The US presidential elections are fast approaching. Most of us are familiar with the candidates– former Secretary of State Hillary Rodham Clinton (D), and Orange Meatball real estate mogul Donald J. Trump (R). Though Trump’s odds are sinking, this is still a tight race: current polls show the bloated swamp cheeto businessman trailing Clinton by only ~6% nationally.
As the world awaits the results of November’s hunger games presidential election, Clinton and Trump have begun focusing their efforts on elusive swing states, also known as battleground states. These are places where there is, historically, no guarantee of support for one party or candidate over the other. The roster of swing states changes every election, but some states such as Florida, Michigan, Pennsylvania, and Ohio are more likely than most to have a determining role in a presidential election.
Ohio is particularly crucial– the Buckeye state has picked the candidate accurately all but twice since 1896 (28 out of 30 times!), the most recent failure was 55 years ago, in the 1960 election. A study by Smart Politics analyzed 1,400+ data points spanning the last 29 presidential elections. The data showed that since 1900, “Ohio’s vote for the winning presidential candidate has deviated a nation low average of 2.2 points from the national vote, an average of just 1.3 points since 1964, and 1.2 points since 1980.”
Most analysts would warn against conflating accuracy with predictability. Yes, Ohio is accurate predictions are historically predictable (say that 10 times fast!), but there’s no way of predicting which party the state will back. In the last 10 elections, Ohio voted blue 5 times and red 5 times. You can never predict which way the political wind will blow.
…Or can you? Recent research suggests battleground states may be more predictable than previously thought, arguing that personality traits have a decisive role in political preferences. People who identify as more conservative in their political affiliations are likely to be more industrious and polite, whereas those who identify as liberal appear to be sympathetic and egalitarian. It has been hypothesized that the underlying causes for these differences are rooted in people’s ability (or inability) to deal with ambiguity and threat– Conservatism is associated largely with uncertainty, avoidance, and resistance to change, while Liberalism poses the exact opposite trend. Fascinatingly, scientists have even observed differences in the brain activity of democrats and republicans in regions of the brain related to risk-seeking.
Observing the numerous associations between personality traits and political affiliation, (and fielding an influx of news stories on the pressing influence & deciding role of swing states in the presidential election), we wanted to contribute to the usual polls and predictions by analyzing our Good&Co data.
Using data from 9,132 app users in 29 US states, we compared the personality traits of people residing in 11 Clinton-polling states with those of users in 11 Trump-polling states. We compared the average scores of specific personality traits within the two groups and looked at four different areas: openness, politeness, empathy, and ambition.
Our data reflects relatively similar previous research findings. We found that app users from Trump-polling states (including Texas, Alabama, and Missouri) scored higher on measures of politeness when compared to people from Clinton-polling states (including Wisconsin, Massachusetts, and Illinois). People in Trump states were likely to possess more conservative sentiments such as believing in the rule of law and being more resistant to change. On average, app users from Trump-polling states were 5% more polite than those from Hillary-polling states. It is noteworthy that “politeness” here is related to notions of order and traditionalism.
The data also revealed that users in Hillary-polling states scored significantly higher on a number of personality traits when compared to users from Trump-polling states. Those residing in Clinton-supporting states were likely to score more than 5% higher on openness, open-mindedness, and adventurousness. As shown in previous research, people who are less open to new experiences are likely to hold more conservative ideals on matters of authority and sanctity. The same outcome was observed in relation to empathy and compassion to others, previously shown to be associated with liberalism; app users from Hillary-polling states were on average 2% more empathetic than those in the Trump-polling states. Though we do not know our app users’ political affiliations, it seems that there are systematic differences across states between the average scores of key personality traits associated with liberalism and conservatism.
Some of our other findings were not quite as in line with previous research. We found that app users from more liberal states (Hillary-polling) were likely to be more ambitious. On the whole, people from states such as California, Washington, and New York were likely to be more determined– be it by seeking challenges or by believing in their own competencies and abilities. Compared to app users from Trump-affiliated states (such as Oklahoma, Kentucky, and Utah), people in Hillary-polling states scored, on average, 3% higher on measures of ambition.
In order to try and determine the undecided states, we compared personality traits of app users hailing from strong Hillary and Trump polling states with those of users from five differing swing states. The battleground states we focused on were Nevada, Michigan, North Carolina, and the infamous Florida and Ohio. From our analyses, we found that user scores for all four personality dimensions did not significantly differ across the five swing states.
App users from Nevada did not differ significantly from users in Trump and Clinton-polling states, which begs the question: is Nevada a true swing state? People in Nevada seem to fall somewhere in the middle between conservative republican states and the more liberal democratic states. This is consistent with current polling numbers. Politico has the average at 44% Clinton, 42% Trump.
App users from the battleground state of Michigan differed significantly on openness when compared to app users from Clinton-polling states. Michigan users were almost 7% less open-minded and adventurous than users from Hillary-polling states. Though polls in the Great Lake state point to a much larger support gap than Nevada, the personality rift seems not to be in Trump’s favor after all– Clinton is polling at 42.6% to Trump’s measly 33.8%.
Significant differences were also found between the personality traits of app users from North Carolina and those in Hillary-polling states. North Carolinians scored over 7% higher on politeness compared to users in Hillary-polling states– an even larger margin than the one observed between Trump and Clinton-polling states for the same trait. This may imply that, on average, people in North Carolina show more conservative or traditional attitudes, which perhaps fits in with the so-called “Religious Freedom” bill recently introduced in the state. Despite its conservative leanings, North Carolina is still very much a battleground. Clinton leads trump by just 4 points (44.4%-40.4%), with some polls tallying a virtual tie.
App users from Florida differed significantly from app users in Clinton-polling states in three out of the four traits, but did not differ at all from Trump-polling states. These app users were on average 7.5% less open to new experiences, 5% more likely to be traditional in their views, and 4.5% less ambitious than those residing in Clinton-polling states, whereas no significant differences were observed between app users from Florida and those from Trump-polling states. People over 60 make up 23% of Florida’s population, which may be a root explanation for these differences from other states with a less prominent elderly population. Like North Carolina, however, Florida’s traditional leanings are no Trump guarantee– Clinton is edging ahead by 5.2%. The battle continues.
App users from Ohio showed the most extreme differences. Ohioans were likely to score lower on openness compared to both Trump-polling and Clinton-polling states; on average 8% less open-minded than app users from Trump-polling states and 13.8% less than those from Hillary-polling states. Similarly to Florida, app users from Ohio scored higher on politeness, scoring on average 5% higher on traditional ideals compared to app users from Hillary-polling states. However, Ohio is not a very diverse state, with 80% of the population being caucasian. This may explain the lack of open-mindedness, as well as the penchant for more traditional ideals. In true Buckeye battleground fashion, the state is still distinctly undecided, with Trump behind Clinton by just 3.2 points.
At this time, these states are still undecided. Looking at the personality traits of app users in battleground locations, we might speculate that North Carolina, Florida, and Ohio will lean towards the Trump camp, as they are more traditional and conservative in their views when compared to Hillary-polling states. However, as some of the polls already show, it should be noted that this is a very crude method of trying to predict which way a battleground state may swing; our data does not take into consideration socioeconomic factors, gender, ethnic background, religiosity, or any of the external characteristics which are likely to affect one’s vote. Nevertheless, our data did show meaningful differences in the way personality was manifested in two groups of potential voters differentiated only by geographic position. This suggests that there may be some systematic differences in personality –and thus behavior– across different US states, which may translate into political affiliations.