Candidates and campaigns win elections in the environment that exists while voting is conducted. While that statement isn’t terribly revealing it points out the three primary components of an outcome. Activists like myself can’t change the candidates (at least after they’re nominated), the environment is largely out of our control, but we do impact campaigns by volunteering our time, our treasure and our talents. What happened on Tuesday at the local level should be a wake-up call for conservative activists, and there’s a path ahead that might lead to a different outcome if we learn from this defeat.
We’ve heard for a while now about how Democrats have been building a database of voters and making sure those voters get to the polls, and Tuesday we got a chance to see a second iteration of this rather impressive effort. A hallmark of the ‘08 election was a higher than usual turnout that included a lot of new voters. As hard as it may be to think Democrats could eclipse that remarkable performance, they did. Turnout once again jumped, and the differential voters showing up were even more likely to be Democrats than they were in ‘08. Forget stories about “voter intensity,” “buyers remorse” and the other largely anecdotal narratives that were spun to encourage people hoping that Obama might get kicked out of office. It was all about a decidedly non-sexy effort driven by data and hard work.
Doesn’t the Republican party have data that they do something with? Sure. At the local level it’s called “Voter Vault” (UPDATE: which has been re-vamped and re-branded as “GOP Data Center”, a change that makes all the difference. Virginia and Florida were the wind test dummies for this change to noticeably underwhelming effect). The data starts with a voter list from the State Board of Elections, and RPV expects campaigns to fill out that list with information they collect during a campaign about voter preference. That hardly ever happened in my experience, as campaigns almost always considered their data a jealously guarded secret not to be shared with anyone else. As a result, about the only voter preference information in that database is derived from who shows up to vote in which primary election. That’s fine if you’re dealing with a low-turnout election where those folks constitute a huge proportion of the electorate, but you have no visibility at all of voters who only show up occasionally and never vote in primaries. As a tool to win elections in Virginia, it’s nearly useless.
Democrats started building a quality database in ‘08 and have been consistently working to refine it ever since. Instead of one data point — participation in a party primary — that list is rechecked several times to make sure voters are who Democrats think they are. The effort is strictly controlled from the top and candidates don’t have the option to build their own data and not share it with anyone else, and given the quality of that data no candidates would ever want to. Data comes from canvassing run by the national party, a dedicated online interaction effort with voters, and from information collected from volunteers. If you’re a Republican, you didn’t get one phone call, one piece of mail, or any other contact from Democrats. They focused their effort on those voters they knew. It’s sharpshooting with a rifle at a specific target instead of blasting away with a shotgun at anything that moves.
In the Manassas area, the only area where additional turnout favored Republicans was in the Brentsville District (adjusted for redistricting) where about 52% of the additional votes where for Romney with a net partisan increase of 173 votes. In Coles, 71% of additional votes were for Obama for a net gain of 721, in Gainesville 57% of additional votes went to Obama for a net gain of 387. In Manassas, 66% of additional votes were for Obama for a net gain of 474 and in Manassas Park a whopping 87% of new votes were for Obama for a net gain of 351. If “buyers remorse” was a consideration in any of these localities, it was easily overcome by identifying and driving new Democrat voters to the polls. When you look at individual precincts the picture gets even more stark.
In Manassas every precinct substantially increased the number of Obama votes, exceeding the number of additional Republican votes over 2008 except in the absentee precinct where Democrats unexpectedly dropped the ball and got 434 fewer absentee votes than in 2008. The effect is more pronounced in typical Democrat-leaning precincts, but appeared everywhere. Here’s a table that shows the increase in vote totals between 2008 and 2012:
This same story plays out in Manassas Park and western Prince William County (except in the Brentsville District), so this isn’t a case of a party organization in one locality doing better or worse than others. Manassas Park’s local GOP left at least half of their precincts unmanned and virtually without signage on election day, which probably is consistent with how hard they worked leading up to election day, while Manassas City had a very organized effort with fully staffed precincts on election day. The Manassas City Democats spent all their time in Manassas Park trying to get Jeanette Rishell elected (the succeeded) and didn’t devote all that much time to Manassas City. Despite this, there’s really not much difference in the results between these two. Clearly something bigger than local committees were driving the results.
Brentsville is the outlier in all this. Republicans managed to perform better than in 2008 and did better than Democrats by 173 votes. That’s not huge performance increase, but it is a performance increase where everywhere else there was a decline. It’s not readily apparent what precincts were driving this as the precinct lines underwent significant changes between ‘08 and ‘12 but it does seem that it was pretty even across the board. The difference in Brentsville may be due to the effect of Jeanine Lawson’s campaign against Wally Covington. Lawson’s campaign might have ended but the organization that was built has remained, giving these activists a substantial data set of people they know well and those activists didn’t sit on their hands this election cycle. It’s also possible that Democrats decided to bypass the strongly-Republican area to avoid “riling up” Republicans but since Obama’s turnout increased by 24% over 2008 while Coles and Gainesville increased by only 9 and 15% respectively, Democrats clearly weren’t ignoring the district. It could be argues that they focused on it instead. It’s important to note here that the precincts used to calculate all these numbers have been adjusted based on redistricting, using the current magisterial district boundaries, as much as it’s possible to do without splitting precincts.
The solution suggested by all this is for local activists to start building their own databases at the magisterial district level to hopefully overcome the shortcomings in Voter Vault, but even that is going to likely come up short in a lot of places. RPV has to make a dedicated effort to improve the quality of that database and put it to good use, or we’re going to see this happen in every presidential election in the future. Democrats will turn out increasing numbers of marginally-interested voters and in a few places local efforts might blunt that effort, but not very often.
The 2013 elections will be rather different, as Democrats are relying on marginally-engaged voters who might be interested in a presidential election but far less likely to turn out at other times. Their rather considerable challenge is to make this happen in off years without the kind of vast budgets they had to work with this year, and with candidates like Terry McAuliffe at the top of the ticket who are much harder to sell. The challenge for Republicans is to spend the next four years building the kind of high quality database Democrats used this year, forcing elected officials and previous candidates to share their voter lists, and using that database well instead of simply buying whatever services entrenched vendors can sell to candidates and the party at the highest profit. That will take time, money and effort, but there isn’t much choice.
If this doesn’t happen, we better get used to a socialist national government because that’s all we’re ever going to have. Activists can force this change at RPV and at the national level, as well as working at the local level as was done in Brentsville this year to implement that change. Fortunately with Ken Cuccinelli running for Governor and favored to win the Republican nomination, who is arguably the best grassroots organizer in modern Virginia history, there’s a very unique opportunity to do just that. His list won’t be as comprehensive as what Democrats are using, but it will be very high quality and a great place to start.
For a fascinating look at behind-the-scene data operation the Obama campaign was running, check this Time Magazine piece out. Here’s an excerpt:
The new megafile didn’t just tell the campaign how to find voters and get their attention; it also allowed the number crunchers to run tests predicting which types of people would be persuaded by certain kinds of appeals. Call lists in field offices, for instance, didn’t just list names and numbers; they also ranked names in order of their persuadability, with the campaign’s most important priorities first. About 75% of the determining factors were basics like age, sex, race, neighborhood and voting record. Consumer data about voters helped round out the picture. “We could [predict] people who were going to give online. We could model people who were going to give through mail. We could model volunteers,” said one of the senior advisers about the predictive profiles built by the data. “In the end, modeling became something way bigger for us in ’12 than in ’08 because it made our time more efficient.”
UPDATE: Apparently the Romney campaign tried their take on big data, but the program was a disaster. Instead of a data warehouse to identify voters and get them to the polls, they developed an “app” that no one was able to use, and insanely focused on robo-calling, which was probably the biggest waste of money and effort a campaign could have made this cycle.
“So, the end result was that 30,000+ of the most active and fired-up volunteers were wandering around confused and frustrated when they could have been doing anything else to help. Like driving people to the polls, phone-banking, walking door-to-door, etc. We lost by fairly small margins in Florida, Virginia, Ohio and Colorado. If this had worked could it have closed the gap? I sure hope not for my sanity’s sake.
“The bitter irony of this entire endeavor was that a supposedly small government candidate gutted the local structure of GOTV efforts in favor of a centralized, faceless organization in a far off place (in this case, their Boston headquarters). Wrap your head around that.”
UPDATE 2: I suspected facebook was a major part of OFA’s efforts. Turns out that suspicion was warranted.
UPDATE 3: Republicans try to figure out how they will respond. Since they’re still stuck in a “list management” solution implemented from the top down, I’m not encouraged at all.
The opinions expressed here are solely the views of the author, and not representative of the position of any organization, political party, doughnut shop, knitting guild, or waste recycling facility, but may be correctly attributed to the Vast Right-Wing Conspiracy. If anything in the above article has offended you, please click here to receive an immediate apology.
Leave a Reply