In this study, we empirically investigate the long-run relationships between education-health expenditures, defence expenditures and economic growth in Turkey by using annual time-series data for the time period 1950-2005. To estimate these relationships, we use the cointegration method, which allows for the existence of structural breaks. The results show that there are two cointegrating vectors in the system with statistically significant structural breaks. According to the identified long-run equations, education and health expenditures affect economic growth positively, while defence expenditures have negative effects. In addition, the results show that there is a budgetary trade-off between education-health and defence expenditures.

This study aims to analyze the perception of voters. It attempts to answer whether the power of purse defined at the constitutional level has been internalized and used functionally by the voters or not. A scale of 45 statements has been used to measure the perception of the voters. Factor and regression analyses have been conducted on the data obtained from the scale and the power of purse perception not functi level of the voters has been investigated.The results of the study put forward that voters in Turkey can onally use the power of purse that they get by the constitution. Also it has been discussed in the study that 6 factors have been effective for the result and that they have the same and positive effects on this perception.