Marx has never been very influential among Western economists, but the coll
apse of all types of socialist movements has reinforced the belief that Mar
x is irrelevant for economic analysis. At the same time, some heterodox eco
nomists have claimed that neoclassical theory is sufficiently flexible to p
rovide a foundation for post-Keynesians, institutionalists, evolutionary an
d feminist economists. We argue that both conclusions are incorrect and tha
t Marx's treatment of agents' choices and constraints, and of systemic coop
eration and conflict, is far superior to that of orthodoxy in several cruci
al respects and can provide a better grounding for non-neoclassical analyse
s.