Asymmetric output growth effects of government spending: Cross-sectional and panel data evidence

Abstract

Using a model that accommodates asymmetric adjustments of output growth to changes in growth of government spending, the effects of aggregate and disaggregate government spending variables on output growth are examined. Both cross-section and panel regression estimations are conducted. Robustness of results to joint-endogeneity of output and government spending, alternative conditioning variables, and cointegration is investigated. Aggregate government spending appears to have positive output growth effects particularly in periods of below-trend growth in this variable. The government sector's productivity appears to be higher than non-government sector's productivity when spending growth is below-trend growth and only for non-OECD countries; no differences in sectoral productivities are detected for OECD countries. Government consumption spending has no significant output growth effects, but government investment spending has positive output growth effects particularly when its growth falls below its trend-growth; this favorable effect turns negative when government investment spending growth exceeds its trend-growth. Results are robust across alternative model specifications and estimation methods.