Income Inequality: The problem of cost

Critics of the US point to its higher level of income inequality arguing that this proves that America is an 'unfair' country.  Indeed using the Gini index - a measure of income dispersion, the US scores a taxes and transfers adjusted 39 versus a 31 for more egalitarian Sweden.

In an earlier post I explained why comparing income inequality in a smallish, homogeneous country like Sweden to the US is flawed and showed that by comparing individual US states to European countries a lot of the so called 'income inequality' gap vanishes.  Indeed on a more 'apples to apples' comparison of the entire EU versus the USA, the EU scores a 41 which is worse than the US.

But there's another factor that affects both the US and Europe and that tends to seriously overstate income inequality:  differences in the cost of living.  For example, here is the cost of living among the ten largest metro areas:

New York-Newark-Bridgeport, NY-NJ-CT-PA CSA	220
Los Angeles-Long Beach-Riverside, CA CSA	148
Chicago-Naperville-Michigan City, IL-IN-WI CSA	113
Washington-Baltimore-Northern Virginia, DC-MD-VA-WV CSA	137
Boston-Worcester-Manchester, MA-RI-NH CSA	134
San Jose-San Francisco-Oakland, CA CSA	172
Dallas-Fort Worth, TX CSA	92
Philadelphia-Camden-Vineland, PA-NJ-DE-MD CSA	124
Houston-Baytown-Huntsville, TX CSA	91
Atlanta-Sandy Springs-Gainesville, GA-AL CSA	97

So a mid-level executive in NYC who is moving to Houston (which are the number 1 and number 2 most common Fortune 500 headquarters cities) would find that his standard of living had more than doubled.  The difference is a little more pronounced when  you include all metro areas.  The cheapest metro area in the nation (Harlingen, TX) has an index of 80.

This means that a significant proportion of the difference in wages in the US may be due to differences in the cost of living.  For this to be true wages would have to track cost of living which is broadly but not completely true (for example, TX has a higher standard of living than NY or CA despite having lower average wages).  On the other hand, we don't have information on the cost of living of rural areas which is no doubt lower than that of most if not all metro areas.  So in general I would say that the differences in costs reflected above are representative of the difference in wages nationwide (and no I'm not up for doing the damned analysis -  do it yourself if you think it's so important).

So what?  Well, I calculated the Gini coefficient for cost of living among the 366 MSAs in the US (don't be impressed, it took about 1 minute and no technical skill).  In other words, I measured the dispersion of cost of living in the US and compared it to the index for dispersion of incomes.  Not surprisingly, it was lower, the index was 7 versus 39 for income inequality.

Another way to say this is that a little over 20% of the variation in after tax and transfer incomes in the US is accounted by the difference in costs of living between the various places that people live. This helps explain why some rich small countries typically have lower Ginis:  the costs of living vary far less within a Denmark or Sweden and therefore so do wages.

There is also evidence that this cost of living differential has been growing over time.  The biggest driver of cost differentials between cities is the cost of housing.  Cities like San Francisco and New York have seen explosive housing inflation over the past 30 years.  Cities like Dallas and Houston have seen very little increase in the cost of housing despite a population that was growing much faster.  This reflects very different philosophies of governance - a 'Blue' versus 'Red' model, if you will.

There's a lot more I would like to say about the difference in costs of living between cities and how it represents the Blue versus Red approaches to governance and inequality but I will save that for later.