Bard College president Leon Botstein resigned, effective June 30, after an independent review commissioned by the college’s Board of Trustees found that he had a closer relationship with convicted sex offender Jeffrey Epstein than was previously known. The lesson for higher education, however, runs beyond Bard. Long-tenured presidents can accumulate authority that operates beyond meaningful oversight. Botstein’s own defense illustrates the problem.
“I would take money from Satan if it permitted me to do God’s work,” Botstein told WilmerHale, the law firm that conducted the review, in his defense. The ends-justify-the-means argument obscures who pays the cost. The young women Epstein continued to abuse during the years of Botstein’s affiliation did not get to make that calculation.
That a sitting college president offered this statement on the record, to investigators, as exoneration, is itself evidence about how Bard has been led over the five decades of Botstein’s tenure. The review found that Botstein made roughly 25 visits to Epstein’s town house, took a two-day visit to Epstein’s island and hosted Epstein on the Bard campus—conduct that, the firm concluded, “could have alerted President Botstein to the possibility that he and Bard would be facilitating Epstein’s continued abuse of women, legitimizing Epstein, or exposing Bard students to a person like Epstein.” At least one senior faculty member warned him directly against affiliating with Epstein, according to the report. In recent weeks, Botstein assured the Bard community that he expected to be cleared.
I did not want to write this story. I am a Bard alumna. I worked for nine years in the college’s admission office, left to earn a math doctorate and returned to teach in Bard’s Language and Thinking Program. The institution shaped me, and I remain shaped by it. I have spent months watching with growing unease as the Bard community proved unable to challenge its longtime president. Many faculty and staff quoted in stories in The Guardian, Inside Higher Ed and The Daily Catch, which provided substantial local coverage, did so anonymously, “out of fear of retaliation.” The New York Times reached out to board members for comment, and they did not respond, and the student government representative initially agreed to an interview but then “changed her mind without explanation.”
Trustees with access to the same investigative reporting the rest of us read, and to the public Epstein files those reports drew on, did not feel empowered to act without cover from an outside law firm. A president who served more than half a century and recently raised a billion dollars appeared, functionally, unaccountable.
This problem extends beyond Bard. Higher ed might look to the aviation industry for an analogue example. On March 27, 1977, KLM Flight 4805 took off from Tenerife in the foggy Canary Islands without clearance from air traffic control. Its captain, Jacob Veldhuysen van Zanten, was the literal face of the airline, a chief training pilot who had logged 11,700 flight hours. His first officer raised a concern, but Van Zanten dismissed it, and the first officer did not press. Seconds later, a Pan Am jet—one with takeoff clearance—emerged from the fog on the same runway. The crash, which killed 583 people, was the deadliest accident in aviation history.
Investigators eventually concluded that the cockpit’s governance structure was the problem. Van Zanten was, as one analyst later put it, the man “everybody at that airline looked up to.” In cockpit culture at the time, his subordinates did not have authority to override him. Tenerife was not an isolated failure of one captain’s judgment; similar dynamics had brought down Eastern Airlines Flight 401 and United Flight 173, also in the 1970s. In each case, a senior captain’s authority effectively silenced crew members who knew something was wrong.
The industry’s response, developed at a 1979 NASA workshop, was crew resource management. CRM recognized that a pilot operating in an authority-saturated cockpit could eventually fail catastrophically, regardless of skill or experience. The fix would not be better captains. The fix had to be a system in which junior crew were trained to escalate concerns from suggestion to assertion to direct challenge, and they needed assurance that they would be protected when they did so. Captains were trained to treat questions as data rather than dissent, and they were instructed on soliciting input. Hierarchy was preserved, but it was no longer the only safety mechanism. CRM is both a philosophy and a training program, and the safety record since indicates it works.
Higher education’s governance cockpit looks a lot like aviation’s did before 1979. Long-tenured presidents who excel at fundraising accumulate authority that boards become reluctant to challenge. The longer the tenure and the larger the endowment, the steeper the cost of confrontation and the more diffuse the responsibility for oversight. Faculty governance, where it exists, depends on individual courage rather than systemic protection.
Those at Bard who were quoted anonymously were, after all, behaving rationally. Trustees, often donors themselves, are typically selected for alignment with the president rather than independence from him. When the president has the best information and good judgment, the system works. When he does not, there is no second safety mechanism.
Bard’s board will now select a successor. Eventually, so will every other institution led by a president of long tenure. Search committees may be tempted to treat this as a casting problem: Find someone of better character, sounder judgment, surer instincts. Aviation rejected that approach for good reason. Van Zanten was, by every available measure, an excellent pilot. The problem was not him. The problem was a cockpit in which no one could stop him.
The harder question, and the one I cannot answer from outside any particular institution, is what the equivalent of crew resource management would look like in higher education. The answer will differ at a small liberal arts college and a large research university, at an institution with a 50-year president and one with a five-year president, at a board composed largely of donors and one with structural independence. But the question is the same everywhere: When something goes wrong, how can a college change who can speak up, when and to whom?
That is work for trustees, faculty leaders and presidents themselves to do now. Aviation did not wait for a second Tenerife. Higher education should not wait for a second tragedy like the one at Bard.
Susan D’Agostino is a mathematician whose stories have been published in The Atlantic, the BBC, Scientific American, The Washington Post, the L.A. Times, Wired, Quanta and other leading media. Her next book, How Math Will Save Your Life, will be published by W. W. Norton.
