Rising economic inequality in the United States is closely tied to the high concentration of capital asset ownership, especially corporate stocks and real estate, and to increases in the price of those assets in recent decades. These in turn are closely related to the structural transformation of the economy over 50 years, especially the decline of unionized manufacturing in the Midwest and the rise of finance on the East coast and of the technology sector – mainly information and aerospace – in the West.
At the national level, this process has had two main effects on American political life. One is the rise of oligarchs and their designated agents, especially in the Democratic Party, initially in the Clinton era, to the point today that billionaires openly contest the nomination for the Presidency. Oligarchs have long dominated the Republican Party, so now American politics has become, to a large degree, a contest between billionaires of different stripes, mediated by other billionaires in control of the major media, both traditional and social. This much is obvious to any observer.
Much less obvious has been the effect of the new pattern of inequalities on the outcome of American presidential elections. The peculiar institutional context of those elections is that they are indirect, conducted through an Electoral College apportioned roughly by population and elected state-by-state, mostly on a winner-takes-all basis. While rising national inequality has had no clear-cut effect on the division of the popular vote between the two major parties, we have shown in a new paper that, in close elections since 1992, rising inequalities within American states have become a decisive factor in determining the state-by-state outcomes, and thus the Electoral College and the presidency.1
The logic of this development lies in the peculiar economic alignments of the two main American parties. The Democrats were once an alliance of multi-racial Northern labor with Southern whites in the era of Jim Crow. They have become a coalition of well-to-do city-dwellers, mostly professionals and office workers, and low-income minorities, both Black and Hispanic. The party thus broadly dominates the two tails of the income distribution, high and low. The Republicans, though always in thrall to the very wealthiest, draw their voting base from suburbs, smaller cities, towns and rural areas, overwhelmingly white, and broadly in the middle of the national income distribution.
Our approach to this issue draws on the techniques developed for measuring inequality within countries, using sectoral and industrial data on payrolls and employment, and applied over more than twenty years in the work of the University of Texas Inequality Project. Adapting these techniques to the data-rich environment of the United States permitted us to develop good estimates of changing inequality within states on an annual basis, going back to 1969, and up to 2014 so far. Previously, measures of inequality within states had only been available for years before 2000 on a decennial basis, as many states are too small for the regular Current Population Survey to provide reliable inequality estimates. We were thus able to assess the relationship between changing economic inequalities after 1969 in each state and their Electoral College outcomes for all of the relatively close elections of this century, specifically 2000, 2004, 2012 and 2016.
Up through the 1980s, inequality within American states was generally greatest in the Deep South, reflecting the racial divide, economic underdevelopment and the legacy of plantation slavery. In more recent years, the locus of rising inequality has shifted to the North and West. California, a major example, was once mainly white and suburban, reliably raising up Republicans from Nixon to Reagan. Today it is a checkerboard of tech wealth, lower-income Hispanic Americans and immigrants, solidly Democratic.
Our annual measures of inequality in each American state show that the largest increases from 1989 to 2014 occurred in California, New York, Connecticut, New Jersey, Maryland, Nevada, Rhode Island, Massachusetts, Hawaii, New Hampshire, Washington, Illinois and the District of Columbia. All of these states voted for Hillary Clinton in 2016. And of the twenty states with the smallest increases in inequality, all but two (New Mexico and Minnesota) voted for Donald Trump, and in the case of Minnesota, Mrs. Clinton’s margin was a bare 1.2%.
This clear relationship foretells ongoing developments in American politics. In the Upper Midwest, the states that are credited with electing Trump in 2016 – Michigan, Pennsylvania and Wisconsin – are drifting away from their longstanding Democratic allegiance, as their cities decay and their working class and minority populations age and shrink. These states were very close in 2016 and could be recovered in 2020 with a small shift in overall public opinion, but things will become increasingly challenging for Democratic contenders in the years to come. Per contra, across the South and Southwest, and especially in Arizona, Texas and Georgia, the cities and non-white populations are growing relative to the suburbs and the countryside. Arizona could tip to the Democrats (as California and Nevada already have) as early as 2020; Texas and Georgia are further off, and subject to statewide campaigns of voter suppression aimed at discouraging minority voters and prolonging Republican control. But the demographics are inexorable and those obstacles will fall over time.
The current dilemma for the Democrats is that the Roosevelt era is long over, and the Clinton coalition is not sufficient any longer, having been worn down by deindustrialization and deunionization – but the Southern transition is not yet mature. So the Democrats in 2020 face a choice between an appeal aimed at retrieving the upper Midwest, and working to accelerate the dawn of a Democratic South. Each strategy has particular policy characteristics – especially with respect to trade, infrastructure and climate change – that may cause problems in the other region. And there is no guarantee that policies and campaign appeals chosen for 2020 – and possibly implemented in the event of a Democratic victory – will still be appropriate for 2024 and beyond.
It is possible, of course, that the 2020 election will be resolved on other issues, perhaps the dangerous questions of war and peace, perhaps the public’s divided view of the idiosyncratic incumbent, Donald Trump. It is also possible – though we believe very unlikely – that an economic crisis or a recession could supervene and drive the result. But in the event that the world survives the rocky opening of the present election year, and the American economy continues on its long-standing slow-but-steady course, the most likely thing would be that the battle lines of 2016 will form anew, and the election will be fought over similar territory and similar issues. In that case, we expect the result to conform to the pattern of recent years, with the South somewhat more contested by Democrats than in the past and the Midwest somewhat more difficult for them. As in 2016, a Democratic advantage in the popular vote could again prove worthless, for under the American system presidential elections are fought and decided in the contestable states – and these are neither the most egalitarian nor the most unequal.
