Educational attainment explains two-thirds of the variation in economic success among metropolitan areas.

Each additional percentage point increase in the four-year college attainment rate increases metro per capita income by $1,250

For a long time, we’ve been exponents of what we call “The Talent Dividend,” the idea that raising a metro area’s educational attainment is the key to raising productivity, living standards and incomes. Our core metric for assessing the importance of a well-educated population is to look at the relationship between per capita incomes and the four-year college attainment rate.

We’ve been tracking these data, and today we’re updating them to reflect the latest information from the Census Bureau’s just-released 2016 American Community Survey.  We’ve paired this information with the Bureau of Economic Analysis’ estimates of per capita income in each metropolitan area for 2016. The following chart plots the relationship between per capita personal income (on the vertical axis) and the fraction of the adult population who have completed at least a four-year college degree (on the horizontal axis).  Each dot on the chart represents one of the nation’s metropolitan areas with at least 1 million population (53 of them, according to the 2015 Census tabulations).  You can mouse-over a dot to see the corresponding metropolitan area and its educational attainment rate and per capita income.

As you’ll immediately notice, there’s a strong, positive correlation between educational attainment and per capita income.  The metro areas with the highest levels of education have the highest levels of per capita personal income.  Cities like San Francisco, Boston and Washington have the highest levels of per capita income and the best-educated populations. Cities like Riverside and Las Vegas have low levels of educational attainment and correspondingly lower levels of per capita income. The coefficient of deterimination of the two variables–a statistical measure of the strength of the relationship–is .67, which suggests we can explain two-thirds of the variation in per capita personal income among metropolitan areas, simply by knowing what fraction of their adult population has a four-year degree. Most cities lie close the to regression line; a few outliers have plausible explanations for their over or under performance. San Francisco and San Jose lie far above the regression line, and are super-charged (and expensive) high performers.  Raleigh and Austin have incomes lower than their educational attainment would predict, but also have populations that skew very young, and therefore have lower incomes.

This chart tells you the most important thing you need to know about urban economic development in the 21st century: if you want a successful economy, you have to have a talented population. Cities with low levels of educational attainment will find it difficult to enjoy higher incomes; cities with higher levels of educational attainment can expect greater prosperity. As Ed Glaeser succinctly puts it: “At the local level fundamentally the most important economic development strategy is to attract and train smart people.” And critically, because smart people are the most mobile, building the kind of city that people want to live in is a key for anchoring talent in place. And, importantly, the economic research shows that the benefits of higher educational attainment don’t just accrue to those with a better education: people with modest education levels have higher incomes and lower unemployment rates if they live in metro areas with higher average levels of education.

The data presented here imply that a 1 percentage point increase in the four-year college attainment rate is associated with about a $1,250 per year increase in average incomes in a metropolitan area, an increment we refer to as the Talent Dividend.  This cross-sectional relationship suggests that if a metropolitan area were to improve its educational attainment by one percentage point on a sustained basis, that it would see a significant increase in its income.

Over time, the strength of this relationship, and the size of the talent dividend effect has been increasing.  When we computed the relationship using 2010 data, the correlation coefficient was .60 and the size of the talent dividend was $860 (in current dollars).  These data suggest that educational attainment has become even more powerful in determining economic success than just a few years ago.

Education is a stronger predictor of economic success today than ever before. That’s true for individuals, for private businesses, for communities, and for metropolitan economies.  The better educated you are, the more likely you are to be prosperous in a knowledge-based economy. Not only do well-paid and fast growing technology jobs go disproportionately to the better educated, but better educated workers tend to be more adaptable and more innovative, which better prepares them to cope with a changing economy.  The policy lessons for city leaders are clear: a successful economy depends on doing a great job of educating your population, starting with your children, and also building a community that smart people will choose to live in.