The Quarterly Journal of Economics has a good paper out called "Human Capital Risk and Economic Growth." I found a working version of the paper here. The basic argument is that households must choose what to do with their money, and investing in human capital (new skills, extra degree etc.) is riskier than investing in physical capital. Surprisingly, it seems that government policies that make severance payments to displaced workers lower this human capital investment risk, thus increasing investment in human capital, which leads to growth. Much of the paper is math analysis, but read the intro if you are interested. By the way, this doesn't seem to work in every case, only those where markets are for some reason incomplete.