In this paper we argue that little is known about either the geographical o
bjectives or the spatial outputs of the welfare state. Conclusions of geogr
aphical inequality are problematic for three main reasons. First, the geogr
aphical aims of the welfare state, "the spatial strategy of equality", are
unclear. Second, the geographical distributional paradigm is rarely placed
in the wider context of focal and national welfare states, and the tension
between spatial equity and local autonomy is ignored.;Third, the geography
of welfare, "the spatial division of welfare" is often based on simplistic
and confused evidence. Much of the existing work implicitly takes a central
ist perspective, assuming that all geographical inequalities are defects. I
ssues of local government, local politics and local welfare states are igno
red. All detected inequality may not be "bad", and greater spatial equity m
ay not necessarily be "good". The spatial division of welfare should not be
examined in an analytical vacuum, isolated from the wider contextual issue
s of national and local services and the trade-off between focal autonomy a
nd territorial justice. If the "default value" is that all detected geograp
hical variations are assumed to be defects, then the arguments for localism
are doomed to failure.