A Critique of the ACCRA “Average” Cost of Living

One of the more perplexing regional economic puzzle is that for many decades Americans appear to have been moving from high pay areas to low pay areas. This was first distressingly obvious when, after the Second World War, there was a “resettlement” of the South as people returned to Dixie, reversing the massive out-migration to the north during the Great Depression and the Second World War. The problem is that those return migrants were leaving the high paid industrialized northeast and moving to the lowest paid region of the nation, the Deep South.

A similar problem was obvious in the 1990s as people poured into the Mountain West despite the fact that 6 of those 8 states had below average pay and income and half of them, including Montana, were in the poorest ten. Meanwhile what appeared to be the richest states in the county were losing population to out-migration. That included New York, Connecticut, Massachusetts, New Jersey, and California.

It is possible that people are just plain stupid or economically irrational. Alternatively, there may be something wrong with the numbers we use to judge local economic well being. Given that these migration patterns have persisted for many decades, it seems unlikely that people are repeatedly making mistakes or are just terribly slow learners.

One obvious explanation for this economic puzzle is that there are geographic differences in the cost of living between New York City or Southern California and, say Missoula or Bozeman, and that migrants are taking that into account. This amounts to saying that there is in fact something seriously wrong with the pay and income figures we regularly use to compare ourselves with the rest of the nation. We do something we would never do if we were comparing pay levels in 1975 to pay levels today. We ignore the difference in the purchasing power of the dollars we are talking about.

Large, densely settled areas almost always have higher costs of living. As the population of cities grows, the demand for centrally or conveniently located land rises but the supply does not. So land costs rise, pushing housing costs and the costs of almost all economic activities that need to make use of land higher. No mystery there.

But data that is regularly released by private economic research firms that study geographic differences in the cost of living seem to indicate that the cost of living in places like Missoula, Kalispell, and Great Falls, are just about at the national average. The high cost city in the state, Bozeman, is only 9 percent above the national average and the low cost city, Billings, is only about 5 percent below the national average. The cost of living in Montana’s cities is reported to be clustered closely around the national average.

That, however, is simply due to a classic arithmetic mistake. In calculating the average cost of living, the fact that some urban areas have 20 million people, the New York-New Jersey metropolitan area, while other urban areas, like Kalispell, have as few as 14,000 residents was ignored. So the cost of living in Kalispell and that in the New York metro area were just added together as if the same number of people experienced those costs of living in each location. The fact that there were 1,500 times as many people in the New York metro area was simply ignored. There are many more small cities in the US than there are large cities. It takes 260 of our small metropolitan areas to make up half of the metro population but only 16 large metro areas to make up the other half.

Let’s round those numbers a bit. Suppose that half the population lived in 300 small cities the size of Missoula or Great Falls and had a cost of living of 100. Suppose that the other half lived in 15 large cities and had a cost of living twice as high, 200. (New York City’s cost of living was almost that high, 190.) If we just average across the 315 cities, the average cost of living would be about 105. Small cities like Montana’s would appear to have approximately average costs of living. But since the average cost of living is really the 200 faced by one half of the population and the 100 faced by the other half, the average cost of living is actually 150, not 105. The smaller cities actually have a cost of living that is 30 percent below the average.

It is very important to get these numbers straight. If, using actual data, we compare relatively small metropolitan areas with relatively large ones, for instance the 125 cities like Missoula, Great Falls, and Billings, with populations less than 250,000 with the 23 metro areas with more than 2 million people, we find that average pay in the larger cities was almost 50 percent or $11,000 per year higher: a startling pay gap around which one would expect considerable political discontent to be mobilized. If, however, we adjust for differences in the cost of living between those different sized cities, the gap declines to about 10 percent, a level that might not call for drastic public policy intervention.

A good deal of the political rhetoric about the state of the Montana economy is built around the assumed reality of just such a huge pay gap. It is reinforced by the same economic and arithmetic errors: not adjusting for cost of living and not calculating averages correctly. That would not seem to be a very firm platform from which to launch major economic policy initiatives.