The midwest is the most equal region of the United States. Can you help us explain why that is?
We’ve been trying to figure it out since David Wiczer, an economist at the Federal Reserve Bank of St. Louis, published his findings on the geographic spread of income inequality in the US.
Income inequality is rising everywhere in the US, but something else immediately stands out about this chart: There is a prominent and persistent gap between the midwest and the other regions, starting around 1993. Rich and poor midwesterners are measurably less different, in terms of income, than people in the rest of the country.
At first, we thought this might just be an artifact of how Wiczer chose to measure income inequality, as the gap between median and mean incomes. But we’ve found similar differences in other indicators.
Consider the gini index, a standard measure of income inequality that ranges from zero to one. Numbers closer to one represent increasing inequality. Though the US Census Bureau has only published the gini index by region for the last few years, that data reveal similar variations. The midwest has been consistently more equal since at least 2010.
A third way of looking at it comes from a 2015 analysis by the Economic Policy Institute. Their data show the midwest had the slowest income growth among the top 1% of earners over nearly three decades.
Because the definition can be contentious, we should define what is meant by “midwest.” It includes Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin.
These states are clearly unlike the rest of the country. The difference isn’t huge, averaging to about 5.4%, but it’s unusual to see such a consistent gap. It’s also unclear why the gap would have developed suddenly in the early 1990s.
There are a lot of reasons why inequality between different parts of America is growing. But those factors don’t necessarily explain inequality within any particular region of the country. We have probed some explanations and consulted economists, including Wiczer, to help explain this phenomenon. So far, we’re all stumped.
Experts at the Census Bureau warned that by only analyzing family incomes, we might be missing potential regional variation in living situations. Plenty of people live in households that aren’t formally families. However, charting the same period by household income, we see the same gap.
The gap also exists, though it is somewhat less pronounced, with personal incomes.
Cross that explanation off the list.
Economists such as Thomas Piketty have persuasively argued that top executive salaries are the primary driver of income inequality. Perhaps, because midwestern businesses are further from coastal economic centers, they pay less competitively.
It’s difficult to conclusively determine if executive incomes are a cause of the gap because there just isn’t that much state or regional data on high incomes. What we can say is that salaries for those identified by the Bureau of Labor Statistics (BLS) as chief executives are lower in midwestern states, probably by about 15%. (BLS data is not granular enough to allow us to make a perfect regional estimate.)
That difference would drive average incomes down. However, the same dataset also tells us that there are significantly more chief executives in the midwest, 2.28 per thousand jobs, compared to 1.56 per thousand jobs in the northeast. That should drive the average up.
Moreover, an analysis of CEO pay by the AFL-CIO found that midwestern states had the second-highest average pay. Unfortunately, because there is no complete source for executive pay statistics, we can’t rule out the possibility that this is only true for the CEOs of the largest companies.
One cultural assumption about the midwest is that it is home to more “mom and pop” small businesses. This is often accompanied by the belief that these businesses are more economically fair to their employees. But data from the BLS show that the midwest does not, in fact, have the most small businesses in the US.
And even if there were significantly more of these firms in the midwest, research has shown that small businesses tend to pay worse, not better, than large ones.
There is substantial evidence that unions can reduce income inequality. Illinois, Michigan, and other industrialized midwestern states have reputations for being bastions of union membership. It’s at least plausible that a higher concentration of union activity has contained inequality. However, available data do not support this conclusion.
As this chart shows, the midwest fell behind the northeast in union membership in the early 1970s. Since then, it has declined the most of any region. Today only 12.3% of midwesterners are estimated to be in union jobs.
If union activity were responsible for the lower income inequality in the midwest, that gap should have appeared well before 1993. There is also no clear reason why such a gap would have held steady over 25 years while union membership steadily declined. And finally, union participation has been highest in the northeast for more than 40 years, but that region consistently has the highest income inequality.
When there’s a change in the way data are measured, it’s a called a “break in series.” Generally, that means data before and after the break can’t be compared because they aren’t really the same. And there was just such a change in the way the Census Bureau’s Annual Social and Economic Supplement was surveyed in 1993.
Particularly relevant to this analysis, the Census Bureau increased the maximum annual income that could be reported, up to $999,999 per year. Census Bureau documentation for that year states, ”The effect of the revised upper limits on median incomes was negligible.” The increase in top incomes could still have moved measures of inequality, but even if it did, this would not explain why the midwest diverged from the other regions.
Census officials told Quartz there are no regional differences in how the survey is administered. It is possible that there were simply so many more people earning a million dollars a year outside the midwest that the change in methodology led to a spike in every other region. However, it’s important to keep in mind that Chicago, with plenty of high incomes, is in the midwest. The data can’t conclusively rule this theory out, but doesn’t clearly support it, either.
We may be stumped, but perhaps you have some ideas. What do you think explains this regional difference in income inequality? What changed in the early 1990s? Does one of our existing theories persuade you, or do you have others to suggest? Use the comments section below to offer your own ideas. Bonus points if you include a chart!
To get you started, we’re sharing all of the data that we analyzed as a series of Google $GOOGL spreadsheets. Have at it.
