Testing the validity of efficiency and compensation effect hypotheses for G20 countries

Local economies are more exposed to external risks with the effect of global trade. This pushes governments to tend to protect their citizens. It is seen that governments have to intervene in the economy by increasing public expenditures in order to compensate for the external risks. Therefore, countries with greater trade openness may experience greater government expenditures. In recent years, the issue of protecting the welfare of society against external risks has been one of the issues discussed in the literature in terms of economic theory by economists. This relationship is accepted in the literature as the "Compensation Effect Hypothesis". Although the foundation of the studies in this field was laid by Cameron (1978), the study, which has kept its importance until today by developing the relationship between external openness and public, has been carried out by Rodrik (1996). In this context, studies that shed light on the literature are regarded as the studies by Cameron (1978) and Rodrik (1996). In both studies, the hypothesis that public expenditures will be increased in order to compensate for the negative economic effects in externally open countries is suggested. After the Compensation Effect Hypothesis was accepted in the literature, a study that made an important contribution to the literature in this field with a different perspective was carried out by Cusack (1997). According to Cusack (1997), contrary to the Compensation Hypothesis, there may be a negative relationship between external openness and public expenditure. This situation is accepted as the "Efficiency Hypothesis" in the literature. Finally, in the study of Clark (1940) and Iversen & Cusack (2000), they state that there is no causal relationship between public size and openness as an alternative to the first two hypotheses, as a result of deindustrialization in economies, which is accepted as "Deindustrialization Hypothesis" in the literature. As a result of the findings obtained in the research, cross-section dependence and heterogeneity are determined in the tests performed for the variables and the panel. Bai and Ng (2004) Panic Panel Unit Root Test, one of the unit root tests, is applied to measure the stability of the series. Westerlund & Edgerton (2007) LM Bootstrap Panel Cointegration Tests are used with ECM Error Correction Model, cointegration tests, in order to see whether there is a long-term equilibrium relationship between the variables and whether they act together. Dumitrescu & Hurlin (2012) and Emirmahmutoğlu & Köse (2011) Panel Causality Tests are used to determine the dependence between variables and the direction of this dependence. Finally, if there is a long-term cointegrating relationship between the variables, the parameter estimators AMG (Augmented Mean Group) and CCE (Common Correlated Effects) estimators are used, which show how much a change in the series affects the variables. As a result of the findings obtained, it is understood that the Compensation Effect Hypothesis is valid in USA, Argentina and Japan; the Efficiency Hypothesis is valid in Germany, Australia, China, Indonesia, France, South Africa, South Korea, India, England, Italy, Canada, Mexico, Russia, Saudi Arabia and European Union; and the Deindustrialization Hypothesis is valid in Turkey and Brazil. Key Words: Compensation Hypothesis, Efficiency Hypothesis, Deindustrialization Hypothesis, G20, Panel Data Analysis