This paper examines the effect of IMF programs on poverty, income inequality and government expenditure on education and health. It distinguishes between non-concessional and concessional programs and examines the effect of contingent factors; program completion, IMF resources, the size of the initial economic problems and aid dependency. Our findings suggest that, in general, IMF programs have not significantly increased poverty and income inequality; in some instances they are associated with reduced poverty and income inequality. Moreover, we find little evidence that IMF programs lead to severe cuts in social expenditure.