Much of the work which has led to a widely held view that the income e
lasticity of health care spending exceeds one has been based on intern
ational cross-section data, or on pooled cross-sections and time serie
s. In this paper we re-examine this view in the context of long-run eq
uilibrium relationships between non-stationary time series, possibly i
ncluding autonomous trends. Our results cast doubt upon the usefulness
of pooling and upon the notion of an elasticity above one.