How We Could Have Won the Election for Deeds and Byrne

By: Lowell
Published On: 11/12/2005 2:00:00 AM

Yeah, right, you're thinking, Raising Kaine could have "won the election for Creigh Deeds and Leslie Byrne."  And you could have eliminated the national debt, too?  Cured the common cold?  Made computer software that actually did what it was supposed to do?  Captured Osama bin Laden?  Anything else?  Ha ha.

Well, I'm not kidding.  Back in the summer, we -- actually, one of the most brilliant people I've ever met, Greg Priddy -- came up with a detailed, sophisticated plan for targeted political marketing using Yahoo banner ads.  Here's the concept of the plan and how it would have worked exactly.

* Off-year elections in Virginia have low turnout, around 44% this time and 46% in 2001.  This compares to 68.5% in the 2004 Kerry vs. Bush presidential election.  That's a difference of more than 20 percentage points.

* The difference in turnout was huge in a place like Arlington, where only  56,972 showed up this year. compared to 94,650 in November 2004.  That's a drop of 37,678 votes just in Arlington County alone.  Also, it is important to point out that 73% of votes in Arlington were cast for Leslie Byrne over Bill Bolling, and Creigh Deeds over Bob McDonnell.  Quick math:  that's 27,505 votes, ENOUGH TO HAVE WON THE ELECTION for Creigh Deeds and Leslie Byrne -- just in Arlington County alone.

* OK, you're saying, that's all wonderful on paper, but how in hell would you guys have increased turnout to 2004 levels in Arlington?  Answer, we wouldn't have.  The point of the preceding bullet was simply illustrative of the power of increased turnout.  Our plan aimed far lower, to increase turnout by just 1%-2% statewide.  Out of 2 million votes cast, we believe that this could have resulted in an extra 20,000-40,000 votes for Kaine/Byrne/Deeds statewide.  For the record, Byrne lost by 23,016 votes in this election.  Deeds...well, who knows, but right now he's trailing by only 620 votes.  Hmmmm.....

* Now, some specifics of our plan.  First, this was internet-based GOTV using Yahoo, the goal of which was the passive absorption of a simple message while browing the web.  Yahoo is the largest online portal in Virginia, with around 4.8 million registered users in the state.  This includes, for instance, 103,326 users in Arlington County, which went overwhelmingly for Kaine/Byrne/Deeds this past Tuesday.

* Yahoo allows for TARGETING by zip code, county/city, age, gender, income level and even occupation.  In other words, you can target your ad at women under 40 in Arlington County, a group that overwhelmingly votes for Democrats but also turns out in FAR lower numbers in "off year" elections.  That's why these people are called "federal voters," or more negatively, "lazy voters."  They WOULD vote if they a) knew about the election; and b) understood that it mattered to them.  And that's exactly where the Yahoo ad campaign would have come in.

* How costly would this have been?  Not very.  We're talking a minimum ad buy of $10,000, and a cost of $10 per thousand impressions ("CPM rate").  The concept here would have been to do repetitive saturation coverage, allowing for passive absorption of message, with an average of 5-10 impressions per targeted user, over seven days ending on Election Day.  The size of the campaign was scaleable, but $50,000 would have funded over 5 million ad impressions, or 5 impressions per person for 1 million people!

* Right, ok, but does this thing WORK?  Well, the Republicans used something very similar in the past, in Wayne Allard?s come-from-behind win in Colorado?s 2002 U.S. Senate race.  So, no, we are not 100% sure it works.  But for $50,000 (or less), we sure could have found out!

So why the heck didn't we do this?  Simple answer:  money.  Raising Kaine formed as a PAC in August, and started raising money via small donations.  We raised a few thousand dollars in this way.  To do the Yahoo strategy, however, we would have needed at least $10,000 (for perhaps just Arlington/Alexandria/Falls Church/Charlottesville) or $50,000 for the entire state.  We shopped this proposal around, and we received definite interest, but in the end, we didn't manage to come up with the cash.

Why not?  Good question, but my theory is that there were at least two main reasons: 1) funds were already committed towards other purposes, largely TV ads; and 2) people may have failed to fully understand the potential power of this relatively new concept using the internet.  Maybe it also was our fault for not pushing harder, or more effectively.  I'm not sure. 

Anyway, the point here is not to look backwards at what "might have been" or to point fingers of blame, but to think about the future, about moving forward.  New technologies, new media, new methods of communicating with voters keep coming along in this world of ours.  The question for the Democrats is this:  do they want to be on the cutting edge of all this innovation, or do they want to fall behind to the Republicans?  Do they want to maximize the power of the "netroots," or not?

A related issue, as Greg points out, is that it's high time for Democrats to seriously rethink how their campaigns compensate media consultants.  Currently, most campaigns pay media consultants percentage commissions based on television ad purchases, direct mail, etc.  Obviously, this can result in a conflict of interest in the consultants' advice to the candidate, since they have a vested interest in making money on the standard techniques that they know about. 

Greg points out that one danger of this system is that it can stifle the development of "new media" approaches which haven't been part of the standard playbook. Virginia Democrats did use a small amount of Internet-based advertising, but it was in places targeted at the party base (Daily Kos and other liberal blogs, Google AdWords) for the purpose of raising funds -- NOT in places reaching a wider audience for GOTV purposes.

In contrast, Greg notes, the Bush/Cheney 2004 campaign was smart enough to do away with the commission-based model entirely, and instead to pay their consultants on a monthly retainer basis.  This removed the conflict of interest, and not surprisingly, Bush/Cheney '04 spent a lot more money than did the Kerry campaign on innovative approaches, particularly the $3 million worth of  targeted marketing data and analytical services they purchased from ChoicePoint.  And, not coincidentally, Bush/Cheney won the election.  Something to thing about moving forward even as we celebrate Tim Kaine's victory this past Tuesday...


Comments