S9 Ep54: Does sacking the manager work?

Your football team keeps losing. The season is slipping away. One thing can be changed by Friday, and the crowd is already singing about it: sack the manager.

Jan van Ours (Erasmus School of Economics, CEPR, Feyenoord fan) has looked at seven seasons of the Dutch top flight and 31 managers who were sacrificed mid-season. To work out whether any of them deserved it, he uses bookmaker odds and expected goals to distinguish bad play from bad luck.

The new-manager bounce is real: results improve after a new manager walks in. But also, not real: clubs that don't sack the manager have an upturn too. The message for the boardroom, in football and business, is that not doing anything might often be the best course of action.

The research behind this episode:

van Ours, Jan C. 2026. "Dust in the Wind: Causes and Consequences of Managerial Replacements." CEPR Discussion Paper DP21850, Centre for Economic Policy Research. The paper is gated.

To cite this episode:

Phillips, Tim, and Jan van Ours. 2026. "Does sacking the manager work?" VoxTalks Economics (podcast).

About the guest

Jan van Ours is Emeritus Professor of Applied Economics at Erasmus School of Economics, Erasmus University Rotterdam, and a core member of the Erasmus Centre for Applied Sports Economics. He is also Adjunct Professor at the Center for Health Economics, Monash Business School, and a Research Fellow of the Centre for Economic Policy Research. His research spans unemployment dynamics, labour market policy, health and well-being, and the economics of professional sport, where match data offer a rare chance to watch a labour market in the open.

Research cited in this episode

The Eredivisie sample. Seven seasons of the top league of Dutch professional football, from 2018/19 to 2024/25, covering 4,136 match observations and 126 club-seasons. In that period 31 managers were replaced during a season; about 15% had gone by mid-season and 25% by the end of it.

Points surprise. The cumulative gap, from the first match of the season, between the points a club has won and the points the bookmakers implied it should have won. Bookmaker odds are used as the benchmark because they already contain everything the market knows about squad quality, home advantage and the opposition; van Ours confirms this in the data, finding that recent results add nothing to the odds as a predictor of the next match.

Performance surprise. The same cumulative gap, but measured with expected points derived from expected goals rather than actual results. Points surprise catches a club that is losing. Performance surprise catches a club that is playing badly. A club can be one without being the other, which is how luck gets separated from ability.

Expected goals. The probability that a given shot becomes a goal, estimated from thousands of comparable attempts and conditioned on the distance and angle of the shooter, the body part used, and the type of pass and attack. Van Ours converts expected goals scored and conceded into a distribution of match outcomes, and from that into expected points. His data come from fbref.com; the paper notes that match-level expected goals were discontinued in January 2026 after a dispute between fbref and Opta.

The counterfactual replacement. The device that carries the paper. For each actual sacking, van Ours searches the same club in a different season for a moment when the sum of points surprise and performance surprise was almost identical, and the manager survived. Of the 31 replacements, 22 have a counterfactual, 19 of them unique. Clubs that sacked the manager gained 0.21 points per match afterwards. Clubs in the same trouble that did not sack the manager gained 0.38.

Scapegoating. William Gamson and Norman Scotch set out the idea in "Scapegoating in baseball" in the American Journal of Sociology in 1964, describing the sacking of a manager as an anxiety-reducing ritual that participants treat as an improvement whether or not anything improves. Van Ours returns to it to explain why a decision with no measurable effect keeps being taken.

Managers in regular firms. Stuart Gilson's 1989 study in the Journal of Financial Economics found that replaced executives were not employed by another firm for at least three years. Football managers are frequently back in work within weeks, which is one reason the job pays what it does. Hilger, Mankel and Richter reviewed 91 studies of top executive dismissals published between 1960 and 2010 and concluded that the effects of managerial turnover are not statistically different from zero.

Related reading on VoxEU.org

What we can learn about economics from professional sport during COVID-19, a VoxEU column on why sport keeps producing clean natural experiments for economists.

Racial bias in newspaper ratings of professional football players, a VoxEU column using match ratings to test for discrimination in a labour market where output is measured in public every week.