Preferences as a Determinant of the Optimal Level of Decentralisation in Health Care Resource Allocation : Theoretical Insights and an Empirical Application

From an economic point of view, decentralisation is expected to increase
social welfare through better matching of service delivery to preferences. The
latter have been a central piece of the economic rationales for decentralisation but
only indirectly. Thus, at the theoretical level, the main question addressed in this
dissertation is: might preferences in themselves influence the impact of
decentralisation on allocative efficiency, in the context of health care resource
allocation?
Regardless of which model (public choice theory or principal-agent
theory) is used to explain the positive outcome above mentioned, in any case, the
benefits generated by decentralisation depend on the assumption of variation in
preferences across jurisdictions. However, there is little empirical evidence
regarding this matter. Consequently, at the empirical level, the main question
addressed in the current work is: does geographic variation in preferences, in the
context of health care resource allocation, exist? To answer this question we
developed and administered the same questionnaire (eliciting preferences) to two
independent samples drawn from two Portuguese municipalities.
Within our framework, central and local decision-makers are seen as
alternative agents acting on behalf of local populations. Given the different
capabilities possessed by agents, decentralisation of resource allocation generates
some trade-offs between objectives. Depending on the trade-offs that local
populations are willing to make, they will be better-off with one or the other
agent. Therefore, we conclude that the specific preferences held by individuals
might also determine in themselves whether or not decentralisation is optimal,
when compared to centralisation.
Concerning the empirical work, the principal conclusion is that the results
do not corroborate the hypothesis of geographic variation in preferences, meaning
that the theoretical discussion about the impact of decentralisation on allocative
efficiency might have to be revisited. The empirical results further suggest that the
geographical dimension of (in)equality in treatment matters to people and that a
maximum opportunity cost of equality, in terms of health gain foregone, is likely
to exist.