The deal that Chad and I have with Campaigns & Elections is that we get to do whatever we like with our monthly columns sixty days after they’re published. So here, a little overdue, is my column from the May issue. Note that I refer to “Demzilla” in the column to describe both Demzilla and Datamart, two separate software packages that are often described with the umbrella name of Demzilla — I’ve stuck with the nomenclature used by the DNC, but I really should have tossed in the Datamart name there somewhere.
While I’m on the topic, who was nice enough to give my parents a gift subscription to C&E?
The Demzilla Downfall?
One of the narratives that emerged from the 2004 presidential race was that the Bush campaign and the Republican National Committee had better voter data than the Kerry campaign and the Democratic National Committee. The DNC’s newly created, much-touted “Demzilla” database, 170 million records strong, had lost out to the RNC’s “Voter Vault.”
By late 2005, Demzilla had been effectively abandoned–it had been a failure. The software was comically slow. It hadn’t been designed for real-world use. The data provided by many states was so bad as to be useless.
Now all is as it was pre-Demzilla. State parties’ lists are balkanized, there’s no flow of data to or from the DNC. A twenty-year voter with a detailed record on file with the Michigan Democratic Party is a total unknown if she moves to Virginia. The DNC can’t even say who comprises the Democratic Party.
Perhaps the most fundamental obstacle to the effort of building a master list is the existing economic structure. Many state parties jealously guard their lists, hoping to offset the costs of building and maintaining those lists by selling them to presidential candidates every four years. If they give those lists away to the DNC, they lose the very revenue stream that makes the lists possible. Though some states may be willing to take the hit in exchange for receiving enhanced data and analysis from the DNC, they can be forgiven for doubting that the promised benefits will materialize.
If they’re serious about addressing the list gap, the DNC will need to address each dubious state’s economic and trust concerns. The voter file doesn’t need to be everything to everyone–it’s fine if it’s nothing more than a data warehouse, provided that the party is prepared to assist in the creation of bidirectional conversion modules for every state’s existing voter file. In addition, a lightweight, Internet-based precinct- and district-level list management package marketed directly to local chairs would prove a tempting incentive for state participation; if states want to be privy to their own precincts’ data, they’ll have to join up.
The DNC says they’re taking another shot at creating a master voter file with their National Voter Database Project. DNC representative Luis Miranda optimistically describes the system in the present tense, saying that it’s being introduced gradually on a state-by-state basis. “It’s the responsibility of the DNC that we have a voter file at a state and national level,” Miranda said. “We’re working closely with the state parties to make this happen.”
We’ll see.
I will say that I have seen the Voter Vault lists for the 44th District, and they are pretty dreadful — filled, frankly, with Democrats. Don’t know about the entire rest of the world, however.
From what I understand, at least in Virginia, the Democrats lists are significantly better than the GOP list. The GOP voter file is a joke, while many Dem committees lampoon the DPVA voter file, it’s a million times better than the GOP file and with the new staff at the party it will get even better.
Effective list management may be one reason why Tim Kaine was able to turn out his voters more effectively than Jerry Kilgore was able to turn out his. A complete voter management system complete with voter history, volunteer history, and donor history is absolutely critical to the future success of the party– and it starts at the local level. A fundamental problem with the DNC that the GOP does not have (not just in list management but in everything) is that the Dems are much more decentralized in everything from beuracracy to ideas. For better or for worse, the Republican party has less breadth and is more top down– which enables a nationwide voter file. Perhaps Democrats would be better served if the state parties perfected their files first. Terry Mac raised much needed funds, and now Howard Dean is focusing the DNC to be a state first party. I think this is good. Will it work? We’ll see.
Tim Kaine was able to rely on some good files, but for the most part, he created them. This was one of Tim’s biggest priorities for his campaign: infrastructure. And I believe we’ve already seen success with the excellent Democratic performance in special elections after November.
I forget the exact numbers, but the number of new contacts that were gathered by the coordinated campaign alone was staggering.
I was VERY well acquainted with the DPVA 2005 database. As I have been doing canvasses this year in the same areas I worked last year, seeing the results of our work has been great. The 2006 database is streamlined from the 2005 version, with a lot of bad addresses and other garbage gone. Voter IDs got a lot more depth to them as well. Now that I know who we’re targeting and who we’re not targeting this year, and that all of our work was utilized well, the whole process is a whole lot nicer.