Harvard sent out a painful letter earlier this month.

“I write to invite your advice on the search for a new president of Harvard” said James R. Houghton, chairman of a presidential search committee for the Harvard Corp. Houghton’s plea, which was addressed to Harvard alumni and other Harvard friends, noted that the university was seeking “a person of high intellectual distinction” who demonstrated a “capacity to guide a complex institution through a time of significant change.”

The painful part is that this is not the first time Harvard has mounted such a search. The university was also on the hunt in 2000, when it sought a president “with intellect and breadth of knowledge.” Harvard recently forced out the man selected in that search, Lawrence Summers. Summers was a star, a famous economist, and the former Treasury Secretary. His sin was trying to make the quality of Harvard’s product, education, live up to the quality of its brand.

Institutions, whether universities or companies, know when they need to change. They spend a lot of energy researching candidates while hiring. Indeed, the greater the anxiety, the higher the premium the institution is willing to pay for a star.

But once the star is in place, the search committee recedes, and the insiders rule. This old crowd is often the one whose limits generated the trouble in the first place. They ensure that the newcomer has a hard time succeeding. In the end the star often stumbles.

Soon enough, the board meets to begin a new search amid mea culpas. This is also the point where you hear the board drop jokes about the unbearable flaws of the same executive it once courted so assiduously.

Why the cycle? Some argue that the star figure is at fault – a bad match for the company. This month, about the same time that Houghton sent out his letter, the Harvard Business Review published an article on the corporate version of the Harvard problem. In the article, “Are Leaders Portable?,” Harvard got an answer: less than you think.

Authors Boris Groysberg, Andrew N. McLean and Nitin Nohria looked at what happened to 20 companies that were struggling, in one way or the other, and invited stars from General Electric Co. to save them.

The expectations for these transplants were high. Every company is considered lucky when it lands a GE executive, with the hiring companies seeing an average gain in share value of $1 billion at hiring. Home Depot Inc.’s market value rose by almost $10 billion when it hired GE’s Robert Nardelli.

Over the longer term the success of the GE turnaround artists was mixed. Tom Rogers, for example, famous for his work at GE creating CNBC, stumbled at publishing house Primedia Inc. Another GE veteran, John Trani, had what the authors call a “troubled tenure” at toolmaker Stanley Works.

The authors compared stock market returns over the years. Nine out of the 20 companies did better than competitors that hadn’t hired GE executives. The other eleven did 40 percent worse than peer companies. Many former GE stars resigned or got the axe.

Another Review article by Groysberg, Nohria and researcher Ashish Nanda, titled “The Risky Business of Hiring Stars,” looked at the performance of 1,052 stock analysts from the mid-1980s to the mid-1990s. The scholars found that “when a company hires a star, the star’s performance plunges.” So does the performance of his team.

“It seems hiring firms systematically overestimate the value of a star’s innate ability and underestimate the contribution of the platform he stood on,” Nanda says.

But of course the trouble often inheres in the institution, and not with the star. Colleges or companies often don’t want change itself. They just want to give the appearance of change.

Companies where the founder or his culture still dominates are especially averse to shifting and suspicious of stars who would change them. The employees of Stanley Works didn’t stand by Trani when he advocated a sensible move to cut its taxes by reincorporating the company outside the U.S. Trani retired shortly after.

Harvard has such a strong culture that it behaves like a family institution. Summers himself is “of Harvard,” – he taught there – but maybe insufficiently so. His years at the U.S. Treasury, the World Bank, and Brookings Institution may have made him feel like a stranger.

Conclusions? For stars, moving to a new company is riskier than it might seem. Life is so short that such a move may not be worthwhile. Sometimes all that spectacular risk offers in the end is the chance to go down as Summers did.

The second thought is for the stakeholders. The sort of pre-hire consultation that the recent Harvard letter promises is not enough. The board, shareholders, and other interested parties need to be involved after the hire as well. Otherwise, all too soon, will come another humble search letter on quality letterhead, asking for “general advice” on a search for a new CEO.

Miss Shlaes is a columnist for Bloomberg News.