Some of the very most interesting political leaks aren't the illegal acts of crusading disaffected insiders. Rather, they're simple boneheaded misjudgements, made in haste and repented at long leisure - such as a SuperPAC's highly revealing internal
chat server logs inadvertently left open to the public and indexed on the web. Or, as in today's example, a presidential campaign's casual name-dropping on its moneybomb site...at first blush seemingly harmless, but when repeated tens of thousands of times and statistically analyzed by a curious bystander...priceless.
Last week the embryonic Rand Paul campaign unwittingly put its foot into just that latter sort of leak. And the insights their mistake affords us all are interesting indeed.
Kentucky senator and presidential candidate Rand Paul’s strategy for winning the Republican nomination hinges, in large measure, on his claim that he alone among all GOP contenders attracts new blood - particularly young voters, women, and people of color - and can thus counter Republicans’ otherwise inevitable demographic slide into irrelevance.
It’s a claim which, at first glance anyway, might seem credible. His avowed “libertarian-ish” positions on NSA-style surveillance, marijuana decriminalization, and foreign military intrigues, to name but a few, are perhaps more in line with Millenials’ sentiments than are those of old-school Republican hawks. But then again, with regard to minorities he is perhaps on shakier ground, where his occasional feel-good campaign talk is more than amply negated by embarrassing revelations regarding neo-Confederate staffers as well as the candidate’s own past words.
Still, according to Politico anyway, Paul’s big-tent strategy might just have a shot:
The [Galt House] rally on Tuesday was a manifestation of his attempt to broaden and boost his standing, particularly among young and minority voters. It’s hard to remember a Republican candidate who has so highlighted diversity in a campaign launch: on the stage behind him were women, blacks, and young people
Women, blacks, and young people?
Holy crap! The Democrats are
toast. Paul's gonna run the table.
But of course, any competent political operative can pack a stage with whatever sort of human resources an event’s script might call for. What really counts are hard numbers regarding hard-core followers - the little guys, by the hundreds of thousands, who will follow their leader into the jaws of hell. Not merely the 'likely voters' that are the life blood of opinion surveys, nor the breathing stage props of carefully orchestrated media events, but the true believers.
By flashing onscreen the names and home towns of small grassroots donors by the tens of thousands, the current Rand Paul moneybomb drive inadvertently provides us a powerful tool with which to get to know his supporters much more intimately than the campaign probably intended. Tell me a person’s name and his or her city and state of residence (see the screengrab at the top of this diary) and I can, in many cases, quickly determine that person’s sex, age, race, political party affiliation, and much more...all without overstepping any ethical boundaries. Tell me enough of those names and hometowns - say, tens of thousands, as Paul's moneybomb does every day - and a statistical analysis of that nice big juicy data set can reveal a lot about that group as a whole. In short, thanks to his overeager web designers and their surprisingly lax supervision by campaign managers, we can rigorously test Paul’s claim that he brings new blood to the Republican party.
How This Works
First, we tested the reliability of the donor information displayed on the Paul moneybomb page by making a few small donations and observing that the names and hometowns we submitted were displayed in near-real time. Then we used commercially available software to visit the page at one- to three-minute intervals (a rate chosen to avoid interfering with the site’s function), around the clock, from April 7th through 9th, automatically capturing and databasing the names, cities, and states of a more-or-less random sample of the many donors identified on that page. After purging the database of duplicate records, this left us with a sample of 939 unique donors from a population (i.e., all those who donated during the sampling period) of just under 50,000, for a 3% margin of error. We then submitted each name/location pair to a leading online people-search tool that draws on aggregated public records, in order to determine each individual’s age. Finally, we assigned a likely sex to each of these names via automated analysis (no sex was assigned in ambiguous cases, such as given names like ‘Chris,’ or ‘Pat’). In order to test the reliability of this analysis and to gain additional insight, we added on a follow-up observation period during which we collected only the names of donors reported to reside in North Carolina - a state which is unique in that makes its voter registration data (including name, address, sex, race, and party affiliation) freely available online.
Results
Age: As the graph below illustrates, Paul's grassroots supporters are, on average, very much older than the voting-age population of the U.S. as a whole, and dominated by supporters 55 or older (more than half of all donors in our sample, but only a third of the voting-age U.S. population).
Particularly noteworthy is the near total absence of college-age individuals (18 to 24 years old) from our sample of Paul supporters.
With respect to the striking absence of young people from the ranks of Paul supporters it is, admittedly, important to consider that the nature of our data might skew this result to a certain extent. Elders, as a class, would be expected to have higher net worths than youngsters, and so might be a little more likely to make donations. But this alone can't explain the stunning geriatric shift we see here. Plenty of online fundraising drives that engage the real concerns of young people draw in large numbers of youthful donors. Young folks just aren't that into Paul.
Age distribution of Paul donors in our sample (black bars) and in the voting age U.S. population overall (grey bars)
Sex: We were able to assign a likely sex to a donor’s name in 92% of cases.
Males comprised a whopping 81% of those donors, while they comprise just 49% of the U.S. population as a whole.
To double-check this rather stunning result, we searched official state data for a total of 56 North Carolina Paul donors, using the state’s NC Public Voter Search web site. In good agreement with the results from our nationwide sample, males comprised 87% of North Carolina supporters, while making up just 48% of the state’s population overall. Apparently Paul's confused libertarianism, combining abhorrence of government-sponsored social safety nets with strong support for government intrusion in women's healthcare, isn't exactly driving the ladies wild.
Ethnicity: Among residents of North Carolina (the only state for which reliable ethnicity data is publicly available, to our knowledge), a near-perfect 96% of Paul donors classified themselves for voter registration purposes as ‘White,’ 2% as ‘Black,’ and 2% as other ethnicities. By contrast, whites and blacks comprise 71% and 23%, respectively, of North Carolina’s registered voters. Paul’s appeal among non-white voters appears to be very nearly undetectable.
Distribution By State: A plot of donor frequency by state, like that below, makes it clear that Paul supporters are largely concentrated in the Duck Dynasty: a handful of thinly populated deep red states like Alaska, Arizona, Idaho, Kentucky, Montana, Nebraska, West Virginia and the Dakotas - all of which went for Romney in 2012. Among blue states in 2012, standout concentrations of Paul supporters are found only in Nevada, Vermont and New Hampshire (the latter - Paul's number-one fanboy state - being home to the small but noisy libertarian Free State Project). Importantly for Paul’s early primary prospects, Iowans are not particularly strong supporters, while New Hampshire Free Staters give him a better shot there.
Frequency of Paul donors in our database, by state (per 1 million state population)
Conclusion
According to the Paul campaign’s own data, his actual grassroots supporters look nothing at all like the carefully staged backdrop of human flesh on gaudy display at his Galt House announcement rally last week. Instead, his supporters are overwhelmingly old white men from a mere handful of what are sometimes ungenerously called fly-over states. Paul's supporters don't really 'stand with Rand'...it's more like they hunch over their walkers with him. Which is to say, they are largely indistinguishable from plain vanilla Republican voters.
Women, young voters, minorities and purple states - key demographic segments in which the GOP must make inroads if it is to survive - aren’t buying Paul’s libertarian-lite line. Despite his posturing, Rand Paul appears to bring nothing to the GOP table other than smoke and mirrors.
Wed Apr 15, 2015 at 9:04 PM PT: About 60 hours after this diary went to press, the design of the moneybomb page at randpaul.com was changed, deleting the javascript widget that displayed donor names and home towns. Nice to know that the Paulists read Daily Kos too (just, you know, kinda slowly....)
Thu Apr 16, 2015 at 8:00 AM PT: And, less than 12 hours later, the name & home town widget is back again. Seems Rand is having a hard time deciding which is worse: being embarrassed, or looking embarrassed.
The raw data underlying this analysis - plus much, much more that still remains unanalyzed - is available to reputable parties on a case-by-case basis, for non-commercial use only. Contact DocDawg666 at gmail-you-know-what. Please include your professional affiliation, job title, work email address, work phone number, and briefly describe your intended use of the data.