Comparison of identification schemes. Exogeneity of identified technology shocks. Unit root versus stationary hours. Habit formation in consumption, adjustment costs on investment. Bivariate model, stationary hours specification. Log of hours per capita.