The demand for certain types of health care services depends on decisions o
f both the individual and the health care provider. This paper studies the
conditions under which it is possible to separately identify the parameters
driving the two decision processes using only count data on the total dema
nd. It is found that the frequently used hurdle models may not be adequate
to describe this type of demand, especially when the assumption of a single
illness spell per observation period is violated. A test for the single sp
ell hypothesis is developed and alternative modelling strategies are sugges
ted, including one that allows for correlated unobserved heterogeneity. The
results of the paper are illustrated with an application. (C) 2001 Elsevie
r Science S.A. All rights reserved.