Since the late 1970s, the received wisdom has been that government size (measured as the ratio of total government expenditure to gross domestic product (GDP) or government consumption to GDP) is detrimental to economic growth. We conduct a hierarchical meta‐regression analysis of 799 effect‐size estimates reported in 87 primary studies to verify if this assertion is supported by existing evidence. Our findings indicate that the conventional prior belief is supported by evidence mainly from developed countries but not from less developed countries. We argue that the negative relationship between government size and economic growth in developed countries may reflect endogeneity bias.