I have a hypothesis that productivity growth is tied to the capital / income ratio. That is: if the percentage of income that goes to capital vs labor increases, the productivity of the laborers rises at a slower rate. But I obviously can't substantiate that.