Claudine Gay's Harvard presidency ended after a single fall term. Liz Magill left Penn 17 months into the job. Ben Sasse walked away from the University of Florida after 18 months, and Minouche Shafik resigned from Columbia 13 months after taking office.
A trustee asked me this month why so many smart presidents keep leaving before a campus learns their names. It isn't a mystery. In each case the governing board was as central to the failure as the person who resigned.
Presidential turnover has become a board governance story. Search committees keep hiring talented leaders. Boards then undercut the same leaders with vague authority, direct donor pressure, and public reversals that arrive by text message. The result is a pipeline of short-term presidents, search fatigue, and institutions that cannot hold a strategy long enough to execute it.
The tenure numbers have been falling for years
This is not a run of bad luck. The American Council on Education's American College President Study, published in 2023, put the average tenure of a sitting president at 5.9 years. In 2006 the same measure was 8.5 years. Any search firm will tell you that five years is barely enough time to change a budget, let alone rebuild a campus culture.
The average hides something more volatile. Elite private research universities are now cycling through leaders in 18 months, 13 months, or some terms that last less than 200 days. Public flagships are not far behind. When a president leaves early, the campus pays twice: once in severance and search costs, and again in the decisions that wait for a permanent leader.
Boards are now the centre of the story
At Harvard, the Corporation praised Claudine Gay on December 12, 2023, after a disastrous congressional hearing, then accepted her resignation on January 2, 2024. A three-week reversal like that is not governance. It is a board deciding in public, and it tells every stakeholder that no decision is final.
At Texas A&M, President M. Katherine Banks resigned in 2023 after the botched hiring of journalism professor Kathleen McElroy. The board's behind-the-scenes interference in a single offer cost the university a president and a respected faculty member. At West Virginia University, the faculty assembly passed a no-confidence vote in President E. Gordon Gee after deep programme cuts; the board backed him, so the campus spent the next year with a governing authority and a faculty arguing about legitimacy rather than strategy.
The pattern repeats. A board hires to solve one problem, then yanks the decision after a donor objects, a legislator tweets about a faculty member by name, or a protest makes the front page. The president is not leading. The president is performing.
Four governance failures shorten presidencies
I have watched the same four failures across public systems and private boards. They are not exotic. They are mundane enough to fix.
- Search-stage dishonesty. Trustees say the next president will have autonomy, then withhold the real constraints: a donor must approve the provost, the athletics director reports to a board committee, and the law school dean has a direct line to the chair.
- Authority on paper only. Presidents sign the employment contract, but the board refuses to delegate hiring, budget and capital approvals, or communications decisions. Every major call goes back to a committee that meets quarterly.
- No due process in a faculty no-confidence vote. A faculty union or senate can pass a symbolic motion, and the board treats it as a personnel decision without saying so. The president stays, but every administrator knows the board is shopping.
- Donor and political vetoes. A president announces a plan, and an external funder or governor-appointed regent signals opposition. The board rescinds the plan, and the president takes the blame for a decision she did not make.
None of these requires a malevolent board. They arise from boards that have never agreed on what the president may decide alone, and that recruit without saying so.
Search-stage dishonesty sets presidents up to fail
The most dangerous phrase in a presidential search is 'we are open to change.' It can mean the board is open to new ideas. It can also mean the board wants a fundraiser who will not touch the medical school. The gap becomes apparent only after the contract is signed.
Search committees routinely hide real constraints from finalists. Candidates hear about ambition and innovation. They hear less about a board chair who has run the endowment committee for 20 years, a system office that controls capital budgets, or a governor who considers the university a political agency. The president accepts the role, raises tuition, cuts athletics, or challenges a dean who is protected by a trustee, and discovers where the real power sits.
When that mismatch becomes public, the narrative is always 'leadership failure.' It is more precise to call it selection failure. A board that cannot describe its own constraints cannot hire a president who will survive them.
When a board flips, the president rarely survives
Governance is stable only if the people who hired the president are the people who judge the president. In many systems that is no longer true. State boards change with gubernatorial elections. Private boards add trustees who arrive with a specific policy agenda. The president hired by one board majority reports to a different one eighteen months later.
This is not abstract. The link between state DEI bans and university hiring has moved political priorities directly into personnel and curriculum decisions. A president who launches an initiative under one political climate can be summoned to a board meeting and told to dismantle it under another. The same dynamic now surrounds campus divestment debates, where boards are expected to take sides in geopolitical arguments and then blame the president for the fallout.
Short-term boards produce short-term presidencies. The solution is not to avoid politics. It is to write down, before the search, which board commitments survive a change in membership. If trustees cannot commit, the next president should know that.
What steadier boards do differently
The institutions that keep presidents past the five-year mark tend to have one habit in common. They delegate. They set a narrow set of reserved powers, communicate them to candidates, and then let the president make decisions between board meetings.
This is a governance design question, not a personality question. In many German universities, the rector is elected by the senate and serves a fixed term; the faculty share in the appointment, so the mandate is broader from day one. In Australian universities, the governing council delegates operational authority to the vice-chancellor more explicitly than many American boards do. The United States board model can work well, but only when trustees accept that the board governs, the president manages, and the two functions are not interchangeable.
The Association of Governing Boards of Universities and Colleges has spent decades urging boards to adopt exactly this division. The best boards also build a presidential evaluation calendar that does not coincide with a crisis, so the president is not first told about dissatisfaction through a newspaper story.
For candidates, read the board before the offer
If you are considering a presidency, the board is not an afterthought. Ask to see the last three board meeting minutes, the written delegation of authority, and the precise composition of the executive committee. A board that will not share its minutes with a finalist is telling you how it will treat a president.
Ask who has informal veto power. Ask what a failure looks like at six months, at two years, and after the first full budget cycle ends. Do not accept a vague answer. The difference between a five-year presidency and an 18-month one is often settled before the offer letter is drafted.
The American Council on Education's American College President Study still offers the cleanest national picture of who serves, for how long, and under what pressures. Candidates who skip the data end up relying on the version of the job the search committee wants them to believe.
For boards, the practical step this month is one afternoon exercise: list every decision currently requiring board approval that a competent president should be able to make alone. If the list is longer than a page, the next search will produce another short tenure. Fix the governance, and the president stays.
