On the ecological fallacy in discrete-choice models

In linear regressions, the ecological fallacy—the erroneous belief that aggregate-level coefficients coincide with individual-level coefficients—arises when individual outcomes depend on the group environment. This paper suggest that such “group effects” determine also the circumstances under which the ecological fallacy vanishes from basic discrete-choice models. In particular, when controlling for group effects, it is shown that the same coefficients arise from a conditional logit model, which is a popular framework to analyse individual choices, and a Poisson regression, which provides a framework to analyse the aggregate number (or count) of such choices across groups. •An ecological fallacy arises if individual differ from aggregate-level coefficients.•In linear regressions, the ecological fallacy is caused by group effects.•In a discrete-choice setting the ecological fallacy is also caused by group effects.•The fallacy can vanish from simple conditional logit and Poisson count models.•In a discrete-choice setting, it is easier to control for group effects.