Abstract
Small and Rosen’s (Econometrica 49(1):105–130, 1981) method for measuring consumer surplus using discrete choice models has been widely adopted in public policy analysis. For the case of a price change, the present paper elucidates five theoretical assumptions inherent within Small and Rosen’s measure, and employs indifference maps to demonstrate that this measure is only applicable to the context of a single discrete choice free of non-linear income effects. The paper argues that, where non-linear income effects are present, the aforementioned theoretical assumptions should be relaxed, and the consumption context revised from discrete choice to discrete–continuous demand. Furthermore, the paper proposes a simple analytical method for approximating the expected Hicksian compensating variation in the presence of non-linear income effects, and compares the empirical performance of this method against existing methods using data from Morey et al. (Am J Agric Econ 75(3):578–592, 1993). As well as offering a simple approximation, the proposed method yields insights on the potential range of the compensating variation depending on the extent of switching between choice alternatives, and on the attribution of the compensating variation to the relevant choice alternatives.
Highlights
Discrete choice models are routinely commissioned by planning authorities, as a tool for predicting the impacts of policy interventions on demand and welfare
We began the empirical analysis by estimating a simplified multinomial logit (MNL) model with utility specified to be linear in residual income, but omitting alternative specific constants, quality and other socio-economic variables as previously discussed
Adopting the presentation by S&R, we introduced a problem of discrete–continuous demand whereby an individual is offered a discrete choice between two mutually exclusive goods; conditional upon that choice, he/she consumes a positive quantity of the chosen good and potentially a positive quantity of a numeraire good, subject to the constraint of budget
Summary
Discrete choice models are routinely commissioned by planning authorities, as a tool for predicting the impacts of policy interventions on demand and welfare. Such interventions may take many forms, but fundamental to economic interests are those which affect prices and/or incomes. Discrete choice models are conventionally estimated on individual-level preference data, whereas planning authorities are focussed more on the preferences of the market or society as a whole. This provokes an interesting question of how discrete choices. Dekker can be defensibly aggregated, so as to yield market- or societal-level forecasts of quantity demanded and welfare following price and/or income changes
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call