New State-Level Price Data Shows Smaller State Real Income Differences
May 23, 2014
After my series on interstate migration ended last week, I received a number of questions from people curious about a few points I made. One in particular was noteworthy: if people migrate to improve their incomes, why do people seem to leave high-income states and move to many lower-income states?
It’s an important question, and one of the curiosities of much migration-related research. Higher state income per capita is consistently associated with lower in-migration, rather than higher.
When I received these questions, my first response was to say the effect is an illusion: prices in DC and Mississippi are different, and those price differences eliminate much of the apparent income differences. Unfortunately, I didn’t have the data at the time to prove this point. From states like Mississippi, West Virginia, and Idaho with less than $32,000 in average disposable personal income (that is, after-tax income actually earned by real people that they can spend), to the District of Columbia with over $65,000 in average disposable income, a simplistic look at state incomes could make you think there are huge geographic income disparities and inequalities around our nation.
As a native of a poorer state (Kentucky), this narrative has always bothered me. Sure, incomes are higher here in DC, but so is the cost of living. From food to rent to taxes to gas, it’s just more expensive to live in some places than others. So real incomes, or what a person can actually buy with their money, may vary less than nominal incomes, or the dollar amount people earn. Unfortunately, standardized, official data on statewide price levels has not been available in the past.
But last month, the Bureau of Economic Analysis released a dataset producing state- and metropolitan-level relative price parities. This new data provides an answer to the question about migration and incomes. Prices are in fact far higher in some states than others, and states with higher incomes usually also have higher prices. Indeed, once state-level disposable incomes are adjusted for state price levels, 40 percent of the variation in state incomes vanishes. More importantly, the correlation between migration and incomes changes directions. Once you adjust for price differences, higher real incomes do indeed have a positive association with migration.
Here’s a concrete example of what this means: before adjusting for inflation and price differences, New York’s average disposable income is $45,463, the 8th highest in the nation. After the adjustment, it falls to $39,396, or 29th. Meanwhile, Louisiana’s pre-adjustment income is $36,806, the 30th highest in the nation. But after the adjustment? $40,269, or 21st. Louisiana goes from being almost $10,000 poorer than New York, on average, to having a “real income” almost $1,000 higher.
So what conclusions can we draw from the new price-adjusted income data? I thought of three:
- As I noted, once you adjust for price levels, the correlation between the migration replacement rate and disposable (after-tax) income changes from -0.4 to 0.25. Those are low correlations, but they do show the huge effect of price differences. Higher real incomes do indeed appeal to migrants.
- Measures of inequality based on unadjusted incomes may incorrectly measure nationwide inequality. If researchers don’t adjust for the fact that price levels are very different, then they will tend to overestimate the real impact of inequality in terms of peoples’ actual standard of living.
- Federal tax progressivity has strange consequences. People who are “poor” in one state could be “rich” in another without changing the dollar amount of their income. So the progressive nature of the federal income tax can lead to poor- or middle-class people in high-price states paying taxes equivalent to what significantly richer (in real, standard-of-living terms) people would pay in low-price states.
Whatever conclusions we may draw, the data is extremely useful, and fills what was previously a major hole in the tools available to economists working at the state level.
|Nominal and Real Disposable Income by State|
|State||Per Capita Disposable Personal Income||Real, State-Price Adjusted Income||Difference|
|District of Columbia||$65,770||$55,643||-$10,127||-15.4%|
Follow Lyman on Twitter.