We presented our most recent study on US wealth inequality during Annual Lithuanian Conference on Economic Research. Tough audience, great comments!

This paper is not a standard one in our portfolio. We do a novel decomposition. We work with observational data without a structural model. And yet, we find some pretty puzzling results. We start by documenting that the increase in U.S. wealth inequality has been to a large extent driven by rising between-cohort wealth inequality and that its role has been rising over decades. With our decomposition we identify and quantify three key factors driving changes in U.S wealth inequality: (i) an adjustment in saving behavior of successive birth cohorts with rising longevity, (ii) changes in the population structure of existing cohorts that occur because the relative share of each birth cohort is changing, and (iii) the generational exchange as members of older birth-cohorts die (exit) and members of younger birth-cohorts enter the population.

We present two novel findings. First, among the three key factors, the behavioral adjustment in savings is reducing wealth inequality, while the changing population structure and generational exchange are amplifying it. We show that the rise in wealth inequality has been driven by the two demographic factors due to rising longevity. The distribution of wealth within birth cohorts has actually been reducing US wealth inequality throughout the post-war period. As we discuss presently, this finding has important implications for both policy and research. In terms of policy implications, our findings indicate that concerns for equity based on raw measures of wealth inequality have been exaggerating the relevance of non-equitable processes in wealth accumulation. In terms of academic research, our findings indicate that models with infinitely-lived agents are not appropriate in this area of research because by construction they ignore rising longevity and changes in the population structure.

Our second novel finding is that the three factors largely offset one another in affecting between- and within-cohort inequality. This contradicts a misguided inference that between-cohort inequality is not particularly relevant for explaining changes in U.S. wealth inequality. In fact, the actual change in wealth inequality attributable to the three factors is much larger than the observed overall decline in U.S. wealth inequality in the first three decades after World War II and also in its substantial rise since then. Hence, while the public debate focuses on what is perceived as a major rise in overall wealth inequality, we demonstrate that this observed rise is in fact “modest” relative to the massive underlying effects brought about by the rise in longevity.