Tuesday, November 11, 2014

Are We All Ferguson?

How unique is the situation in Ferguson, Missouri - racially speaking?

What does the shooting of Michael Brown in Ferguson, Missouri and the ensuing debates and protests say about race relations in America?

Salem State's Center for Diversity and Multicultural Affairs recently organized a panel and discussion on the issues raised by Ferguson, which I had the privilege of being on, along with three co-panelists: Dr. Tiffany Chenault – Sociology, Dr. Forrest Rodgers – Criminal Justice, and Deputy Director Rahsaan D. Hall Esq. - Lawyers' Committee for Civil Rights and Economic Justice.

Each of the panelists was given full license to talk about any aspect of the issues raised, and I focused my 15 minutes on the issue of segregation. As a geographer, I thought it appropriate to look at how racial tensions play out spatially.

Before I get too far, let me lay out the more counter-intuitive take-home points I want to make:

  • The US is an increasingly diverse nation, and it is still very segregated
  • The most segregated racial group is Whites
  • Race relations in America are not Black and White; they are multi-racial and multi-ethnic 
  • School children are the most diverse and the most racially isolated amongst us
  • Our regional stereotypes of where segregation is a problem, and where it is not, are probably wrong
  • Ferguson, Missouri is no different than any other community in America, and that's the problem

Measuring Segregation

What is segregation?  At its most basic, segregation is the separation of groups of people - by race, class, religion, etc - within a country, community, or establishment. Actually identifying and measuring segregation, however, is not so simple.  In fact, researchers have been debating how to measure segregation for decades, and dozens of segregation indices have been proposed. In the late 1980s, Douglas Massey and Nancy Denton, two lions in the field of race studies, cleared the deck somewhat by performing a systematic analysis of 19 of these segregation indices. They concluded that segregation is not a singular phenomenon, but rather can be understood as consisting of 5 dimensions of spatial variation:

  1. Evenness
  2. Exposure
  3. Clustering
  4. Concentration
  5. Centralization

Nearly two decades later, Sean Reardon and David O'Sullivan proposed an even simpler conceptualization of  residential segregation consisting of only 2 dimensions:

  1. spatial exposure/isolation: the extent that members of one group encounter members of another group (or their own group, in the case of spatial isolation)
  2. spatial evenness/clustering: the extent to which groups are similarly distributed
Reardon and O'Sullivan offered the following graphic to illustrate this two-dimensional idea of segregation.

Reardon and O'Sullivan. 2004. “Measures of Spatial Segregation.” Sociological Methodology 34(1):121-162.

The authors explain the graphic this way:

"In the upper half of the diagram are two patterns where black and white households are evenly distributed throughout space. Both of these patterns have low levels of spatial clustering (or high levels of spatial evenness). In the pattern on the upper right, however, there are more black households in the local environment of each white household (and vice versa) than in the pattern on the upper left; this means that the white-black exposure is higher on the right, and the white isolation is higher on the left. In the bottom half of the figure, both patterns show greater clustering—but roughly the same levels of exposure—than the corresponding patterns above."

Here is another way of looking at it.


Imagine a town (large box on the right) in which 3/4 of the people are blue and 1/4 are red. The town's overall population is represented by the larger people figures on the left. Now, assume that the town is made up of 4 neighborhoods, each of which is represented by the smaller, interior boxes. Imagine that each of these neighborhoods is also made up of 3/4 blue people and 1/4 red people.  The populations of the individual neighborhoods are represented by the smaller people figures. In such a scenario,we can say that the groups are evenly distributed across the town because each neighborhood has the same proportion of blue to red people as the town as a whole. Red people have a high degree of exposure to blue people, but not to other red people. In any given neighborhood, the average red person is exposed to three blue people. By contrast, the blue people have less exposure to red people - only one out of three other people - and more exposure to their own kind.

Here is an alternative scenario with the same overall town population of 3/4 blue people and 1/4 red.

In this case, three out of the four neighborhoods consist entirely of blue people, and all the red people are confined to the fourth neighborhood.  It's the same overall town population as the previous example, but the distribution of the groups amongst the neighborhoods is much different. In this scenario, we can say that there is a very uneven distribution (i.e. clustered). It is uneven because the group proportions in each neighborhood are very different from the proportions of blue and red people for the town as a whole. In addition, there is even less exposure of the two groups to one another, and a lot more exposure to their own kinds (i.e. they are isolated). This latter scenario is what I think most of us imagine when we think of segregation.

Growing Diversity, Persistent Segregation

How segregated is America, and is it becoming more or less so?

In evaluating the degree of segregation, we start by looking at the aggregate demographics of the country as a whole.

PolicyMap http://www.policymap.com/

The US is a mostly White country - approximately 63% non-Hispanic White in 2013, and conversely, 37% non-White (i.e. Hispanic, non-Hispanic Black, Asian and others).


PolicyMap http://www.policymap.com/

If there was a completely even distribution of these groups (i.e. not segregated), we would expect every state in the country to have the same proportions of White and non-White people as the country as a whole. This is clearly not the case.  If we change the scale of our analysis, looking at counties rather than states, we see a more complex pattern, but the basic unevenness of the distribution is still plainly evident.


PolicyMap http://www.policymap.com/

The country's demographic profile is always changing of course.  The US population continues to grow larger, albeit increasingly slowly. Both the White and non-White groups are growing, although that growth is also uneven, both geographically and mathematically. The greatest growth of the White population is occurring largely in the West and Southwest, largely as a result of internal migration.


PolicyMap http://www.policymap.com/

By contrast, growth of the non-White population is greatest in those areas that are currently most White - central and northern parts of the country and Northeast.


PolicyMap http://www.policymap.com/

Equally important, the non-White population is growing much faster than the White population, as a result of both greater natural increase (i.e. births outpacing deaths within the country) and immigration. These two facts - the geographic spread and greater mathematical increase of the non-White population - have led many to conclude that the US is becoming an increasingly diverse nation.


PolicyMap http://www.policymap.com/

This is correct. The US is increasingly diverse because the number of racial-ethnic groups in the population is increasing and the sizes of these groups relative to each other is evening out.  However, increasing diversity of the nation is not the same as decreasing segregation of its neighborhoods.

The US2010 Project, which sponsors research to analyze demographic trends, including segregation, has identified the following trends in its most recent reports on segregation:

  • Almost all communities have grown more diverse, although there is still a wide range of diversity profiles, from predominantly white communities (a shrinking number) to minority-majority and no-majority ones (an increasing number).
  • Black-White and Asian-White residential segregation have been declining consistently since 1990
  • In racially diverse metro areas, Whites increasingly occupy diverse neighborhoods
  • In racially diverse metro areas, about half the Black population and 40 percent of Hispanics still live in neighborhoods without a White presence
  • Since 1990, the highest residential segregation has remained between Blacks and Whites
  • Hispanic-White residential segregation has been on the increase since 1990, especially in the West.

Segregation At a School Near You

When we talk about segregation, we tend to talk about residential segregation. However, segregation is a pernicious problem that manifests in a variety of spaces, like schools. Segregation in schools is surprising for at least two reasons: 1) this is the part of the population where demographic change is really most evident and 2) this is a space that the country has been deliberately trying to desegregate for more than half a century!


The Civil Rights Project at UCLA. 2014. Brown at 60: Great Progress, a Long Retreat and an Uncertain Future.

While a little over 1/3 of the US population is non-White, half of school children are now non-White. The school population is the leading edge of a demographic wave that is sweeping this country.

However, the experience of the average school child belies this monumental fact.


The Civil Rights Project at UCLA. 2014. Brown at 60: Great Progress, a Long Retreat and an Uncertain Future.

The racial exposure of children of different races is still very uneven, despite the increased diversity of the population of school children as a whole.  In the graphic above, note that the average White student (first column) attends a school where 72.5% of the children are White. Black students (second column) attend schools where about 48.8% of children are Black and 27.6% are White. Latino students (fourth column) are in schools that are 57% Latino and 25.1% White.  Asians (third column) are the only racial group (albeit a very large and diverse one) that is exposed more to other groups (especially Whites) than to their own. Asians are a special case who warrant a separate discussion.

In general, we can say that school children in America (except for Asians) have relatively low exposure to other racial groups. Latino school children are the non-White group that is most isolated; least exposed to the dominant group (i.e. Whites) and most exposed to their own racial group. However, it is White school children who are least exposed to other racial groups, and thus most racially isolated. We might even go so far as to say that White school children are actually the most segregated group. But we need to be careful with how we interpret the latter statement.

As the authors of Brown at 60: Great Progress, a Long Retreat and an Uncertain Future point out, "Segregation is typically segregation by both race and poverty. Black and Latino students tend to be in schools with a substantial majority of poor children, but White and Asian students are typically in middle-class schools." This poverty is not just of the individual children, but of the school and the community and their resources and opportunities. The impacts of segregation are not equal, which is one big reason why the Civil Rights Movement has fought so hard against the idea of "separate but equal."

The crusade to desegregate American schools has resulted in major change, especially in the South. In the wake of court-rulings forcing desegregation and other desegregation policies starting in the mid 1960s, schools in the South have become much less segregated.


The Civil Rights Project at UCLA. 2014. Brown at 60: Great Progress, a Long Retreat and an Uncertain Future

However, these accomplishments are fragile. Since the late 1980s, school desegregation policies have been deliberately dismantled as a result of new court rulings, and changes in attitude by the political leadership. As a result, schools have been re-segregating. However, the current state of affairs is still significantly better than it was 50 years ago before these policies were implemented. All of this speaks to the power of policy.

What's equally interesting is how these trends play out on a regional basis. While the South has long suffered under the burden and stigma of its history of racism and segregation, it is now the model of integration, and it is the rest of the country that needs to take a look in the mirror.


The Civil Rights Project at UCLA. 2014. Brown at 60: Great Progress, a Long Retreat and an Uncertain Future

The graphic above shows the level of racial segregation for Black children in schools for different regions of the country between 1968 and 2010. Over the last 40+ years, the South has experienced the largest drop in segregation, followed by the West, and these two regions now have the lowest levels of school segregation for Black children. The Midwest and Border regions (i.e. the historic border between "free" and "slave states") have had modest decreases in segregation. The Northeast, by contrast, has only grown more segregated since the late 1960s and now represents the most racially segregated region of the country for Black school children. This may be a little upsetting to people who still harbor the myth of the liberal and tolerant Northeast (too often set in contrast to the stereotype of an irredeemably racist and intolerant South).

Ferguson, Missouri

How segregated is Ferguson, Missouri, and is it becoming more or less so?

Ferguson is a smallish suburban city (population ~21,000)  in St. Louis County, Missouri. It is part of the Greater St. Louis metropolitan area and a part of the Midwest.  While most of St. Louis County is White, Ferguson and neighboring towns are predominantly Black. According to the ACS 2008-2012, Ferguson is 67% Black and 31% White.

PolicyMap http://www.policymap.com/

Blacks were once a minority in Ferguson. In fact, in 1990, White residents made up 74% of the population, while Blacks were only 25% - the opposite of today's situation. But the city's demographics were shifting; White families were moving out to surrounding suburbs, and Black families were moving in.  By 2000, Ferguson's White population had declined by nearly 50%.  Whites were still a majority in 2000, but the trend was clear. This process by which communities change racially is sometimes described with the benign label of "neighborhood succession," although it may often be described as "White flight."  It is a an all too familiar phenomenon, one process by which the geography of segregation is created and maintained.

The most commonly used measure of segregation is the dissimilarity index. The dissimilarity index reflects the relative distributions of two groups across neighborhoods within a city or metropolitan area; a measure of evenness. It can range in value from 0, indicating complete integration, to 100, indicating complete segregation.  According to the Population Studies Center at the University of Michigan, the greater St. Louis metropolitan area had a 2010 Black-White dissimilarity score of 72.3, meaning that 72.3% of White (or Black) people would need to move to another neighborhood to make Whites and Blacks evenly distributed across all neighborhoods. 

CensusScope http://www.censusscope.org/us/m7040/chart_exposure.html

The level of segregation, as measured by this dissimilarity index, has actually dropped by about 6% since 1990. However, the St. Louis metro region is still the 7th most Black-White segregated major metropolitan region in the US.


CensusScope http://www.censusscope.org/us/m7040/chart_exposure.html

Not surprisingly, the racial exposure of different groups in the St. Louis metro region is also uneven. In the figure above, the rightmost column shows the racial composition (or diversity) of the metro area as a whole. The first five columns represent the average racial composition of the neighborhood of a person of a given race. Note that the racial diversity of the St. Louis metro region is not reflected by the racial exposure of its various groups. If racial exposure matched racial diversity, you would expect all six columns to be identical; that a member of any racial group was exposed to roughly 77% White residents and roughly 22% Black residents in her neighborhood. This is clearly not the case. Amongst non-White groups, Blacks are the most racially isolated. However, Blacks are still less racially isolated than Whites. Whites, once again, are the most racially segregated group; the group with the least exposure to other racial groups.

Are We Ferguson?

How segregated is my community - Salem, Massachusetts - and is it becoming more or less so?

Salem is a midsized suburban city (population ~42,000)  in Essex County, Massachusetts. It is part of the Greater Boston metropolitan area and a part of the Northeast.  Like most of Essex County, and the Commonwealth of Massachusetts, Salem is mostly White, although less so than the county or the state; Salem is 74% non-Hispanic White compared to 80% or more for the county and state. Within Salem, the largest non-White group is Hispanic. Hispanics make up 15% of the city's population, and Blacks about 5%.


CensusScope http://www.censusscope.org/us/m7040/chart_exposure.html

According to the Population Studies Center at the University of Michigan, the greater Boston metropolitan area had a 2010 Black-White dissimilarity score of 64, meaning that 64% of White (or Black) people would need to move to another neighborhood to make Whites and Blacks evenly distributed across all neighborhoods. This level of segregation has declined since 1990. Out of the 102 largest metro regions, the Boston metro region is the 27th most Black-White segregated metropolitan region in the US.

What about Hispanics? The Boston metro region had a 2010 Hispanic-White dissimilarity score of 59.6, meaning that 59.6% of White (or Hispanic) people would need to move to another neighborhood to make Whites and Hispanics evenly distributed across all neighborhoods.While this segregation score is less than that between Whites and Blacks, this Hispanic-White score makes the Boston metro region the 5th most Hispanic-White segregated metropolitan region in the country. Salem is not unlike the rest of the Boston metro region in this respect.

CensusScope http://www.censusscope.org/us/m7040/chart_exposure.html

As with the St. Louis metro region and Ferguson, the racial diversity of Salem is not reflected by the racial exposure of its various groups.  If racial exposure matched racial diversity in Salem, you would expect all six columns to be identical; that a member of any racial group was exposed to roughly 74% White residents and roughly 15% Hispanic residents in her neighborhood. This is clearly not the case. Amongst non-White groups in Salem, Hispanics are the most racially isolated. However, Hispanics are still less racially isolated than Whites. Whites, once again, are the most racially segregated group; the group with the least exposure to other racial groups. 

Again, we must keep in mind that the impacts of segregation, and its meaning, are not equal.

Are We All Ferguson?

The country is certainly diversifying, but segregation is a persistent and real phenomenon. It is not an accidental thing, and its impacts are not the same for each racial or ethnic group. Are we all Ferguson? In too many respects, yes.

The biggest questions that we are left with are familiar ones: why does this continue to happen and how can it be addressed for the long term harmony of society?

Friday, September 26, 2014

Overpopulation and climate change - too many Americans?

In a recent article timed to coincide with the UN Climate Summit in New York City, The Economist offered a ranking of “20 policies and courses of action according to how much they have done to reduce the atmosphere’s stock of greenhouse gases." A number of their choices were unexpected, such as the implementation of China’s one child policy.

China's 1-child policy ranks 4th in the list - allegedly preventing 1.3 billion tons of CO2 from entering the atmosphere. This estimate is taken from a 2007 Reuters article in which "Su Wei of China’s foreign ministry said that his country’s one-child policy, by reducing the number of births between the late 1970s and the mid-2000s by 300m, had reduced carbon emissions by 1.3 billion tonnes in 2005 (because there were fewer people to consume goods which generated greenhouse gases in their production)."

If you do the math, this estimate assumes per capita emissions of 4.33 tons of CO2 for each of those 300 million people whose births were avoided, which is a number consistent with the World Bank's estimate of China's per capita emissions in 2005. Although the The Economist article by no way says this, the implication is that the world is better off with 300 million fewer Chinese.

However, one offensive example deserves another. Had we arrested population growth in the US over the same period, 1970 - 2005, there would have been 95 million fewer Americans in the world. Using the per capita emissions of the US in 2005 - 19.7 tons per person - would have prevented 1.9 billion tons of CO2. This is 1.5 times as much GHGs as that avoided by the Chinese!

One take-home message could be that more would have been accomplished by reducing the number of Americans than Chinese. Maybe America needs a 1-child policy - or the oppressive government that such a policy would require. The last is meant sarcastically.

As far as I can tell, no one is seriously offering population control as an explicit climate change mitigation policy. Nevertheless, I think the issue of population growth and its environmental implications is something worth highlighting. Indeed, fear about overpopulation is a perennial worry. There are legitimate concerns about population growth, but these are mostly to do with its relationship to the rights and status of women and local issues of economic development.  The assumption that population growth is a leading cause of global environmental degradation is really problematic because it misses (or deliberately hides) the radical inequalities in consumption between different populations. Some births have more environmental consequences than others.

It is frequently pointed out that China and India are the most populous countries on the planet (1.4 and 1.3 billion, respectively), but it is not often pointed out that the US is the third most populous country on the planet (319 million). But what's more important, from an environmental perspective, is the relative impact of these populations, not their sizes.

Consider the impact if America reduced its per capita emissions to levels similar to China, or better yet, India.
  • In 2010, American emissions of CO2 were 17.1 tons per capita, while in China they were 6.2 tons per capita, and in India, 1.7 tons per capita.
  • Each American, on average, emitted CO2 equivalent to 2.75 Chinese or 10 Indians.
  • If America reduced its per capita emissions to levels similar to that of China (holding Indian emissions steady), there would be roughly 3.5 billion tons less of CO2 in the world annually. If we treated this 3.5 billion tons of avoided CO2 as "room" for other people to exist, this would be equivalent to making room for another 500 million Chinese, or another 2 billion Indians! 
  • If the US reduced its per capita emissions to that of India (holding Chinese emissions steady), there would be "room" for another 790 million Chinese or 2.9 billion Indians!

This idea of "making room" for other people is not so radical as it might sound. Indeed, "making room" is implicitly what arguments about population growth are often really about, at least when invoked in the context of global environmental problems. But the question is: making room for who and for what?  For wealthy countries to continue a particular way of life with its high environmental impact? For poor countries to develop economically and reduce mass poverty? 

This question touches on a core debate that has plagued climate change negotiations since the 1990s; how to fairly allocate responsibility for climate change mitigation, especially between more developed and less developed countries. A large part of the problem is that consumption (as in per capita emissions of CO2) is still a reliable proxy for wealth. Wealthier countries (and individuals) consume more (emit more CO2 per capita) than less wealthy countries (and individuals). So long as this relationship between wealth/development and consumption hold, we have an environmental and a moral dilemma. This is the core challenge.