A tremendous number of clues are encoded in our names: our gender, age, ethnicity, even from which region of the United States we hail. A person named Schmidt is much more likely than average to have German ancestry, and therefore to have ancestors who settled in the Midwest. If a woman is named Helen, which peaked in popularity in the 1910s, she is more likely to be of retirement age than if her name is Ashley, which was all the rage in the ’80s.
Applied to a single person, this is called stereotyping. Applied to a group, it’s called microtargeting.
I first noticed that names seemed to loosely correlate with politics when thumbing through a list of delegates during one of the political conventions. The Republicans seemed to have a hold on the Donalds and Sharons, while the Democrats were rich in Angelas and Willies. But that was a sample size of a few thousand people, and not much of a sample at that: Those positions go to elected officials and local party bigwigs.
But the Federal Election Commission has a much larger database of names and political associations. Candidates for federal offices are required to report the name of any person who has given them at least $200, whether it came in one large check or dozens of smaller donations. (A psychotherapist in Chapel Hill, N.C., for example, has given $850 to President Obama in 81 installments to date.) The Obama campaign has reported 1.75 million individual donations through Aug. 31, while the Romney campaign has 353,000. The other Republican primary contenders reported a combined 320,000 donations. These records include some people who have not hit the $200 threshold but whose donations the campaigns reported anyway.
These donations come from roughly 452,000 unique people who gave to Barack Obama and 315,000 who gave to Mitt Romney and the other Republican presidential candidates. (The dreary methodology and code for how I computed that is at the bottom of this article.)
That adds up to 21,186 Johns, 7,081 Susans, 1,108 Harrys, and 41,400 other unique first names.
The data show that names are a strong predictor of support for one party or the other. Looking at names that occurred at least 1,000 times on the donor rolls, it is immediately evident that women give in much greater numbers to the Obama campaign, a fact that the site OpenSecrets.org has also observed. Within a gender, however, some names have much stronger correlations than others. While people with very common names—James, David, Michael—are roughly evenly split between the two parties, names like Brent, Tyler and Clayton are considerably skewed toward the Republicans.
Nicknames and diminutives also correlate to party preference. Christopher—just to pick a random name out of my hat—tilts slightly to Republicans, while “Chris” trends slightly Democratic (though that could be because there are some women nicknamed Chris out there). People named William have a 57 percent chance of supporting the Republicans, while Willies are the most Democratic name on the list at 93 percent.
Along the same lines, people named Liz are extremely Democratic, with only 11 percent donating to Republicans. But 26 percent of Elizabeths give to the GOP, and Betty is one of the most Republican women’s names on the list, with 37 percent of women who share their name with Barney Rubble’s wife ponying up cash on behalf of a candidate who wanted to replace President Obama.
Of all names that appear at least 25 times, the most Republican men’s name on the list is Brent. (Willard is No. 4!) The most Republican women’s name on the list is Ashley. But that could include some men, as could the next two women’s names, Kelly and Courtney. So let’s declare Patsy, the No. 4 contender, the most Republican women’s name.
We’ve already mentioned that if you want to raise a Democratic son, name him Willie. Democrats expecting a daughter should go with Gwendolyn, the most pro-Obama girl’s name on the list.
So, while we have to be careful not to generalize too much from the list of presidential donors in 2012, it’s clear that in broad strokes, names can be useful for candidates who want to identify potential supporters and donors. Candidates who can’t afford fancy market research can get a lot of the same data from the phone book.
Follow Chris Wilson at @chriswilsondc or email him at email@example.com.
Methodology: FEC data is not the cleanest in the world. The task here was to take a bunch of names, many of which are listed multiple times with small variations (like a period after a middle initial), and reduce them to unique individuals.
I reduced each record to a first and last name only, and assumed people were unique by these two names, their ZIP code, and the candidate to which their donation was made. Thus, two John Smiths in different ZIP codes will not be confused with one another. However, two distinct John Smiths in 22901 each giving to Mitt Romney would be counted as one person. (In browsing the data for common names, I didn't once find an example of this.)
You can view the source code here.