PhD defence: From Data to Decisions in Labour Economics

On Friday 10 April 2026 at 12 pm, Bas Scheer will defend his PhD thesis From Data to Decisions: Policy-Relevant Empirical Studies in Labor Economics.

In his dissertation, Bas Scheer investigates how major long-term changes affect the labour market and what these changes mean for labour market policy. To this end, he focuses on three developments: the population ageing, the rise of non-standard forms of work, and technological and data-driven changes. 

Demographic developments

First, the dissertation examines demographic change. As populations age and birth rates fall, many countries are concerned about future labor shortages and slower economic growth. Using detailed Dutch data, Scheer shows that the growth of the labour force in the past was mainly driven by population growth and increased participation rates of women. Looking ahead, Scheer expects labour supply growth to slow as population growth slows and women's participation rates level off to the same level as men's. This implies that future economic growth cannot rely on an ever-increasing labour supply.

Payrolling

Secondly, Scheer analyses changes in the way work is organised. He has looked in particular at "payrolling", a form of outsourcing in which employees are hired through intermediary agencies. His research shows that employees in these situations tend to have less job security, fewer permanent contracts and lower pension contributions. These results highlight how flexible working arrangements can increase the disparity between secure and precarious jobs. 

Data and decision-making

Thirdly, in his dissertation, Scheer investigates how advances in the availability of data and computing power affect decision-making. He shows how predictive models can be used to select students for studies that are subject to a limited number of places: selection for admission to the study of medicine, for example. By reserving places for specific groups, you prevent individuals in these groups from having a lower chance of admission due to the chosen selection method. The research shows that in this way, more diverse student populations can be reached with the help of data-driven selection with only limited costs for efficiency, which can be measured, for example, by the success rate. 

Prediction models and turbulence

In the final sub-study of his thesis, Scheer shows that the prediction errors of economic prediction models move along with the state of the economy. He specifically looks at prediction models for the unemployment rate up to a maximum of two years ahead. Forecasting errors tend to increase when the economy develops 'turbulently': when the development of unemployment suddenly changes. For example, from steadily decreasing to rapidly rising. These are difficult periods to predict, which increases forecasting errors. In quieter periods, for example when unemployment is steadily falling at the same rate, forecast errors are lower. 

This pattern can be seen in many different types of prediction models, but the extent to which this pattern occurs differs between models. Some forecasting models do especially well in quiet times, but struggle when the economy turns. This is useful to know for institutes that make estimates of the economy, because it gives you something to choose from: a forecasting model that does especially well in quiet 'normal' economic times, or a forecasting model that is more robust and does relatively well in turbulent economic times. 

Conclusions

Labour market policy benefits from thorough empirical research, especially in a changing world. Policymakers can draw on the results of the studies in this dissertation when drafting legislative proposals or the national budget.

For example, the labour supply is expected to grow less in the future than we are used to. It is wise to adjust the national budget and policy accordingly now, so that state finances remain sustainable in the long term.

In addition, reserving study places for underrepresented groups can contribute to the diversity of the student population, resulting in only a limited expected decrease in the graduation rate. Examples of these types of groups are students with a migration background in medicine or women in technical programmes. The law can be amended to make this type of reserved seats more possible.

Bas Scheer is scientific researcher at the Netherlands Bureau for Economic Policy Analysis (CPB) and a PhD student at the Utrecht University School of Economics (U.S.E.).

Streaming video

You can follow this PhD defence via streaming video: