrew J. Leone, accounting professor at the University of Miami School of Business Administration, and Michelle Liu, assistant accounting professor at Pennsylvania State University’s Smeal College of Business, studied a sample of 96 public companies that announced accounting restatements between 1997 and 2006. They divided the sample into two groups — companies with CEOs who were founders and companies with CEOs who were not — and then tracked the postrestatement fates of both the CEOs and CFOs.

Not surprisingly, founder CEOs were fired a lot less frequently (29% of the time) than nonfounders (49%). But that fact apparently only made their CFOs more vulnerable. “CFO turnover around an irregularity is generally high anyway, around the 65% range,” Leone tells CFO, but when the CEO is a founder, the CFO is fired more than 80% of the time after a restatement.

To be sure, both executives may be asked to leave after a restatement. But in cases where a founder CEO survived, CFOs departed about 60% of the time, compared with about a 30% departure rate for CFOs working with nonfounder CEOs who remained.

The sample included many well-known companies that were accused of corporate fraud or backdating, including Lucent Technologies, Delphi, Brocade Communication Systems, and Refco. The study appeared in the January/February issue of The Accounting Review, published by the American Accounting Assn.

Leone and Liu tested an obvious explanation for the results: perhaps founder CEOs had less-than-average accounting acumen, thus putting more of the responsibility, and the blame, on the CFO. Using CEOs’ MBAs or prior experience in accounting as a proxy for their general accounting knowledge, though, the professors say that line of logic didn’t pan out. In a multivariable regression, that factor accounted for none of the differences between the two groups, they say.

Leone and Liu hypothesize that founder CEOs may be viewed as being more important to the firm than nonfounders, and that ostensibly independent board investigations of wrongdoing may be subtly biased to make a less-valued executive the scapegoat for the problems.

One particularly poignant case was that of Bradley W. Harris, former CFO of GMH Community Trust. In 2006 Harris sent a memo to his board of directors expressing his concern about the pressure his boss was applying to take aggressive accounting positions. “Although GMH’s audit committee concluded that no wrongdoing had occurred, the firm decided to restate its earnings and fire its CFO on the same day,” write Leone and Liu, noting that CEOs may “successfully use their influence to (1) avoid getting fired and (2) convince the board to fire the CFO.”

Leone and Liu offer one piece of good news, though. Consistent with the idea that CFOs were being fired as scapegoats and not because of true culpability, they found that fired CFOs of founder-managed firms were less likely to be the subject of a Securities and Exchange Commission enforcement action and more likely to find subsequent employment in a private or public firm, compared with fired CFOs of nonfounder firms.