Ballotpedia noted an interesting trend from the 2016 election among singles. Specifically, they shifted strongly to Donald Trump relative to the 2012 and 2008 contests. However, in my humble opinion, Ballotpedia’s analysis of why this occurred is lacking.
According to Ballotpedia, the “Marriage Gap” shrunk significantly from 2012 to 2016. In 2012, the marriage gap was a whopping 41 points (Romney won the married 56-42). The article posits that the institution of marriage might lead people to be less dependent on government (certainly plausible) and that single men and women might have become less trusting of government after 8 years of Obama (also plausible).
Hillary Clinton did not just run as Obama’s 3rd term but she also ran as the most progressive candidate in history. Meanwhile, her husband ran as the most centrist Democrat in a generation, well, a generation ago.
But, let me posit some more likely reasons for why the marriage gap shrunk in 2016. It has little to do with ideology and more to do with education, income and electoral geography. Afterall, from the 2016 election we have seen the parties somewhat resort themselves away from their traditional coalitions.
The Marriage Gap 2012 and 2016
|Share of electorate 2012/2016||Marriage status||Obama ’12 /Clinton ’16||Romney ’12 /Trump ’16||Independent ’16||GOP margin|
|58% / 60%||Yes||42% / 43%||53% / 53%||3%||+14 / +10|
|42% / 40%||No||62% / 55%||35% / 38%||6%||-27 / -17|
|29% / 29%||Men Yes||38% / 37%||60% / 58%||4%||+22 / +21|
|30% / 31%||Women Yes||46%/ 49%||53% / 47%||3%||+7 / -2|
|19% / 18%||Men No||56% / 46%||40% / 45%||8%||-16 / -1|
|23% / 23%||Women No||67% / 62%||31% / 33%||3%||-36 / -29|
First-off, let’s start with the table above. It is clear from national exit polls that Trump did slightly worse than Romney among the married and significantly better among single men. Trump even did better among single women. Again, what probably drove this was education and geography.
Trump’s victory was fueled by his surge in the Midwest. He captured dozens of counties the GOP had not won in a generation and held down Clinton’s margins in many other ancestrally blue areas. This is what fueled his surge among singles, especially single men.
The proof can be found in statewide exit polls for many of these states. In Michigan, Trump performed better than Romney, losing singles only by 23 points compared to Romney’s 34 points. Notably, Trump did better among single men than women by 11 points. In Wisconsin and Iowa the same trend shows itself. In Iowa, Romney lost singles by 28 percent, Trump a mere 11 percent. In Wisconsin, Romney lost singles by 24 percent, Trump a mere 8. Again, Trump did better among single men than women.
These states are not unique among the Blue Wall states Trump captured. He managed to garner bigger margins among singles than Romney virtually everywhere in this Democratic leaning region. In downscale regions of many states, including those Trump lost, he ran better than Romney according to county level results.
Now, in politics, for every reaction there is at least an opposite reaction. That reaction was Trump doing worse than Romney among married voters. Trump ran 4 points behind Romney nationally among these voters and there are a few reasons we can hypothesize for this shift.
First, Trump’s coalition was primarily down-scale and Trump did much better among younger voters than Romney. Both down-scale and younger voters are much less likely to be married.
Secondly, Trump’s entire candidacy was based on shock politics. He said crass things, attacked people, etc. To a married voter ensconced in a relationship this might rub them the wrong way. Additionally, a thrice-married candidate might not have the same curb appeal to traditional, socially conservative married families (Utah being a prime example of this).
Thirdly, Trump did better among blacks and Latinos and worse among Asians compared to Romney. Asians are much more likely to be married than blacks and Hispanics. So, those small differences (according to exit polls) could have had an impact at the margins.
Lastly, education was highly correlated with the results of this election. Marriage and education also are highly correlated. Married couples tend to have higher education levels than singles. As a result, Trump performed marginally worse among married individuals.
The last point might not continue to carry over however. Some debate whether marriage will continue to alleviate social ills and ensure educational and economic progress for individuals. Additionally, demographers have found the number of singles who are simply cohabitating has dropped over the last decade (to say nothing of married couples).
Admittedly, this is all very, very preliminary analysis. Until more Voter File Data becomes available we will not know the exact characteristics of many voters. Exit polls are good to a degree but self-reporting tends to have its own built-in biases. Still, for now it will have to do.