Most research has focused on growing inequality within and across nations. What’s less clear is to what degree income inequality has grown across different parts of the U.S.
With the help of my Martin Prosperity Institute (MPI) colleagues Charlotta Mellander and Karen King, I examined the change in income inequality across the 50 states between 1979 and 2012. We used U.S. Census data to track the Gini coefficient, the standard measure of income inequality that measures the distribution of income across a population on a scale of 0 to 1. Across the nation as a whole, the Gini coefficient increased by roughly 15 percent (from .415 in 1979 to .476 in 2012).
The map below, by MPI’s Zara Matheson with support from the geospatial mapping team at ESRI, visualizes the change in income inequality over this three-decade period for each of the 50 states. (A larger map with even more functionality is available on the MPI website.)
In 1979, eleven states plus the District of Columbia had rates of inequality that exceeded that of the nation as whole. Most of these were less advantaged states in the South with large poverty populations—Mississippi, Louisiana, Arkansas, Alabama, Georgia, Kentucky, and Tennessee—as well as Oklahoma, New Mexico, Florida, and New York.
By 2012, the map as a whole has clearly gotten much darker, indicating increased inequality across the board. But now the states with the highest levels of inequality—ten of them, plus D.C.—are those with more advanced, knowledge-based economies. New York and Connecticut join D.C. at the very top of the list, while California and Massachusetts also number among the top ten most unequal states. Southern states that began with high levels of inequality remain near the top of the list, with Mississippi, Louisiana, Florida, Georgia, and Texas all among the top ten. By 2012, each and every state in the union had a level of inequality higher than the national average in 1979.
More than half of U.S. states—26 plus D.C.—saw inequality increase at a rate higher than the national average over this period. Connecticut led the way with a 26 percent increase, followed by Massachusetts, New Jersey, Ohio, and New York. With the exception of Ohio, these are all larger states with robust knowledge economies. And all but four states— South Dakota, Alaska, Arkansas, and Hawaii—saw inequality rise by ten percent or more
Inequality has risen consistently across all states. The level of inequality in 1979 was closely correlated with the level of inequality in 2012 (.74). In other words, states that started off with high levels of inequality remained highly unequal 30 years later, but inequality did not increase more in these states over time.
The map below shows the rate of growth in inequality by state over the period from 1979 to 2012. Inequality grew more in the Northeast, Midwest, and California.
What lies behind the growth in inequality across the states?
To get at this, Mellander ran a simple correlation analysis between state inequality at different points in time, as well as the change in income inequality across states, and a series of key economic, demographic, and social factors which would seem to bear on it, including income and poverty, race, education, population, and density. As usual, I point out that correlation does not equal causation and points only to associations among variables.
Perhaps surprisingly, income inequality does not appear to have risen more in states that had higher median incomes to begin with. There was a negative and significant correlation between inequality and median household income in 1979, but no significant relation between the two by 2012.
Some research has suggested that growth at the very top of the income distribution, accruing to the top 1 percent, may have played a significant role in overall inequality. A recent study from the Economic Policy Institute found that, between 1979 and 2007, the share of income going to the top 1 percent of population increased at a rate 10 times that of the income of the bottom 99 percent. The share of income going to the top 1 percent rose across each and every one of the 50 states.
One common explanation suggests that rising inequality is a consequence of an increasingly polarized labor market—the decline in once high-paying blue-collar jobs and the growing divide between low pay, low skill work and higher-paying knowledge work. Inequality, according to this view, is a product of so-called “skill-biased technical change,” which has conferred greater returns for higher skilled work that require higher levels of education.
But Mellander’s analysis finds that the share of the workforce that is high-skilled—which we defined as the percentage of the population with a college degree—appears to play little, if any, role in state inequality levels. For this reason, it may well be that policies that seek to upgrade skills and generate more college grads, as effective as they may be for improving skills and wages, are likely to have little effect at mitigating rising overall inequality.
What seems to matter more is the share of the population that is less skilled. Inequality is positively correlated with the share of adults who have not completed high school across all time periods in her analysis. That said, her analysis suggests that this variable has had little to do with the change in income inequality over the past three-plus decades. In other words, states with lots of unskilled people have had consistently higher levels of inequality over time, but this has not led to a larger rise in inequality over time. Still, this suggests that policies aimed at encouraging young people to stay in high school and at improving the high school graduation rate may prove more effective in helping to combat rising inequality at the state level.
But the biggest factor in the growth of inequality, according to Mellander’s analysis, is simply the size and density of states. The change in income inequality over the past four decades is strongly associated both with the initial size of the population (.52) and, even more so, with population density (.73). This is in line with other studies that have found inequality to be higher in larger cities and metro areas and identified a close connection between urbanization, productivity and inequality across U.S. metros.
Three decades ago, inequality was highest in the South, and it remains high there today. But inequality has grown fastest in larger states with more vibrant knowledge economies like California, New York, and Massachusetts. What this would seem to reflect is the hollowing out of these states’ economies over time and the loss of higher paying middle-class jobs. The job markets of these states have split into high-wage knowledge jobs and even faster growing low-wage jobs. The result is gaping inequality.
In fact, our results suggest that inequality may be a byproduct of the very factors—density and clustering— that drive innovation and economic growth in today’s economy. While clustering is a propulsive economic force, left to its own devices its benefits accrue mainly to the skilled and advantaged, while the remaining "66 percent" fall further and further behind.
The most effective strategy for reducing inequality would thus seem to lie in improving educational and economic opportunities and increasing the wages of those at the very bottom of the economic order. It will take a new social compact to upgrade the 60 million-plus low-skill, low-pay service jobs and add a social safety net. But it is necessary if we want to help create a new middle class and stop the runaway inequality that this entire country and each and every one of the 50 states now face.
Top Image: A demonstrator affiliated with the Occupy Wall Street carries a sign during a rally in New York's Times Square, Oct. 15, 2011 (AP Photo/Mary Altaffer).