We address the noted puzzle that despite increased capital mobility, international consumption risk sharing appears to be very limited. For all possible country pairings, we measure idiosyncratic consumption as the difference between national real per capita consumption expenditures. Using a pair-wise framework based on the time-series properties of idiosyncratic consumption, a probabilistic test for non-stationarity suggests that the extent of risk sharing in fact occurs for a large sample of industrial countries. Further to this, we conduct a probit analysis to confirm a statistically significant positive association between the probability of cointegration between national measures of real per capita consumption and the degree of capital mobility.
