Over the past decades, the economic interdependence of national economies has increased to unprecedented levels in world history. In many countries, regulatory and other restrictions on international trade and capital mobility have been weakened or lifted, while at the same time there have been major reductions in transportation costs, a wave of financial innovations, advances in international banking, the establishment of supranational institutions, improvements in international policy coordination, and many more.
Unfortunately, the anti-globalization movement has hijacked the debate on globalization with populist statements that blame trade and financial globalization for rising inequality and increased poverty. It is therefore important to emphasize the many benefits related to globalization. One of these is the possibility for economies to share risk internationally and, hence, to smooth the consumption of their residents in the face of domestic income shocks.
To counter populist arguments and make the case that trade and financial globalization have increased risk sharing opportunities and have thereby contributed to augment the welfare of households, it is first and foremost important to measure risk sharing convincingly and correctly.
This paper therefore proposes a new approach to measure international consumption risk sharing. To this end, we estimate a rich consumption-income framework for fifteen industrialized economies over the period 1875-2016 that allows for time-varying and country-specific parameters and variances. We propose and calculate risk sharing measures as the share of the variance of country-specific consumption growth that is driven by idiosyncratic (income) risk. These variance ratio measures of risk sharing are more encompassing compared to the conventional measure, which is calculated as the impact of idiosyncratic income risk on country-specific consumption growth, as they take into account all potential sources of time variation in the model. This generality is important as our results show that changes in risk sharing come about through the sensitivity to income shocks (as measured by the parameters in the model) but especially through the size of these shocks (as measured by their variances). We find average degrees of risk sharing (across time and countries) of about 35% for the conventional risk sharing measure, of about 45% for our lower bound risk sharing measure and of about 65% for our new encompassing risk sharing measure. We also find that the average degree of risk sharing across countries has increased both during the first era of globalization (1870-1914) and during the recent globalization period. We further give economic content to our risk sharing measures by linking them successfully to trade and financial openness.
- Professor
- More information