This paper tests the predictions that (i) sub-central government expenditures are procyclical and (ii) sub-central government expenditures are likely to be more procyclical than central government spending. The predictions are based on the importance of ‘voracity effects’ and on the proposition that they are systematically more pervasive if spending is financed by intergovernmental transfers. Evidence from 23 OECD countries between 1995 and 2006 indicates that sub-central government spending is more procyclical than central government expenditure.