Who Supports Sanders, Who Supports Clinton?

Although there is still a decent amount of runway left in the Democratic nomination race, the shape of each candidate’s primary coalitions do seem to be relatively set. These coalitions explain a great deal about the race, who is winning and why.

First, an important caveat. In identifying a voting bloc that supports one candidate or another, it is crucial to remember there were always be individuals within that bloc who will support the other person. Too often in trying to get the categories right, the media forget this basic reality, which upsets a lot of people and justifiably so. Still voter categories are incredibly important for predicting elections. Often times demographic head counts will give a better sense of how an election will go than the available polling. If a poll does a poor job of reflecting the demographic head count than it is a bad poll. This is one reason the polls predicting a substantial Clinton victory in Michigan were so wrong--they contained poor assumptions about the makeup of the electorate. Pollsters assumed it would be older than it turned out to be; when it wasn’t, everything flipped.

The biggest gap in who votes for Sanders and who votes for Clinton is generational. No matter what other demographic one belongs to the older you are the more likely you are to support Clinton, the younger you are the more likely you are to support Sanders. Sanders young supporters have influenced greatly the narrative of the contest. It is not surprising that young voters receive more attention. They are more interesting. They make for bigger crowds and better visuals. They also help power an online giving campaign because they live online. But they also remain a smaller portion of electorate. Under 30s were outvoted by over 65s in nearly every state.

The second biggest gap is a gender gap among almost every demographic and in every state (a few exit polls produced different results but they don’t seem reliable). Male voters vote more for Sanders than for Clinton. This was the case in 2008 as well; men also were better for Obama. Something other than sexism may be at work, or it may just be sexism. In addition, some women may be voting out of gender solidarity. (We have a way to decipher whether the attitude is sexist or not--the person’s willingness to criticize President Obama for the same behavior as he or she is criticizing Clinton. If the person criticizing Hillary for taking large campaign contributions, for example, didn’t criticize Obama for the same practice, they are more likely sexist than a thoughtful critic.) Whatever may be behind it, women’s support has clearly benefited Clinton. Women are a much greater portion of the Democratic primary electorate. To be fair, the younger the voter, the smaller the gender gap. Among voters under 30 the gap mostly disappears.

The third biggest gap is with respect to party self-identification. In almost every state, self-identified Democrats have voted for Clinton. Sanders has done better, and almost always won, self-identified independents. Some of that is attributable to the age gap because young people are far less likely to identify with a party, but that does not seem likely to be all of it. Clearly many independents like Sanders’ message of being outside the establishment. But independents make up a smaller part or primary voters.

The next gap is with respect to class. There are lots of interesting things to note here. In particular, it is interesting that Sanders does in fact do better down the income scale. Again, this is somewhat attributable to age. Educational attainment and income, the two categories we looked at as proxies for class, go up with age. Still, the data seems to suggest that, even controlling for age, either attaining a post-graduate degree or earning over 100k a year makes you more likely to support Clinton. Exit polls in a few states suggested that Clinton might have done better with lower income voters, but those polls didn’t take into account income and race. Generally, having only a high school degree makes you more likely to support Sanders in places with a white population, but less likely in less white places. Again leaving age aside, Sanders’ message is resonating with whites further down the income scale. There’s another crucial point, however. Voters as a group are much better educated and richer. This fact has allowed Clinton to win states, such as Massachusetts, that she likely would not have won had the electorate looked more like the state as a whole.

The final gap is race and it is the one we have covered the most here. African Americans are voting at about a three-to-one or slightly better clip for Clinton. They powered her to very close wins in Missouri and Illinois (despite Missouri being Sanders’ best state with African Americans, losing only two-to-one). They kept things very close in Michigan and mitigated the delegate loss. The evidence is more mixed with other non-whites. Clinton won Hispanics big in Texas and Florida, but Sanders won them fairly comfortably in Illinois. In Michigan, Muslims and Arab Americans went to Sanders. For Asian American/Pacific Islanders, the evidence is so far inclusive. What happens with non-whites other than African Americans will likely determine the margins in coming states.

Conclusion: From these gaps, it should be clear why Clinton is winning: almost every group she claims, older people, women, self-identified Democrats, the better educated, minorities, are a larger part of the primary electorate (less well-off people are the exception).

Her supporters also provide an interesting explanation for what has often been referred to as an enthusiasm gap between her and Sanders. It may be that Sanders supporters are considered more enthusiastic not because they are but because they culturally seem so. It is also not surprising that a candidate whose supporters are whiter, younger, more male, and less typical voters would be described as having the more enthusiastic base, while the candidate favored by older, female, less white and more typical voters would be seen as drawing less enthusiastic support. What is also true is that Sanders supporters in all demographic groups tend to be louder and tend to give more money. This also creates the perception of enthusiasm. Clinton’s base groups are taken for granted partly because they are more reliable.

The odds are excellent that Clinton will win a majority of the pledged delegates but if, on some fluke, she does not and delegates need to decide to whom to thrown their support, they should not be distracted by Sanders’ “enthusiasm.” New does not necessarily mean better and dissing your core base for the new people, especially if the core outvoted the new by over a million or so votes, would be a mistake. Clinton supporters are too easily marginalized but there also are a lot more of them.