Investigating the relationship between government size and economic growth in Iran: An Application of state-space and Auto Regressive Distributed Lags Models

Barro (1990) by adding government spending into the growth models showed that the amount of government activities have a positive impact on economic growth, But if government spending is increased over a certain size, government activities will have a negative impact on economic growth. In this direction, this paper by using theoretical Barro growth model and an empirical model for Iranian economy tries to investigate the impact of current and capital government expenditure on output growth using ARDL approach and state-space models applying quarterly data during 1967-2014. First, using Lumsdaine-Papell (1997) unit root test and Gregory-Hansen (1996) and Saikkonen and Lutkepohl (2002) cointegration test we determine the degree of integration and cointegration of the variables. The results indicate that there are structural breaks in the variables under study and these breaks affect the relationship between variables. Then we use ARDL model, considering structural breaks, to determine threshold level for current government expenditure which is 15.2 percent and for capital government expenditure which is 8.2 percent of GDP per head. Regarding Lucas theoretical critique and empirical structural breaks in the Iranian economy, we use state space model to investigate relationship between growth and the government size and the results indicate that coefficients are not stable during time and they behave differently regarding the source of the shock.