Thursday, March 3, 2011

Mapping Urban Inequality: Using the Gini Coefficient to Measure the Urban Divide

National Income Inequality Using Gini Coefficients, from CIA World Factbook, 2009.

Many of you have most probably seen maps of the Gini coefficient used to measure income inequality in nations.  (See the map above. If you don’t know what the Gini coefficient is, I’ll explain it in a minute.) 
            The United Nations Human Settlements Programme (known as UN-HABITAT) has put together a rather different perspective on the Gini coefficient.  Rather than measuring income inequality as averaged for an entire nation, they have teased apart the income inequality that exists in urbanized areas - an important thing to know, now that more than 50% of the world lives in cities.  Some of the most alarming and extreme statistics for income inequality exist in cities, and these facts are masked when looking at national averages. 
Even more surprising, in light of the extreme income inequality usually found in the developing countries, are the high Gini coefficients (i.e., high income inequality) found in U.S. cities.  This would not be apparent from just looking at the national Gini.  “The most surprising variations between national and city-specific Gini coefficients of income or consumption disparities are found in the United States of America, where around 2005 the national coefficient stood at 0.38, but exceeded 0.50 in many major metropolitan areas including Washington, DC; New York City; Miami; and others.  These values are comparable to the average Gini coefficients of cities in selected Latin American countries, where income inequality is particularly steep,” (UN-HABITAT, 2009:xiv).  And, in developed countries like the U.S., overall the national Gini coefficients have increased fairly steeply from the mid-1980’s to 2005. 

African cities have the highest Gini coefficient in the world, averaging 0.58.   Sub-Saharan Africa has the highest degree of poverty in the world, and the highest prevalence of slum populations in urban areas.  From: UN-Habitat “State of the World's Cities 2010/2011 - Cities for All: Bridging the Urban Divide”

        How is income inequality measured using the Gini coefficient, and indeed, what is the Gini coefficient?  The Gini coefficient is a number between 0 and 1, where 0 corresponds with perfect equality (where everyone has the same income) and 1 corresponds with perfect inequality (where one person has all the income, and everyone else has zero income).  It’s a measure of statistical dispersion – the inequality of a distribution - invented by an Italian statistician, Corrado Gini, about 100 years ago.  A low Gini coefficient indicates a more equal distribution, and a high score indicates a very unequal distribution.  While the Gini coefficient can be used in many disciplines (for instance, in ecology to measure biodiversity) it is most commonly used in economics, and specifically, to measure income inequality.  

Diagram from Southern African Regional Poverty Network

“Graphically, the Gini coefficient can be easily represented by the area between the Lorenz curve and the line of equality. On the figure to the left, the Lorenz curve maps the cumulative income share on the horizontal axis against the distribution of the population [households] on the vertical axis. In this example, 70 percent of the total income is obtained by around 20 percent of the population [or households, in this example].  If each individual had the same income, or total equality, the income distribution curve would be the straight line in the graph – the line of total equality.  The Gini coefficient is calculated as the area A [between the curve and the line of equality] divided by the sum of areas A and B [everything under the line of equality].  If income is distributed completely equally, then the Lorenz curve and the line of total equality are merged and the Gini coefficient is zero.  If one individual receives all the income, the Lorenz curve would pass through the points (0,0), (100,0) and (100,100), and the surfaces A and B would be similar, leading to a value of one for the Gini-coefficient.”  From the Poverty Reduction and Equity website of the World Bank,,contentMDK:20238991~menuPK:492138~pagePK:148956~piPK:216618~theSitePK:430367,00.html

In a nutshell, then, the Lorenz Curve is established by plotting the cumulative percentage of income earned by the cumulative percentage of the population along two axes.  The Gini coefficient is then derived by creating a ratio of the area between the Lorenz Curve and the line of equality, divided by the total area under the 45 degree angle line that represents complete equality.  The lower the Gini coefficient, the more equality of income.

Because of the disparate ways that different countries calculate poverty and collect information on income, the Gini Index takes into account either or both “income” as earnings, and also by estimating “consumption,” or actual expenditures, as a proxy for income.  You will often see both used within the same map, depending on how the individual countries collected the data. 

Asia – relatively low Gini coefficient (0.38) in urban areas. 
From: UN-Habitat “State of the World's Cities 2010/2011 - Cities for All: Bridging the Urban Divide”

So, what are some of the impacts of income inequality in urban areas?  In addition to indicating simply economic disparities, a high Gini index represents barriers to cultural, political, social, environmental, and health equality.  These forms of exclusion continue to marginalize huge swaths of the population.  This marginalization is social as well as physical and translates into spatial inequalities and deeply divided cities.  Since virtually all of the world’s population growth in the coming decades will take place in cities, it is crucial to make cities more inclusive.  In general, countries with the highest per capita income are the most urbanized, and conversely, those with the lowest incomes tend to be the least urbanized.  So, overall, the process of urbanization should be seen as (and has been) overall an effective means of poverty reduction, and urban living has the potential to (and usually does) offer much more opportunity to people. 
Some areas have made progress in reducing the number of people living in slums.  In Northern Africa, for instance, both the proportion and absolute numbers of slum dwellers fell significantly, as it has in China as well.  What constitutes a “slum”?  According to the UN, a slum is defined as a part of the city where the residents do not have durable (permanent) housing, little or no acceptable water or sanitary infrastructure, severe overcrowding (more than 3 people to a room), and no secure tenure and protection against forced eviction.  By this definition, many cities are “slum cities,” with the majority of their residents living in slum conditions, which in Bangladesh includes 70 percent of urban dwellers; in Yemen, 65 percent; in Haiti, 76 percent; and in Bolivia, 61 percent.  Women bear the brunt of slum living, being more often the water bearers and waste disposers, spending more time within the slum than men, and thus being more exposed to the slum’s environmental burdens as well as the higher rates of crime and violence.
            This spatial inequality, the segmentation and marginalization of people and geography, the uneven distribution of amenities and disamenities, and fragmentation of opportunities, generate high social and economic costs for the poor as well as society as a whole.  Unequal income distribution can lead to social unrest and political instability, more than overall high poverty levels alone would produce.  The spatial “poverty trap” of income inequality stems from the physical and social distance between rich and poor neighborhoods, with concomitant adverse impacts: restrictions for the poor on job opportunities, gender disparities, deteriorated living conditions, social exclusion, and high crime rates.  The poor can not participate to the fullest degree in the “urban advantage.”  Considering that by 2050, nearly three quarters of the world’s population will be urban, and in the extreme example of South America, it is estimated that over 90% of its population will be urban by 2050, the fact that many of the world’s cities have such extremes of wealth and poverty does not bode well for our collective welfare. 
            There are a number of drawbacks to using the Gini coefficient as a measure of inequality.  As with any global index, it is an imperfect and imprecise measure for a complicated concept, and many factors are overlooked in its calculation, such as non-monetary economies, bartering, subsistence farming, welfare benefits from the state, governmental subsidies and services, and it measures income rather than wealth.  Income disparities also do not necessarily capture standard of living differences.  For instance, In Brasilia, there is a high Gini coefficient of 0.60, but nearly 95% of the population has piped in water and sewage provisions, indicating a higher quality of life for most of the population than might be assumed from a high Gini coefficient.  Most importantly, the Gini coefficient can not be used to measure poverty, per se, but merely income inequality.  Countries can be overall wealthy or overall poor, and therefore each have very low Gini coefficients, because they are homogeneous in terms of income.  That doesn’t mean, of course, that the two countries have similar standards of living, or poverty rates.  It just means that everyone in that country or city is approximately equally poor or equally wealthy.  The Gini coefficient is perhaps most useful when used in conjunction with other indices, such as the Human Development Index, and the GDP (Gross Domestic Product) of a nation.  Unfortunately, these are calculated at the national level, and can not usually be disaggregated to the municipal level to be compared with the urban Gini coefficients. 
Urban Gini Coefficient in Mexico, the Caribbean, and Central and South Americas,
From: UN-Habitat “State of the World's Cities 2010/2011 - Cities for All: Bridging the Urban Divide”

See these weblinks for further information on poverty mapping and urban income inequalities:

 UN-Habitat “State of the World's Cities 2010/2011 - Cities for All: Bridging the Urban Divide”
(Free download of pdf book)

“More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions,” 2007, World Bank.  (Free download of pdf book)

“Where are the Poor? Experiences with Development and Use of Poverty Maps,” 2002, World Resources Institute (Free download of pdf book.)


  1. On the last map the colors for the countries (chloropleth, right?) and for the cities (little filled dot symbols) don't match.

    Or rather, they do match, but with different meanings. Bothers me.


  2. It is too bad the maps did not predict a high conflict rating for north Africa.

  3. Jonathan - I think what is troubling you can be explained by the following - the over all country color is indicating the AVERAGE urban Gini index for that country, while the the color of the city points and squares are indicating the Gini score for that particular city. At least that's the way I am reading the maps. So they are actually showing different things.

  4. Scott - the Gini coefficent maps (income disparity) actually DO show a very large income disparity for Egypt, which of course partially explains the problems there, but most of the rest of the countries in North Africa/Middle East where there is currenlty unrest have "no data." As far as the peace index maps, the student map DOES show likely conflict in many of the affected countries in North Africa/Middle East, as does the Richard Florida analysis and map. It's really only the GPI (Global Peace Index) map that does not, and in fact, as pointed out in the posting, the GPI map has Egypt and some of the other affected countries shown in one of the most peaceful categories.

  5. So the orange countries (this is the South America map) show a country Gini of .40 - .49, but the dots of the same color show a city Gini index of .60 - .69. Same color, different meanings, one map.

    They could have lined up the colors, or they could have used non-overlapping color gradations.


  6. I agree that whenever you are showing multiple variables on one map, there is bound to be a potential for confusion/lack of clarity. However, the country color part of the map is showing the country's urban average Gini (NOT the national Gini), and as far as using the same color for the countries and the cities (having them represent the same Gini value) it would be difficult to see the points (city locations) against the country colors if the colors were the same. Your suggestions to use altogether different colors - one set for country's urban Gini and a different set of colors for the city-specific Gini - would seem to be one way around this problem.