A note on the Bautin bifurcation in the modified Romer model with endogenous technical change