The relationship between social spending and gdp: Case of Turkey (1980-2014)

In this study, the relationship between public social spending GDP in Turkey 1980-2014 was examined. Annual data covering the years 1980-2014 were used in the study. The real social expenditure series per person is used as a representative of social expenditures and the real GDP series per person is used as the level of economic activity. In the analysis section, the Augmented Dickey-Fuller test was first applied for the stationarity test. According to stationarity test results, both variables were found to be stationary in the first difference. After the stationarity test, the Engle-Granger cointegration test was applied to analyze the long-run relationship between social spending and GDP. As a result of the cointegration analysis, it was concluded that there was no long-term relationship between the variables. After examining the long-run relationship dimension, the Granger Causality test, which examines the relationship in the short term, was applied. Granger causality test results did not find any relation between social expenditure and GDP in the short term. Short and long-term test results were parallel and Impulse-Response and variance decomposition analyzes were conducted at the last stage. According to the results of the Impulse-Response analysis, an exogenous shock that would occur in real per capita social spending was not lead to any increase in the per capita real GDP trend. In the same way, it has been determined that an exogenous shock, which may occur in real per capita GDP, would not lead to any increase in the tendency of real per capita social expenditure. According to variance decomposition results, all of the variables used in the system are explained by itself and the variables act exogenously. Ultimately all the analysis results, concluded that there was no relation between public social expenditures and GDP in Turkey between 1980-2014 in that the variables are not affect each other. Key Words: GDP, Social Expenditures, Engle-Granger, Variance, Impulse-Response.