Abstract
This study aims to estimate an endogenous growth model that will allow the dynamics of the Turkish economy to be followed in the 1990-2020 period and to reveal the gradual structure of this model. The linearity behavior of the macroeconomic indicators, which are the subject of the research, was determined by the Harvey, Leybourne and Xiao (2008) test; stationarity structures were investigated in detail through ADF (1979, 1981), Lee and Strazicich (2004) and Hepsağ (2019) unit root tests and KPSS (1992) stationarity test. In order to deal with the theoretical problem of the estimated macroeconometric model, the residual augmented least squares (RALS) technique was applied in the first step. The overall inferences from this model are that R&D expenditures and export volume have an increasing effect on GDP per capita. For a more detailed and graduated view, the quantile regression technique was used and it was possible to observe the effect of endogenous growth variables on different quantiles of GDP per capita. Findings show that the effect of R&D expenditures on GDP per capita values in lower quartiles is increasing, but it is not statistically significant in the upper quartiles. It has been noted that the export volume has the opposite effect with R&D, and while there is no statistical significance on lower quartiles of GDP per capita, it has an increasing effect in the upper quartiles. Thus, while R&D expenditures are expected to accelerate the GDP per capita, exports are burdened with the acceleration effect in the transition to upper quartiles. Empirical findings provide evidence for the validity of R&D based endogenous growth and export supported growth theory.