With national conversations about inequality and fairness in the air, I’ve been thinking about what economic justice might look like to regions. I find the late John Rawls to be the most insightful philosopher on the subject of justice, so I’ve been re-reading his great works.

First of all, Rawls argues a just society must meet a minimum standard for civil liberties--basically, those specified by the U.S. Bill of Rights. Next, access to what he calls primary goods--things we value like influence, security, income, respect--should be openly available to all. That is to say, people with equal motivation and innate talent should have just as likely a chance at attaining a position of power--like a CEO or President--regardless of the race, occupation, or level of education of their parents. In so far, as existing inequalities remain in the distribution of primary goods--because of differences in luck, motivation, and innate abilities, for example--those goods should be arranged to maximize the welfare of the worst off.

Looking across countries, it is fairly easy to gain at least a rough sense of where nations rank on Rawls’ characteristics of a just society. It’s more challenging to come up with similar measures within countries, but there are ways to evaluate Rawlsian criteria as they might apply to metros. While basic civil liberties have been formally equalized across the United States since the Civil Rights movement, and, in principle, all laws must conform to the U.S. Constitution, more subtle aspects of the democratic process are likely to vary across regions. For instance, consider what the World Bank calls “voice and accountability,” which they define as “the extent to which a country's citizens are able to participate in selecting their government.” This might include measures of voter turnout, corruption, volunteering, and the number of associations that provide public goods.

Another important measure of opportunity has to deal with access to high quality schools. According to data compiled by John Logan, blacks living in San Jose, for instance, attend much better performing public schools than blacks living in Philadelphia. As a result and for other reasons, equality of opportunity is likely to vary considerably across regions.

To evaluate different institutional and social arrangements, Rawls proposes comparing one scheme to another based on the welfare of the least off. One can do this at the metropolitan scale, using purchasing power as a proxy for welfare. For these purposes, I identified the bottom quintile of family income earners as the least advantaged.

To measure purchasing-power (as opposed to income, which is correlated with the region cost-of-living), I created a housing inflation index to adjust incomes to reflect their regional value. I focused on housing instead of all goods because housing accounts for the largest share of consumer spending, must be consumed where it is purchased, varies the most across metropolitan areas, and is the only significant good for which metropolitan price data and quality characteristics are freely available. I used individual responses from the 2010 Census data (from IPUMS) to predict housing costs, based on a number of observable characteristics: the number of bedrooms, whether the unit is owned or rented, proximity to public transit, whether the unit is single detached, attached, or multi-family, commute time, and the decade of construction. Dividing actual housing costs by expected housing costs--based on how the average American values these characteristics--yields a measure of housing inflation, which can then be used to adjust incomes accordingly.

By this method, San Jose has the highest housing costs. They are 75 percent higher than the same hypothetical unit in the average large metro and roughly 144 percent higher than a comparable unit in a low cost metro like Pittsburgh. Focusing only on the 100 largest metropolitan areas, San Francisco is next, followed by Bridgeport, Conn., Oxnard, Calif., Los Angeles, San Diego, Washington, D.C., Honolulu, New York City, and Boston. For just 14 of the largest 100 metros, real incomes are lower as a result of housing prices. Said otherwise, almost everywhere in the country is cheap compared to those 14 large cities and their suburbs.

As for the welfare of the worst off, the poorest 20th percentile of residents in Ogden UT have higher real incomes than their counterparts in any other large metropolitan area in the United States. Provo, Utah and Washington D.C. are next, followed by Madison and Harrisburg.

On the opposite end, Fresno, Miami, Los Angeles, and San Diego make up some of the metropolitan areas with the lowest incomes for the poorest quintile. Other than New York and New Orleans, these metros are disproportionately in California and Florida. All of these metros have high income inequality, with New York standing out as having the largest gap between the top and bottom quintile (see table here).

Of the metros with relatively high incomes for the bottom quintile, Washington D.C. is the only one with high income inequality--measured by the income ratio of the top quintile to the bottom quintile. In other words, there is a strong negative relationship between inequality and the incomes of the worst off.

Of course, this is just one measure of welfare, and it ignores income mobility, equality of opportunity, and the justice of state and local laws. It also might fail to capture what we mean by the least advantaged. Many of the residents in the poorest quintile--an admittedly broad category--in metros like Washington D.C. and Provo are highly educated compared to similarly situated residents in Fresno and Miami, or differ along other characteristics that are associated with advantage in the United States--like age, race, and immigration status. Still, all things being equal, if you were born to one of the poorest families in Ogden or Washington D.C., you’d be likely to have roughly twice the purchasing power than if Miami or Fresno had been your origin.