What factors influence social mobility? Raj Chetty and his co-authors released a first-of-its-kind dataset, the Opportunity Atlas, allowing researchers to compare upward mobility across Census tracts. They are able to do this by linking administrative tax records with census data, and find sharp differences in children’s outcomes at very fine levels of geographical granularity. Neighborhoods can have very different outcomes across racial groups, for instance; but upward mobility is not associated with job or wage growth (see "Top chart" below). The neighborhood effect is clearly causal, because children who move to a better neighborhood early in life have better adult outcomes. A child from a family that moves from a low-opportunity neighborhood to a high-opportunity neighborhood will contribute up to $41,000 more in tax dollars over her lifetime. The Atlas is the first product from the rebranded shop at Harvard led by Professor Chetty, Opportunity Insights, for social scientists their website is now a “must-see”. 

It’s a question perplexing many of us: What difference does it make where you go to college, or what major you select? Using matched administrative data, Rajashri Chakrabarti and Michell Jiang set out to answer this question, in terms of employment and earnings, both medium and long term, and upward economic mobility. Selective schools produce better earnings than non-selective schools, private not-for-profit schools do better than public and for-profit schools, and STEM and business majors beat out arts and vocational majors.They also find an increased selectivity premium in the long-term, which the wage premium of selective schools over nonselective schools almost doubles. There are inequality implications here too: selective schools decrease the earnings gap between the top and bottom family-income terciles by 43 percent, while for-profit schools widen the gap by 117 percent.

Some people are worried about income inequality; others about median incomes; many about both. But what is the relationship between the two? Using data from LIS, OECD, and WID, Stefan Thewissen and his co-authors find that median income growth is negatively correlated with a country’s Gini coefficient (as inequality goes up, median income growth declines). But importantly, given the frequent framing of inequality as a “bottom 99 percent issue,” they find no relationship between median income growth and share of income going to the top one percent. They also find that economic growth (measured as GNI) is strongly correlated with median income growth. This is not a surprise of course, but it useful confirmation that the condition of the middle class and the state of the economy as a whole are tightly intertwined.