On what might be Terminal Tuesday for the Republican Party it seems appropriate to ask the question exactly what kind of people vote for the fascist buffoon who is on his way to become its nominee? This morning the New York Times published ten demographic variables that correlate with voting for Trump, seven positively and three negatively. As every statistician knows, correlations between independent and dependent variables can be spurious, in which case intervening variables have to be identified that explain both the independent and the dependent variables, but in this case there can be little doubt about the causality, and a more detailed study would probably find strong correlations between some of the independent variables as well. The strongest correlation with voting for Trump is shown by people who are white and don’t have a high school diploma. These are the ‘poorly educated’ that Trump often declares his love for. The second strongest correlation goes to people who identify themselves as ‘American,’ rather than by some kind of ancestral ethnicity. They are the ones who start chanting ‘America, America’ when a protestor shows up, and cannot stop declaring that ‘we’re number one.’
The third strongest correlation is for people who live in mobile homes. In the US this group is often derogatively referred to as ‘trailer trash,’ but their housing conditions are probably best explained by their economic status. This is illustrated by the fourth strongest correlation, which shows that Trump voters, if working at all, have ‘old economy jobs,’ in agriculture, construction, manufacturing and trade, and by the first negative correlation, showing that they have a low labor participation rate, which means that many of them are either looking for a job or have given up on finding one. This is the group that cannot be fooled by rosy US employment numbers, which by design camouflage the real unemployment rate. The fifth, sixth and seventh positive correlations tell the rest of the story: A sizable part of Trump’s supporters has a history of voting for segregationists, which would obviously suggest racist inclinations, they tend to be born in the US, and many of them are Evangelicals, an electoral segment that Trump is stealing from Ted Cruz and borrowing from Ben Carson. As for the last two negative correlations, Trump supporters don’t vote for liberal Republicans and very few of them are WASPs, like the Donald himself.
The correlations paint a pretty clear picture that allows us to create a stereotype: People who vote for Trump are in majority uneducated, un- or underemployed, white, racist, consider themselves Christian, and were born in the USA. They are afraid that blacks or Hispanics will take the jobs they cannot get, and that Muslims will come to the US to wage Jihad. Maybe this is unfair to those who simply want America to be great again and don’t even think Obama screwed it up, but that’s what stereotypes do.
I don’t know if it exists, but I assume that historians have compiled a profile of Germans who voted for Hitler in the 1930s, and that there are significant similarities: Under-educated, racist and anti-Semitic, economically disadvantaged and therefore disgruntled, considering themselves Christian and eager to make Germany great again. What ‘the Wall’ is for Trump supporters was the ‘Atlantic Wall’ for them.
The good news is that only 23% of the voters identify themselves as Republicans and that Trump at most gets 50% of their votes. But if he can get a sizeable part of independent voters in the general election it’s still a dangerous situation, so the Democratic Party had better lock in on its strongest nominee soon.