Since 1945, there has been substantial variation among the OECD countries and within them over time in central-government deficits. This paper surveys and assesses the literature on deficit spending in these industrialized democracies, emphasizing positive theories and the empirical support, or lack thereof, for each of them. We begin by considering the degree to which theories about the size of government and empirical tests of these theories about the level of overall government spending can explain deficits. We then survey theories pertaining to the political economy of fiscal deficits per se, including the ‘tax smoothing’ hypothesis, the neo-Keynesian countercyclical hypothesis, the public choice perspective, the ‘left-right ideology’ hypothesis, Roubini and Sachs' (1989) ‘strength of government’ hypothesis, and the importance of divided government in non-parliamentary systems.