Media Briefings

DISCOUNTED TUITION FEES CAN REDUCE CLASS ATTENDANCE: Evidence of the ‘sunk-cost effect’ among Dutch university students

  • Published Date: July 2016

DISCOUNTED TUITION FEES CAN REDUCE CLASS ATTENDANCE: Evidence of the ‘sunk-cost effect’ among Dutch university students

University students who received a surprise discount on their fees for voluntary extra-curricular tutorials responded by being less likely to show up for class. That is the central finding of experimental research by Nadine Ketel, Jona Linde, Hessel Oosterbeek and Bas van der Klaauw, which is forthcoming in the Economic Journal. Their results are an indication that reducing the cost of education can lead to a fall in some students’ commitment to studying.

Can paying for someone’s education hurt them? That is the question explored in this study. Education at all levels is heavily subsidised around the world. Clearly this can help students as it makes education affordable for many more people. But if education is cheap, students may value it less and therefore devote less time and effort to studying. This is known as the ‘sunk-cost effect’: if you put up more of your own money, you want to get your money’s worth.

This study uses a field experiment to investigate the sunk-cost effect. In particular, a random group of students were surprised by a discount on their tuition fee for an extra-curricular course. As a result, the different students are all equally willing to pay the full price for the course, but some end up paying less than others.

During the experiment, the researchers recorded detailed information on class attendance and exam performance. The sunk-cost hypothesis predicts that students who pay less for the course will attend fewer tutorial sessions, with possibly detrimental effects on their performance.

Within the full sample of students, the study finds no substantial effect of the discount on attending lectures and exam performance. But using answers to hypothetical questions, the researchers identify a group of students who are likely to display the sunk-cost effect.

For this group, they do find that the discounts on tuition fees reduce class attendance. When these students do not pay for the course, their attendance rate is 11 percentage points lower than when they pay the full price. These students constitute almost half of the sample. For performance, there seems to be no effect of the discount.

The study focuses on university students who voluntarily signed up for extra-curricular tutorial sessions. While the results do not prove that reduced tuition fees reduce attendance for (a large subset of) the wider population of university students in regular classes, it suggests that this is a possibility.

The researchers conclude:

‘There are serious concerns about the low effort that university students in the Netherlands (and elsewhere) devote to their studies.’

‘To increase study effort, it may be worth considering a reduction in the education subsidies that keep tuition fees low.’

‘This can possibly be done in combination with increased financial aid to students to keep university education accessible.’

Although this study explores the sunk-cost effect in an education setting, the same effect can also affect behaviour in other settings. The researchers only find evidence for the sunk-cost effect for a sub-group who can be identified as prone to sunk costs. This suggests that it is important to examine who is and who isn’t prone to sunk costs before worrying about, or looking for, sunk-cost effects in other settings.


Notes for editors: ‘Tuition Fees and Sunk‐cost Effects by Nadine Ketel, Jona Linde, Hessel Oosterbeek and Bas van der Klaauw is forthcoming in the Economic Journal.

Nadine Ketel, Jona Linde and Bas van der Klaauw are at VU University Amsterdam. Hessel Oosterbeek is at the University of Amsterdam.

For further information: contact Romesh Vaitilingam on +44-7768-661095 (email:; Twitter: @econromesh); Nadine Ketel via email:; Jona Linde via email:; Hessel Oosterbeek via email:; or Bas van der Klaauw via email: