Ken Strasma, recognizable to many 538 readers as Barack Obama’s national targeting director in 2008, is the first participant in a new series of interviews we will be conducting with consultants and experts from the field of campaigns and elections.
In this interview, Strasma describes the voter targeting approaches of the 2008 Obama campaign, explains why he thinks Montana is the top target for Obama to flip in 2012, calls Newt Gingrich an unconventional and thus potentially tough 2012 Republican nominee, and predicts that Virginia Sen. Craigh Deeds will win his state’s Democratic gubernatorial primary tomorrow.
First, give our readers a little bit of background on you and Strategic Telemetry.
Strategic Telemetry specializes in providing microtargeting and other strategic consulting services for progressive candidates and organizations. I founded Strategic Telemetry in 2003 in order to help make sure that Democratic candidates didn’t fall behind in the targeting arms race against the Republicans, who under the leadership of Karl Rove were making big investments in voterfiles and data analysis.
My own background is in campaign management. I was using microtargeting as a manager long before there was a buzzword to describe the process. One of my favorite early successes with microtargeting was when I was able to combine polling data from more than 20 individual legislative races in Minnesota in order to find likely swing voters. The morning after a mailing using that model hit, I was listening to a right-wing talk show host in the Twin Cities who was attacking our strategy in sending out the mailing. He felt that because it was attacking the legislative Republican’s votes on education, it should have been targeted to hard-core Republicans who always vote, and he didn’t understand why he hadn’t gotten the mailing. He went on to say that his girlfriend who voted for some Democrats and some Republicans had received the mailing. So that made my day, having the right-wing talk show host confirm for us that our targeting had succeeded in finding his swing-voter girlfriend.
What was your role as a consultant for the Obama campaign?
I served as Obama’s national Targeting Director, and my firm provided the Obama campaign’s microtargeting in both the primaries and the general election. We started planning for the IA Caucuses in late 2006. The primary process lasted longer than anyone expected, so we shifted right into targeting for the general election even before the last primaries had been held.
Without giving away any trade secrets, can you explain to us how you developed the targeting models for the Obama campaign?
At the most basic level, any campaign is about persuading undecided voters, and turning out supporters. Using telephone IDs, we were able to ask hundreds of thousands of voters who they were supporting, and how they felt about certain issues.Those IDs by themselves were very valuable, but there were still millions of voters who we were not able to reach. Using the ID information we did have, combined with demographic and commercial marketing data, we built statistical models that predicted how voters we weren’t able to reach would have answered the ID questions if we’d been able to reach them. Some of the statistical modeling techniques are well known from the academic and commercial marketing worlds. Other proprietary techniques fall under the category of trade secrets that we’ll have to keep to ourselves for the moment. The Obama campaign was in some ways a wonderful 2-year research and development project, with the most aggressive testing of microtargeting models that I have ever seen. We’ll be continuing intense R&D to make sure that we maintain the technical lead we established in 2008.
The Obama campaign used Catalist, supplemented with some other voter databases. Can you give us a status update on the quality of voter lists today?
While I can’t get into the details of exactly what voterfiles the Obama campaign used in the primaries and the general election, I can say that the voterfiles were the best I have ever worked with. There are always problems with voterfiles, but a lot of progress was made this cycle. Because my work is highly dependent on having accurate voterfile data, I was especially pleased with the advances in the quality of voterfiles in 2008. Another major achievement was the quick turn-around on voterfile updates. The Obama campaign focused heavily on voter registration and on early vote. Getting new registrants on the voterfile and flagging who had voted early was extremely valuable. Because we knew who had voted early, and because we had support scores for those voters telling us how likely it was that they were Obama supporters, we were able to estimate our margin from early vote going into election day, and calculate the percentage of the at-the-polls vote that we needed to get to 50 percent +1.
Obama flipped nine states: three each in the Southwest, Midwest and South. How did the demographic challenges differ in each region?
One of the key demographic challenges was to avoid treating groups as if they were all the same. There are significant differences between Protestant and Catholic Hispanics in the Southwest, between Cuban and non-Cuban Hispanics in Florida, and between Hispanics living in majority Hispanic neighborhoods and those who live in more diverse areas.
Once you start looking at voters on the individual level, rather than as monolithic blocks, there are a lot of similarities between these states. There are pockets of similar voters in all of these states, but the share of the electorate made up by the various groups differed dramatically.
OK, I lost a bet to you about Obama winning North Carolina. In my defense, that bet was made in July, before the September economic collapse and McCain’s “fundamentals are strong” blunder. But, to your credit, you were bullish on NC five or six months out. What made you so optimistic?
The North Carolina primary gave us a lot of insight into how the state was changing, and what was doable in the general election. Winning North Carolina wasn’t just about mobilizing African-American turnout, although that was very important. North Carolina’s electorate is changing, with a significant increase in younger, well educated voters who made up another key part of Obama’s winning coalition. Even when things looked bad for a little while when McCain briefly took the lead following the Republican convention, we knew that we were on track to hit or vote goal in North Carolina. That gave me confidence, even when the polling looked bad.
Indiana might be the biggest puzzle of 2008. Aside from the neighboring state advantage, how the heck did Obama swing Indiana so far in one cycle?
Indiana is one of those states whose dark-red status in presidential elections has become something of a self-fulfilling prophecy. Indiana is Republican-leaning, but still competitive at the state level. We made Indiana competitive by choosing to compete there. It is important to understand that this wasn’t any kind of head-fake. We invested serious resources in building an unprecedented field operation in Indiana.
It’s also worth noting that our primary organization paid huge dividends in the general election. People are still debating whether the long primary process helped or hurt, and I don’t claim to know the answer to that question. One thing I am sure of is that winning Indiana would have been much more difficult if we hadn’t been forced to compete there late in the primary process.
We’re a long way from 2012, but if you had to project forward, what will it take for Obama to flip states like Montana and the Dakotas?
I’m very bullish on Montana. It is currently my number one pick to flip in 2012. Energy, land-management and environmental issues are key in Montana and the Dakotas. If, after four years, voters there see that Obama’s policies aren’t the caricatures that Republicans have claimed, we should do quite well.
Which Republican of the potential field of names presently under discussion-Sarah Palin, Mitt Romney, Newt Gingrich, Haley Barbour-do you think would present the biggest obstacle to Obama in 2012?
I would love to see Sarah Palin as the Republican nominee if only for the entertainment value. In terms of who I think would be the strongest opponent, that depends on what type of environment we’re facing. If the economy is still in dire straits, and voters blame Obama, than I would worry about facing Romney. He is an inoffensive, reasonable-appearing candidate, especially compared to some of the other candidates competing for the support of the right-wing of the Republican Party.
However, I expect that the economy will begin to recover, and that Obama will continue to be a popular and successful president. In that case, the Republicans will not win by nominating an inoffensive candidate like Romney. In those circumstances, Newt Gingrich is the candidate who worries me the most. Gingrich has new and unusual ideas. While those ideas are usually dead wrong, and often quite scary, he is something different. You don’t defeat a popular incumbent with a conventional politician, so an unusual choice like Gingrich would be what would worry me the most.
You got to know Virginia demography pretty well in 2008. Any predictions on what will happen Tuesday in the Virginia primary?
The key in Virginia as that we don’t yet know the demographics of the primary electorate. The Democratic primary electorate in VA has been changing over the last several cycles. The share of the Democratic primary vote coming from Northern Virginal more than doubled between 2001 and the 2008 presidential primary. We’ll see on Tuesday who does the best job of turning out their vote.
Most public polling is showing Deeds and Moran gaining and McAuliffe dropping, but the numbers are close enough that a good GOTV operation could make the difference for any one of the three candidates. I see the most likely outcome as a Deeds win, but McAuliffe could still win if Deeds and Moran continue to split the “non-McAuliffe” vote. If Moran’s supporters begin to defect to Deeds then there is probably no way for McAuliffe to win what would then be functionally a 2-person race against Deeds.