A1090
Title: Systems of multi-factor production functions: Modelling the joint behavior of GDP and emissions
Authors: Sebastian Veldhuis - University of Klagenfurt (Austria) [presenting]
Martin Wagner - University of Klagenfurt and Institute for Advanced Studies, Vienna (Austria)
Abstract: Integrated modified least squares estimation is applied and (fixed-b) inference to systems of cointegrated (multi-output) multi-factor translog production functions. Translog systems are a special case of systems of cointegrated multivariate polynomial regressions. As usual in the cointegration literature, the regressors are allowed to be endogenous and the errors correlated, both serially and across equations. In the application, one of the outputs, GDP, is a good, whereas the other considered outputs, CO2 and SOX emissions, are bad. The considered input factors include capital, labor, and different specifications of total factor productivity. The developed methodology is applied to aggregate data for about 20 countries using annual data ranging from about 1950 to 2020. Emphasis is put on testing economically relevant hypotheses, e.g., constant returns to scale or the Cobb-Douglas functional form of the GDP equation.
