8.4 Million Obama-Trump Voters
Obama Voters Abandoned Clinton
President Barack Obama won two elections with a robust and resilient electoral coalition that propelled him to easy wins throughout the Midwest. His coalition, resilient though for him, did not remain intact for Clinton. According to the American National Election Study, 13% of Trump voters cast a ballot for Obama in 2012. That amounts to around 8.4 million individuals. By comparison, of Clinton’s voters, only 4% voted for Mitt Romney in 2012 (totaling around 2.5 million people).
|Makeup of Clinton/Trump Voters by 2012 President Vote||Obama||Romney|
Extrapolating those numbers to individual states and adjusting for state swing from 2012 to 2016 yields the following Obama-Trump voter estimates and compares that to Trump’s margin of victory.
|Estimated Obama-Trump Voters||Trump’s Margin of Victory|
In each state, Obama-Trump voters more than account for Trump’s margin of victory (though, again, see the caveats). Even accepting flaws in these estimates, it’s readily apparent that a sizable number of Obama voters had to flee from the Democratic Party: How else would Iowa have swung 15 points; Ohio, 11; Michigan, 9.7; Wisconsin, 7.7; and Pennsylvania, 6.1?
These gains came predominately from the white working-class. A pre-election survey in Pennsylvania found that of Obama’s white working-class voters, some 18 percent planned to vote for Trump over Clinton. In Iowa, Obama won the white working-class by around 3 points in 2012 whereas Clinton lost it by 20 just four years later. Macomb County, Michigan, went to Trump by 11.5 points but Obama by 4. Trump took Erie County, Pennsylvania, by 2 points. Obama won it by 17.
The exact reasons Clinton failed to retain white working-class voters who supported Obama continue to be debated. Cultural and economic anxiety quickly come to mind, as does Donald Trump’s demagoguery, critical rhetoric aimed directly at this sprawling constituency. Regardless of why the white working-class abandoned the Democratic Party, this instance of the Obama coalition’s partial collapse spelled disaster for Hillary Clinton as the voters with which she aimed to replace them simply did not reside in swing states.
Coming soon: Part 2 – An Inefficient Electoral Coalition
 Such survey results do come with caveats: Respondents routinely misremember (or lie) about for whom they voted in the preceding election. In this case, 58% of ANES survey takes claimed to have voted for Obama in 2012. Obama only received 51% of the vote. Research posits that individuals tend to say they voted for a socially acceptable answer – in this case, that means saying they voted for Obama (who has a high approval rating) where in following 2004, more claimed to have voted for John Kerry than for George W. Bush.
 These numbers suffer from the same drawbacks explained above. Furthermore, these are just estimates and may well be off (this is also single-party crossover; the numbers don’t look at Romney-Clinton voters). I tried to account for partisan swing by treating the 13% of Trump’s voters who cast a ballot for Obama as a baseline adjusted upwards based on how much the state swung to the Republican Party in 2016. That means states with the largest GOP swing is estimated to be home to more Obama-Trump voters than that state’s share of the nation’s total voting population.