Skill transferability, regret and mobility

Abstract

After graduation many students start working in sectors not related to their field of study or participate in training targeted at work in other sectors. In this article, we look at mobility immediately after graduation from the perspective that educational choices have been made when these pupils had little experience of the actual working life in these professions. We develop a model where students accumulate partially transferable human capital but also learn about their professional preferences at the university and during the first years in the labour market. As a consequence of this newly acquired insight, these young workers might realize that working in another occupational field would better fit their preferences, although they are better equipped to work in their own field. The empirical analysis reveals that if wages are 1% lower due to lower skill transferability, the probability that a graduate who regrets his choice actually switches decreases by 1.4 percentage points, while those who switch on average take 0.3 months additional education.

Acknowledgements

The authors would like to thank Arnaud Dupuy, Ben Kriechel, Edwin Leuven, Philip Marey, Derek Neal, Hessel Oosterbeek, Gerard Pfann, Catherine Ris, Wendy Smits, Maarten Vendrik, Bas Ter Weel, seminar participants at the 2003 Human Capital Workshop at Maastricht University, the 2003 EALE conference in Seville and the 2003 Transitions in Youth conference at Madeira, the editor and an anonymous referee for valuable comments and Christina Lönnblad for editorial assistance.

Notes

¹ We use the term ‘pupil’ to indicate an individual who is in secondary school, ‘student’ to indicate a person who attends college and ‘graduate’ to indicate that one has successfully finished college.

² For simplicity, utility can only be derived from working and not from studying.

³ We assume that the pupil bases his choice on an estimation of I_k and is not yet aware of the true value of his preferences. While his perception of his true preferences might change, his true preferences are constant.

⁴ Individual subscripts are excluded from the model.

⁵ An additive version of the model gives similar results.

⁶ This is due to our assumption that graduates start working in the latest chosen profession.

⁷ This may occur if a profession demands skills from different educations. Changing education then becomes a prerequisite for working in such a profession, not a correction of the original choice. Next to this, switching can also be explained by a consumption effect when some studies provide much direct utility but do not offer interesting job opportunities. For simplicity, we assume in our model that utility can merely be derived from an education while working. Extensions including consumption motives from education do not affect the empirical results.

⁸ Therefore, the duration of further education is not only truncated at zero but also at 3 years.

⁹ The results do not change qualitatively when all graduates are included.

¹⁰ In the robustness analysis, we investigate the effect on the results when switching is defined differently.

¹¹ There is no information available about the intensity of training or education in hours per week.

¹² The concept of regret as defined in regret theory (introduced into economic theory by Loomes and Sugden, 1982; Bell, 1982; Fishburn, 1982) fits closely to our definition. While utility directly depends on the level of regret in regret theory, utility here only depends on actual consumption and job satisfaction. In regret theory, people therefore try to avoid a situation of regret, while in our model, people just try to maximize utility.

¹³ Note that this also implies that s ₁ is a constant.

¹⁴ The results are similar if the transferability parameters are not conditional on regret.