Labor Market Polarization with Hand-to-Mouth Households

Over the recent decades, wide-spread automation has led to a shift of the US labor force from occupations intensive in routine tasks into occupations intensive in manual and abstract tasks. I integrate routine-biased technological change into an incomplete markets model with occupation-specific human capital. I use the model to study the transition between steady states pre and post labor market polarization in general equilibrium. When human capital is occupation-specific and wages in the routine occupations relative to the other occupations fall over time, occupational choices become dynamic investment decisions. When households are close to the borrowing constraint, their occupational choices are distorted and they optimally choose to work in the routine occupations for longer than households who have accumulated a buffer stock of savings. I show that in a counterfactual economy, in which all workers choose occupations as if they were hand-to-mouth, the fall in routine labor is protracted by about three years compared to what was actually observed. I use the model to discuss several labor market policies. Incentivizing experienced routine workers to switch to the manual or abstract occupations, by paying them a government transfer, increases social welfare and average output. Empirically, I show that the friction I study is highly relevant, as about 34% of the households working in routine occupations live hand-to-mouth