How far have we come since the fair housing act? Black-white segregation in the last 45 years

Over the last 40 years, segregation decreased gradually every decade and fell by 27 percent overall. On the whole, this is good news. The 2010 average dissimilarity score of 55.6 effectively means that 55.6 percent of whites would have to move for the mix of whites and blacks to be the same in every neighborhood. While this figure is large in absolute terms, the trend over time has been a steady move toward greater integration. As usual, however, the trend hides large differences between metros.

Larger metros tend to be more segregated than smaller metros. This was true in 1970 and it is true in 2010. In 2010 the top decile, by population, of metro areas had a mean dissimilarity score of 58. The smallest decile had a mean dissimilarity score of 43. America's 10 largest metro areas are also more segregated than the national average. In this group, all three northeast metros (Boston, New York, and Philadelphia) remain highly segregated and experienced slower declines in segregation than the national average. New York actually became slightly more segregated over the last 40 years. In 2010, Dallas-Fort Worth was the only top 10 metro to be less segregated than the weighted national average.

In 2010, metros with small black populations were also likely to be more integrated. In that year, the five least-segregated metro areas were Missoula, Montana; Provo-Orem, Utah; Boulder, Colorado; Las Cruces, New Mexico; and Fort Collins-Loveland, Colorado. None of these areas are more than 2 percent black. The five most segregated metros in 2010 were Milwaukee-Waukesha-West Allis, Wisconsin; New York-Northern New Jersey-Long Island, New York-New Jersey-Pennsylvania; Chicago-Naperville-Joliet, Illinois-Indiana-Wisconsin; Detroit-Warren-Livonia, Michigan; and Cleveland-Elyria-Mentor, Ohio. All of these areas are more than 15 percent black. The decile of cities with the largest black populations had a dissimilarity score of 0.50, while the decile with the smallest black populations had a dissimilarity score of 0.27.

This wasn't always the case. In 1970, Spartanburg, South Carolina, was the least segregated metro area in America at 21 percent black. In that same year, Wausau, Wisconsin, was the most segregated with a grand total of 10 black residents, much less than 1 percent of the population. In 1970, 1980, and 1990, there was no connection between a city's percentage of black residents and how segregated it was. Then, the starting in 2000, metros with small black populations were more integrated than counterparts with larger black populations. This pattern held even more strongly in 2010.

Something-it's unclear what-changed in the last 20 years in metros with small black populations. On average all metros reduced segregation by 10.5 points. The 10 percent of metros with the smallest black populations reduced segregation by 25 points, while the 10 percent with the largest black populations reduced segregation by just 4 points.

Where does all this leave us 45 years later? More than anything it leaves us with more questions. We are a less segregated country than we were 10 years ago, and we are an even less segregated country than we were 45 years ago. That's the good news. But much of this change is being driven by metros with small black populations. It is positive that these places are becoming more integrated. It is also troubling that metros with large populations, and large black populations, aren't seeing that progress. New York, Chicago, and Philadelphia are all hardly more integrated then they were 40 years ago. What changed in metros with small black populations? Why aren't large, diverse metro areas integrating faster? Most importantly of all, what can we learn from areas that are integrating faster and how can we apply these lessons to places where change has stalled?