The close margin in Kentucky’s May 18 Democratic primary underscores some of Hillary Clinton’s potential vulnerabilities in the general election (which we raised earlier, after the Indiana Primary).
Overall, the race in Kentucky was amazingly close, especially considering how well Clinton performed in Lexington, home of the University of Kentucky. Previous contests suggested that Sanders would have an obvious advantage there.
Unsurprisingly, Sanders outperformed Clinton in the coal counties, but also did well in western areas of the state less dependent on coal. As with Indiana, these results expose a real weakness for Clinton with rural and less educated white voters.
The wider impact of this effect could hurt Clinton in Ohio, Pennsylvania, and possibly Michigan and Minnesota during the general election.
A Closer Look at Ohio
Using the Census Bureau’s 2014 Current Population Survey, the estimated shares of Ohio’s Citizen Voting Age Population (CVAP) by race and Hispanic origin break down as follows:
Race and Hispanic Origin | CVAP Share |
---|---|
White (non-Hispanic) | 84% |
African American | 12% |
Hispanic | 2% |
Asian | 1% |
Other | 1% |
In the table below, we include estimates of white and black vote share from the 2012 election, when President Obama carried Ohio by 3 points.
Source | Vote Share, White | Obama Support, White | Vote Share, African American | Obama Support, African American |
---|---|---|---|---|
Exit polls | 79% | 41% | 15% | 96% |
Census Bureau | 83% | – | 13% | – |
According to the 2012 exit polls, Obama won Ohio by carrying 96 percent of the African American vote, and 41 percent of the white vote, giving him a victory margin of 166,272 votes.
Scenario 1: Trump Plurality
How the election in Ohio will play out is far from certain, but using Obama’s performance as an indicator, along with other available data, we can forecast several scenarios that show an uphill climb for the Clinton campaign in the state.
Based on the following turnout assumptions, the race in Ohio could be even closer than 2012:
- An estimated 5.527 million major party votes are cast in the 2016 General Election (a 38,000-vote increase over the 2012 turnout).
- Eligible white and African American voters both turn out at 63.1 percent), which admittedly reflects a drop in African American turnout that we attribute to Obama’s absence from the ballot.
- Hispanic citizens turn out at 56.9 percent (higher than in 2012).
- Individuals in the remaining non-white groups (a very small part of Ohio’s electorate) turn out at 50 percent.
Trump performance assumptions:
- Trump carries 57 percent of the white vote, matching Romney’s share.
- Trump receives 10 percent of the African American vote, a 7-point increase over Romney (but slightly less than George W. Bush’s 2004 support).
- Trump receives 25 percent of the Hispanic vote, a significant drop from Romney’s 42 percent support among the group.
- Trump receives 40 percent of all other votes, a portion of the total vote that is not statistically significant.
Based on these projections, the vote would breakdown like this:
Race | Major Party Vote | Trump | Clinton | Trump Margin |
---|---|---|---|---|
White | 4,679,000 | 2,667,030 | 2,011,970 | +655,060 |
African American | 665,000 | 66,500 | 598,500 | -532,000 |
Hispanic | 115,000 | 28,750 | 86,250 | -57,500 |
Other | 68,000 | 27,200 | 40,800 | -13,600 |
Total | 5,527,000 | 2,789,480 | 2,737,520 | +51,960 |
For this scenario to materialize, Trump would need to capture 10 percent of the African American vote, which would be slightly less than George W. Bush received in 2004. As noted above, in this scenario, we are assuming both a drop in African American turnout and a slight defection of African American voters.
Scenario 2: Clinton Matches Obama’s African American Support
Some observers will be reluctant to accept that Donald Trump could carry 10 percent of the African American vote, given Obama’s vast advantage over the last two presidential cycles. So, keeping other factors the same, here we consider what happens if Clinton matches Obama’s 96 percent African American support.
- Donald Trump carries 57 percent of the white vote, matching Romney’s share.
- Hillary Clinton maintains the same level of African American support as President Obama, and Donald Trump receives 4 percent of the African American vote.
- Trump receives 25 percent of the Hispanic vote, a significant drop from Romney’s 42 percent support among the group.
- Donald Trump receives 40 percent of all other voters, a portion of the total vote that is not statistically significant.
Race | Major Party Vote | Trump | Clinton | Trump Margin |
---|---|---|---|---|
White | 4,679,000 | 2,667,030 | 2,011,970 | +655,060 |
African American | 665,000 | 26,600 | 638,400 | -611,800 |
Hispanic | 115,000 | 28,750 | 86,250 | -57,500 |
Other | 68,000 | 27,200 | 40,800 | -13,600 |
Total | 5,527,000 | 2,749,580 | 2,777,420 | -27,840 |
This scenario changes the outcome considerably. If Clinton maintains Obama’s advantage with African American voters, she will win the state. However this is still tenuous, as a Trump increase of 2 percent among African American voters would result in a virtual tie (in this scenario).
Scenario 3: African American Turnout Increases, Trump Receives 10% of African American Vote
Trump’s divisive rhetoric may actually increase overall African American turnout (as voters turn out in higher numbers to oppose him). If we assume, again, that the vote share is divided more traditionally (as in Scenario 1), we see that an elevated African American turnout may not be enough to deliver the state for Clinton.
- Trump carries 57 percent of the white vote, matching Romney’s share.
- African American turnout increases to 700,000 (inflating overall turnout), and Trump receives 10 percent of the African American vote.
- Trump receives 25 percent of the Hispanic vote, a significant drop from Romney’s 42 percent support among the group.
- Trump receives 40 percent of all other voters, a portion of the total vote that is not statistically significant.
Race | Major Party Vote | Trump | Clinton | Trump Margin |
---|---|---|---|---|
White | 4,679,000 | 2,667,030 | 2,011,970 | +655,060 |
African American | 700,000 | 70,000 | 630,000 | -560,000 |
Hispanic | 115,000 | 28,750 | 86,250 | -57,500 |
Other | 68,000 | 27,200 | 40,800 | -13,600 |
Total | 5,562,000 | 2,792,980 | 2,769,020 | +23,960 |
Scenario 4: Clinton’s White Vote Share Drops
If the Democratic primary has shown one thing, it’s that Hillary Clinton has a problem engendering support with working-class white voters. So now assume that Clinton’s support among white voters drops to 39 percent, but she maintains Obama’s advantage among African Americans.
- White support for Hillary Clinton drops below 40 percent, and Donald Trump carries 61 percent of the white vote.
- African American support for Clinton matches the 96 percent earned by President Obama.
- Trump receives 25 percent of the Hispanic vote, a significant drop from Romney’s 42 percent support among the group.
- Donald Trump receives 40 percent of all other voters, a portion of the total vote that is not statistically significant.
Race | Major Party Vote | Trump | Clinton | Trump Margin |
---|---|---|---|---|
White | 4,679,000 | 2,854,190 | 1,824,810 | +1,029,380 |
African American | 665,000 | 26,600 | 638,400 | -611,800 |
Hispanic | 115,000 | 28,750 | 86,250 | -57,500 |
Other | 68,000 | 27,200 | 40,800 | -13,600 |
Total | 5,527,000 | 2,936,740 | 2,590,260 | +346,480 |
It is clear in this scenario that such a marked drop in share of the white vote in Ohio would be devastating for Clinton. Elevated turnout of Hispanic voters—something proponents often cite—would be unable to overcome that margin.
All in all, it could be a challenging battle for Clinton in Ohio.
Implications for the Election Nationwide
We are not suggesting that Trump has the advantage when analyzing the electoral map. Rather, we are confirming that his path to victory runs through states like Ohio and Pennsylvania.
However, if the demographic advantage favoring Democrats were to deliver Colorado and Nevada for Clinton—which it certainly could—Trump would need to negate that outcome with a victory in Michigan and possibly Minnesota. Other states, like Virginia, that went Democratic in 2008 and 2012, could be potentially devastating for Trump if they did so again. From our perspective, Virginia in particular may be more likely to go Democratic than Ohio.
All of this leaves out Florida, which again presents Trump with serious demographic obstacles, but remains winnable for him. His path to victory is extremely complicated and unlikely should he fail to carry Florida.