## The campaign metric we need to understand

by Taylor Daily

In the aftermath of the 2012 election, we saw the importance of mathematics in both understanding and successfully participating in electoral politics. Not only was Barack Obama the election winner, but so were the Nate Silvers of the world. Their desire to bring more mathematics into politics has proven resoundingly successful.

We already have many campaign metrics for judging campaigns and elections: Cook Partisan Voting Index, margin of victory, dollars spent per vote, and so on, but I would like to propose for consideration a metric which seems to add a depth of analysis to what we currently utilize.

The metric is the marginal cost per marginal vote, or the Spending Inefficiency Index, as I like to call it. To measure a race’s SII, simply take the margin of victory for the winning campaign and divide by the fundraising advantage. For example, if Candidate A won an election by 1,000 votes but spent \$5,000 more than Candidate B, Candidate A’s SII is \$5 per vote. A lower SII indicates spending efficiency, i.e. more marginal votes for each marginal dollar spent, while a high SII indicates that a campaign spent great amounts to capture few marginal votes.

I have calculated the SII for every Minnesota House of Representatives race since 2008 to see what the SII can tell us. I selected Minnesota since it is my home state and it presents a nice variety of upsets, third-party candidacies, and some of the most-funded House campaigns in history. However, I would like to disclose that I worked on two of these campaigns (Tarryl Clark in 2010 and Brian Barnes in 2012).

 Winning Candidate Vote Share Δ\$/ΔVotes 2008 Walz 62.50% \$16.41 2010 Walz 49.34% \$94.24 2012 Walz 57.52% \$31.98 2008 Kline 57.29% \$15.93 2010 Kline 63.31% \$17.20 2012 Kline 54.01% \$52.71 2008 Paulsen 48.48% \$0.67 2010 Paulsen 58.80% \$33.59 2012 Paulsen 58.10% \$41.11 2008 McCollum 68.44% \$5.48 2010 McCollum 59.09% \$11.51 2012 McCollum 66.29% \$7.58 2008 Ellison 70.88% \$9.01 2010 Ellison 67.69% \$12.11 2012 Ellison 74.47% \$9.81 2008 Bachmann 46.41% \$87.26 2010 Bachmann 52.51% \$228.94 2012 Bachmann 50.47% \$4,351.42 2008 Peterson 72.20% \$7.30 2010 Peterson 55.20% \$20.95 2012 Peterson 60.36% \$6.76 2008 Oberstar 67.69% \$10.98 2010 Cravaack 48.18% -\$362.04 2012 Nolan 54.28% -\$34.45

We can immediately see the impact of swing elections on Minnesota races. In 2010, Democrat Tim Walz’s moderate district forced him to spend roughly \$94 per marginal vote as opposed to \$16 and \$32 dollars for his 2008 and 2012 elections, respectively. Even Democrats such as Betty McCollum and Keith Ellison, who reside in Democrat-safe districts, were forced to spend more in 2010.

The lowest positive SII was that of Republican Erik Paulsen, with \$0.67. A third-party candidate who captured almost 11 percent of the vote countered a strong Democratic challenger’s nearly equal fundraising. Paulsen’s next two elections were more difficult. His 2010 and 2012 reelections had SIIs of \$33 and \$41, respectively.

Over this period we have two negative SIIs, which indicate that the worse-funded opponent was victorious. This occurred in the Eighth District, where the seat changed hands in both 2010 and 2012.

The most fascinating example of an SII is the infamous Michele Bachmann. Bachmann is only marginally popular in the district, and she relies greatly on a fundraising advantage. Her elections’ SIIs of \$87 in 2008 and \$229 in 2010 are insignificant compared to her 2012 SII of \$4,351. Bachmann outspent her opponent, Jim Graves, by almost \$19 million, though she only won by around 4,000 votes.

The SII indicates compares a campaign’s relative strength with its ability to fundraise. For example, had Graves received more funding, he could have won the election instead of Bachmann. Yet one could argue that losing opponents in races with high SIIs helps other races by forcing the opposing party to concentrate funding inefficiently. While Graves lost to Bachmann, he also may have prevented millions of dollars from going to other Republican candidates, which could have helped to swing other districts.

The SII may not be the most electrifying subject, the SII can help to predict and win elections.

photo of Michelle Bachmann by Gage Skidmore: http://www.flickr.com/photos/gageskidmore/5435512618/