But productivity is defined as GDP/unit of labor. If it takes less labor to make a unit of GDP (because less inputs were needed), then the productivity statisti

But productivity is defined as GDP/unit of labor. If it takes less labor to make a unit of GDP (because less inputs were needed), then the productivity statistics should show that. They shouldn't miss it. Even if those efficiency gains cause the prices of output to fall, inflation adjustment should, in theory, address that. So I don't think that greater efficiency can explain the productivity puzzle, at least not by itself.