Dr. Paul Koster – Choice modelling for participatory value evaluation experiments

  • Date: Tuesday 3 July 2018
  • Location: Blenheim Terrace SR (G.15) House No. 11-14
  • Time: 11:00-12:00

Paul Koster, Assistant Professor, Vrije Universiteit.

Choice modelling for participatory value evaluation experiments.
(Dekker, Koster and Mouter)



Participatory value evaluation (PVE) is a novel valuation approach to study the optimal allocation of public budgets developed in a series of papers by Mouter et al. (2018), Dekker et al. (2018) and Koster et al. (2018). Citizens are asked to choose the best portfolio of governmental projects with corresponding impacts for society given a budget constraint of the government. PVE differs from CBA in at least three aspects: i) individuals are conceptualized as co-owners of the government (not as consumers of public goods) who derive utility from the extent to which the government allocates taxes in accordance with their preferences, ii) it relaxes the assumption of CBA that value derived from private choices reflects the value citizens derive from public projects, iii) it gives citizens the opportunity to incorporate distributional concerns.

This presentation presents results of two realistic PVE experiments with an emphasis on the challenges for choice modelling. In the first PVE experiment the budget of the government is fixed and citizens choose the optimal portfolio of projects or delegate their decision to an expert. The remaining budget will be shifted to the next year. In the second experiment citizens can also adjust the governmental budget by decreasing or increasing the additional tax per inhabitant. We model portfolio utility as the sum of random project utilities given the governmental and private budget constraints using standard choice models and MDCEV modelling approaches in the spirit of Bhat (2005, 2008, 2018).

The presentation concludes with a discussion on how the obtained results can be used for welfare analysis. This allows us to determine the optimal portfolio of governmental projects given a fixed budget or jointly optimise the budget and the optimal portfolio.



Bio: https://research.vu.nl/en/persons/pr-koster

Paul Koster works as an assistant professor at the Department of Spatial Economics, Vrije Universiteit Amsterdam. He graduated at this department in Economics in 2007 and obtained his PhD in 2012 on a thesis entitled “The cost of travel time variability for air and car travellers”. He is also affiliated as a research fellow at the Tinbergen Institute. His research focuses on the valuation of non-market goods such as travel time and travel time variability and the regulation of externalities resulting from transportation. He works on explaining and understanding individual choice behaviour using advanced discrete choice econometrics and connects behavioural models to economic evaluation of transport policies.


Bhat, C. R. (2005). A multiple discrete–continuous extreme value model: formulation and application to discretionary time-use decisions. Transportation Research Part B: Methodological, 39(8), 679-707.

Bhat, C. R. (2008). The multiple discrete-continuous extreme value (MDCEV) model: role of utility function parameters, identification considerations, and model extensions. Transportation Research Part B: Methodological, 42(3), 274-303.

Bhat, C. R. (2018). A new flexible multiple discrete–continuous extreme value (MDCEV) choice model. Transportation Research Part B: Methodological, 110, 261-279.

Dekker, T., Koster, P.R., Mouter, N. (2018) Choice modelling for participatory value evaluation experiments. Working paper.

Koster, P.R., Mouter, N., Dekker, T. (2018) Counting what counts: Participatory value evaluation and altruistic concerns. Working paper.

Mouter, N., Koster, P.R., Dekker, T. (2018) Participatory value evaluation: an introduction. Working paper.