Features : : Financial Insecurity and the Election of Donald Trump

Financial Insecurity and the Election of Donald Trump

By Diana Elliott and Emma Kalish

December 16, 2016

In an earlier version, we transposed the national voting results for Trump and Clinton. The feature has been revised to correct the error.

The outcome of the 2016 presidential election caught many by surprise, considering the projections in the lead-up to November 8. In the wake of the election, media outlets and pundits have made assertions about which voters drove the election and what motivated their choices.

One of the primary narratives to emerge has been that financial insecurity drove the election results. Was this election about the frustrations of working-class white voters and their increasingly precarious economic status? Or is the explanation grounded in voters’ demographic characteristics, including race and ethnicity, age, and educational attainment?

Data may shed some light on the answers. We looked at county-by-county election results and voters’ financial and demographic characteristics and found that financial insecurity—as measured by credit scores—did not drive voting preferences. The perception of financial insecurity, however, may have been quite important. Before the election, the most favorable ratings of Trump were held by those with the most anxiety about their finances, regardless of income or local economic conditions.

Our analysis considered not only income, but also other components of financial security, including families’ access to credit and their wealth-building potential. Financial security is also intertwined with families’ economic context, like job opportunities and access to homeownership. All these factors paint a more complex picture of Americans’ financial lives than has been portrayed in the postelection narrative.

Among the 55 counties (or county equivalents) with residents with the highest average credit scores (720 and above), Hillary Clinton won just four of them: Falls Church, Virginia (with an average credit score of 729); San Juan County, Washington (722); Cook County, Minnesota (721); and Washington County, Minnesota (720). High credit scores are associated with long, successful credit histories and bills paid on time and are implicit markers of financial security and stability over a lifetime.

High credit scores are also more often held by white consumers. The most diverse areas in this group are Falls Church, Virginia (73 percent of the population is white), and Morris County, New Jersey (72 percent). All but three (Douglas County, Colorado; Falls Church; and Morris County) have median incomes well under $100,000 a year.

Unemployment rates are under 5 percent in each of these areas, and homeownership rates range from 59 to 90 percent. Residents here are not necessarily high income, but they do work, and many are white, working-class people. They are financially secure homeowners who generally pay their bills on time and have access to reasonably priced credit.

Meanwhile, Clinton won all 11 counties that had the lowest average credit scores (below 600, in the subprime range). In 10 of those counties, Clinton won more than 60 percent of the votes. In Tensas Parish, Louisiana, which has an average credit score of 598, Clinton won with 52.3 percent of the votes. Tensas Parish is majority black, but also has the highest percentage of white residents (42 percent) in these 11 counties. The other counties with average subprime credit have high percentages of black or Native American residents, and the median household income is below $32,000 a year.  

Unemployment rates in these 11 counties range from 7.4 to 14.9 percent, and homeownership rates range from 44 to 75 percent. These residents are more likely to be financially insecure with diminished access to credit and fewer job prospects.

Residents in counties with the best credit scores tend to be white, higher-income, middle-class homeowners and overwhelmingly Donald Trump voters. Meanwhile, residents in counties with the lowest credit scores tend to be black or Native American, have low median household income and unstable employment, are less likely to own their homes, and overwhelmingly are Clinton voters.

This seems to indicate that financially secure voters were more likely to cast their ballot for Trump. But let’s take it a step further. We can focus on the effects of financial insecurity by holding counties’ other economic and demographic factors constant (median income, unemployment rate, percentage white residents, percentage with a bachelor’s degree or higher, and homeownership rate). Doing that shows that financial security is no longer a significant predictor of whether Trump won a particular county.

Why? Because higher credit scores are located in counties with higher percentages of white residents, and race was an important predictor in this election: a 10 percentage point increase in a county’s white residents is associated with an 8 percentage point increase in voting for Trump. Education was also important. A 10 percentage point increase in residents with a bachelor’s degree or higher is associated with a 20 percentage point increase in not voting for Trump.

These data suggest that, unlike the educational attainment and race of a county’s voters, financial insecurity was not a strong driving force in this election. Perhaps more important, the data suggest that underlying and persistent racial patterns of financial insecurity prevail in the United States—a fact that all politicians, regardless of party preferences, would be wise to address.