The Spatial Dimensions of Inequality
When we talk about inequality, we are usually referring, directly or indirectly, to the variability of the income distribution or some related characteristic like wealth or access to resources. The rapid increase in income inequality in recent decades is well documented and much discussed. In many ways, however, the day to day experience of inequality is a function of the spatial dimension of inequality. Spatial inequality can occur at many levels. There can be inequality between regions, states, and metropolitan areas driven by variations in the economic bases of those areas. But in the United States, the dominant spatial inequality is between neighborhoods within metropolitan areas. In contrast, discussions of spatial inequality among European scholars tend to focus on regions – Central Europe, Southern Europe, etc. In part, this reflects a far smaller degree of inequality within European cities at the neighborhood level.
Neighborhood inequality has important implications for access to public amenities, the quality of schools, the availability of role models, access to information networks, availability and quality of health care services and many other factors that directly determine the quality of life and opportunities for social mobility. Spatial inequality serves to maintain and reinforce inequality that results from the operation of the labor market, although the degree to which it is a cause rather than just an effect is still open to debate.
Good indicators of neighborhood inequality include racial segregation, economic segregation, and the concentration of poverty. Racial segregation of blacks in the US is still extremely high, despite the slow decline of recent decades. On average across all metropolitan areas, the Index of Dissimilarity of whites from blacks is 60 – meaning that 60 percent of blacks would have to move to achieve an integrated state. This is slightly down from 64 in 1990, but still very high in relation to other groups. In many large cities, 70 to 80 percent of blacks would have to move. In large cities, the vast majority of neighborhoods are either 80 percent minority or 90 percent white. Moreover, school age children are substantially more segregated than adults because the whites who live in urban neighborhoods with minorities tend to be childless. The segregation of school children is further amplified by selection into public and private schools and tracking within schools.
Segregation by income level is increasing within both the white and black communities. Poor blacks are 16 percent more segregated from non-poor blacks than they were in 1990. Poor whites are 8 percent more segregated from non-poor whites. High-poverty ghettos and barrios represent the most extreme form of neighborhood inequality and have now reached unprecedented levels. The concentration of poverty – people who live in very high-poverty neighborhoods – doubled between 1970 and 1990 to 9.6 million people. After the booming economy of the late 1990s, there was a substantial decline to 7.2 million persons residing in high-poverty neighborhoods. However, that decline was illusory. In the most recent data from the American Communities Survey, which covers the period 2007-2012, the number of high-poverty neighborhood residents surged 55 percent to 11.2 million, exceeding the level of 1990. Surprisingly, the fastest growth has been among non-Hispanic whites. High-poverty neighborhoods are less likely to be dominated by a single racial or ethnic group.
The experience of life in such neighborhoods is a different world from the mainstream of American life. These neighborhoods are characterized by high levels of housing vacancy and dilapidation. Nationally, in high-poverty neighborhoods, fewer than 50 percent of men are employed and more than half of children live in female headed families. Children growing up in such neighborhoods observe a different social and economic reality than children in mainstream America. Many of the schools in these neighborhoods are highly dysfunctional, characterized by violence and high dropout rates.
The reconcentration of poverty has gone largely unnoticed. One possible explanation is that, ironically, the concentration of poverty itself has followed a less concentrated pattern. While there are far more high-poverty census tracts now than in 2000, containing far more people, these census tracts are far less concentrated than they were at the least peak of concentrated poverty in 1990. High-poverty census tracts now are more likely to be fragmented into smaller pockets around a metropolitan area, including in older suburbs, rather than in one or two main clusters. The pattern is striking when you compare maps from 1990 and 2009 (see the example of Detroit below), and is also confirmed by the Moran’s I, a measure of spatial clustering.
Why did this happen? The economy matters, of course, as seen by the responsiveness of concentrated poverty to the boom of the 1990s and the Great Recession. However, a major factor is rampant suburban sprawl – in most suburban neighborhoods population was growing by more than 10-50 percent per decade while population in the central areas declined by 20 percent or more. Suburban housing growth far exceeded the rate needed to facilitate population growth, and so it has been cannibalistic. Suburban housing growth was also highly class selective through the use of exclusionary zoning.
As we think about the policy responses to inequality, it is once again useful to return to the distinction between its spatial and non-spatial dimensions. In non-spatial terms, we have to focus broadly on the outcomes of the labor market. To the extent that the inequality of income reflects the distribution of skills and abilities, then policies must address the distribution of those skills on the front end through education and training. Policies can also address compensation directly through policies like the minimum wage, unemployment compensation, and so on. On the back end, the distribution of income is adjusted through redistribution and progressive taxation. But there is to a certain extent an unavoidable degree of inequality that will result in a capitalist economy. Our policies can serve to temper and reduce that inequality and counterbalance the strong tendency for that inequality to reproduce itself.
In contrast, the spatial dimension of inequality, at least within regions, is substantially determined by public policies at various levels of government. Land use policy, zoning, planning, growth management, mortgage lending, highway construction, and infrastructure construction and subsidies are just a few of the myriad ways that government determines the inequality from one neighborhood to the next. Governmental fragmentation combined with a laissez faire approach to local development facilitates sprawl and exclusion. The distribution of housing of various types within and between municipalities and within metropolitan areas is not the outcome of a free market, but is driven by explicit and implicit policy regimes.
We take the current distribution of neighborhoods for granted because this is the way we have always done things. We think of it as inevitable. But other nations do it differently and get vastly different outcomes in terms of neighborhood differentiation. Because the construction of urban space is determined by policy, governments and institutions have a special obligation to address the spatial dimension of inequality. In the US, we have abdicated this responsibility.