We compare trends in mortality, nutritional status and food supply to other living standard indicators for the early years of the Nazi period. We find that Germany experienced a substantial increase in mortality rates in most age groups in the mid-1930s, even relative to those of 1932, the worst...

We use two approaches to examine the macroeconomic consequences of disruptions in global food commodity markets. First, we embed a novel quarterly composite global production index for the four basic staples (corn, wheat, rice and soybeans) in a standard vector autoregression (VAR) model, and we...

What determines the structure of labour market institutions? This paper argues that common explanations based on rent sharing are incomplete; unions, job protection, and egalitarian pay structures may have as much to do with social insurance of otherwise uninsurable risks as with rent sharing and...

We suggest a political economy explanation for the stylized fact that intragenerationally more redistributive social security systems are smaller. Our key insight is that linking benefits to past earnings (less redistributiveness) reduces the efficiency cost of social security (due to endogenous...

argue that the offsetting effect of social security contributions on household retirement saving depends on how closely the social security programme imitates a private retirement saving plan (i.e. the ‘actuarial’ component of the social security programme) the closer the design of the...

Estimating saving and fertility simultaneously by the VAR method, we find that social security cover has a positive effect on household saving, and a negative effect on fertility. In Germany, as in other countries where the hypothesis was tested, social security is thus good for growth. A...

We find evidence in the OECD cross-country data to support the Knightian view that non-diversifiable economic risks shape equilibrium entrepreneurship in an occupational choice model. Differential social insurance of entrepreneurial and labor risk is found to be statistically significant and...