Non linear cointegration approach to the wage ? production relationship in turkey

Since reaching the right predictions, which is one of the essential objectives of econometric models, could only be possible by the right establishment of the model, desired objective was unable to be achieved in models found by the linearity hypothesis, and studies were carried out on nonlinear models and their coefficient predictions. Although there were studies that conducted in the first half of the 20th century on the necessity and predictions of nonlinear models, the requisite of utilizing nonlinear time series models for bringing an explanation to business cycles and conducting the financial time series analyses is now clearly understood. The studies at this point that become prominent are TAR and STAR models.The study aims at analyzing the relationship between the industrial production index and real wages. The starting point is that these variables manifest a nonlinear structure. And in the light of this conjecture, it is held that the testing and developing theories could only be possible through nonlinear models. Thus, the relation will be analyzed by nonlinear time series methods.These methods can be summed under three headings as single-variable nonlinear time series models, stability analysis of nonlinear time series, and the cointegration analysis in nonlinear time series. Following theoretical explanations within this scope, the conducted tests were conventional unit root tests, TAR unit root testing, and then TAR cointegration analysis, all performed in the application section.