This study attempts to investigate the effect of discretionary government spending volatility on economic growth. This impact has been examined at the aggregated level for a world representative sample as well as for the samples disaggregated by development level of the countries. We have augmented the Neo-Classical Solow growth model by incorporating discretionary public spending volatility through total factor productivity. The empirical analysis is based on a panel of 55 countries over the period of 1985-2014. We have estimated the dynamic panel using Generalized Method of Moment (GMM). The overall results confirm that discretionary spending volatility has a negative effect on economic growth. The impact of discretionary spending volatility is negative and significant at aggregated level (i.e the world representative sample) as well as in a sub-sample of developing economies. However, the association of volatile discretionary public spending with economic growth is negative but insignificant in developed economies. This suggests that spending rules (i.e permanent numerical limits on total government expenditure) should be introduce to restrict government from the aggressive use of discretionary policy. Supervisor:- Dr. Nasir Iqbal