Why We Wrote Moving Toward Integration.

• Author: Sander, Richard
• Position: Fair Housing Past, Present, and Future: Perspectives on Moving Toward Integration

CONTENTS INTRODUCTION I. THE OVERLOOKED STORY OF HOUSING INTEGRATION II. BETTER MEASURES A. Block data B. Migration data C. Economic segregation III. THE IMPORTANCE OF CONNECTING DISCIPLINES IV. RACE IS COMPLEX V. BE CONCRETE ABOUT THEORIES, GOALS, AND SOLUTIONS INTRODUCTION

The problem of African-American housing segregation in America has not suffered from a want of academic attention. Hundreds of scholars have published articles on this problem, many of them insightful and some of them path-breaking. The subject has also generated many books, including two--1993's American Apartheid (1) and 2017's The Color of Law (2)--that have found audiences far beyond the ranks of academics and have each achieved something like cult status. Compared to many other racial issues in the United States, such as employment discrimination or the racial dimension of homelessness, housing segregation has received a good deal of attention. So why was our (rather long) book (3) needed, and what gap did we hope to fill?

There are five answers to that question, and together they not only justify the book's existence (we hope), but largely explain our book's innovations and theses.

1. THE OVERLOOKED STORY OF HOUSING INTEGRATION

   When social scientists try to measure the severity of housing segregation, the most common metric they turn to is the "index of dissimilarity," which measures how racial composition varies across the neighborhoods of a city or metropolitan area. An index measure of "0" means that all neighborhoods have the same racial makeup, while an index measure of "1" means that two groups, such as whites and blacks, live in completely separate neighborhoods with no overlap whatsoever. By this measure, segregation in urban America was severe indeed from 1930 through 1970. In 1970, for example, the black-white index of dissimilarity averaged .92 in the nation's sixty largest metropolitan areas, and was above .85 pretty much everywhere. (4)

   Half a century later, very high levels of housing segregation remain the norm in most American cities, and it is increasingly clear that the highly-segregated lives of African-Americans are the single most important driver of racial inequality--in schools, job markets, health outcomes, crime, and poverty. (5) That is why the subject is so important. But the vast majority of scholars and journalists who write about housing segregation tend to ignore the equally important fact that housing-segregation levels declined sharply in a number of metropolitan areas in the late-twentieth century. Among the sixty largest metro areas, about a dozen now have dissimilarity-index measures that are close to .60. This might sound high, but it represents, in those areas, tremendous progress. The typical African American in those metro areas lives in a highly-diverse neighborhood, and very few live-in neighborhoods that feel segregated. The segregation level of African Americans in those cities is, moreover, comparable to the dissimilarity-index measure of Russian Americans (a predominantly Jewish population) or Chinese Americans relative to the general Anglo (i.e., white non-Hispanic) population, and neither Jews nor Chinese Americans generally perceive segregation to be a significant problem for their groups.

   But there is an even more important reason why achieving a black-white dissimilarity index of .60 in some metro areas is encouraging and important: in every region where something close to this level has been achieved, black-white disparities in every measurable dimension have shrunk. If we compare our most--and least-segregated major metro areas, the black-white gap in test scores is 25% smaller in the integrated regions; the unemployment gap is over 50% smaller; and the mortality-rate gap is over 65% smaller. This is not just a fortuitous correlation; a wide variety of evidence shows that declines in segregation dramatically improve a host of African-American opportunities and outcomes. (6)

   Thus, the story of African-American housing segregation over the past fifty years is really two stories: painfully slow declines in most major urban areas, but rapid and consequential declines in a fair number of other areas. Yet most of those writing in this field have ignored or dismissed this second, "increasing integration" path. If one cares about racial inequality, what could be more important than understanding the reasons for these two paths, and how to get a city from path one to path two?

2. BETTER MEASURES

   Many past claims about both the course and nature of housing segregation have been marred by reliance on poor data and faulty measurements. Often scholars in the field talk past each other because they are not being precise about the phenomena they are describing. Throughout our book, we sought to develop better measures of nearly all the phenomena we were trying to explain. We did not always succeed. Where we did, we were often aided by colleagues who generously shared their unpublished research, and by the Bureau of the Census, which allowed us access to "internal" census data through secure data sites. Here are three examples of how we improved on past measurements, and why they mattered to the book.

   1. Block data.

      Through internal census data, we were able to compute dissimilarity indices for entire metropolitan areas at the block level for five censuses from 1970 through 2010. Most prior literature has used "census tracts" as the unit for measuring segregation. But tracts are much larger than blocks, their boundaries often change from decade to decade, and they often mask patterns of internal segregation. Our block indices gave us precise numbers that were more comparable over time--and across metro areas--than in any of the...