Sunday, March 13, 2011

2010 Recap Part III: Recalling the Past

I have put off writing the two remaining sections of analysis of the 2010 election in Michigan in part because I am still waiting on voter file data to arrive from the Secretary of State, as well as receiving corrected voter tallies from November. When that does arrive I will do a final write up on results.

However, the political events of the past few weeks in Wisconsin, as well as the forthcoming school district elections in early May in Michigan (along with the municipal general election for blog readers in Texas), have motivated me to write a third post on a rather timely topic: field operations. With labor confident of recalling a number of Republican state senators in Wisconsin, and local campaigns making promises to bring an unprecedented number of voters to the polls in many states, any realistic campaigner needs to throw the old campaign playbook out the window.

In 2004 Donald Green and Alan Gerber came out with a great little book entitled Get Out the Vote: How to Increase Voter Turnout. The book was significantly updated in 2008 with additional case studies, but the basic message remained the same: traditional campaign methods that have developed over the past thirty years are rather worthless. Green and Gerber scientifically analyzed turnout measures used by over 75 campaigns between 1998 and 2007, and determined the cost effectiveness of various turnout methods. Think sending a robocall with Bill Clinton reminding your base voters to vote will increase turnout? Green and Gerber find that robocalls have no effect on increasing turnout. Does doing three hits of direct mail to flood voters’ mailboxes in the last weekend of the campaign send people flying to vote? Once again, there is no evidence of any effect. In the same way, television ads, direct mail, and commercial phone banking has a minimal impact on increasing voter turnout. Give that the television market is very fragmented among basic and cable channels, the impact of advertisement on the airwaves is lost in a world of information overload. Similarly, Facebook and other social media is an excellent way to repeat your campaign message ad nauseam, but this is not going to do anything to increase turnout.

What does seem to work, and is rather cost effective at that, is extensive door-to-door canvassing. The authors calculate that door-to-door canvassing results in one vote per 14 contacts, which averages a voter contact cost of $29. Many campaigns use direct mail to bring the election bell, and this was certainly the case in 2010. For example, in the 2010 general election, State Senate candidate David Hildenbrand spent almost $23,000 on a mass mailing (which included paying for literature, mailing costs, and developing a voter list). If Hidlenbrand had spent this money on door-to-door canvassing, he would have likely would have gained about 650 more votes on Election Day. Given that Hildenbrand narrowly won on November 2, this decision did not adversely impact his efforts. However, for the five Democratic State House candidates who lost by less than 500 votes on the same day, the redirecting of $25,000 to a more effective ground game would have carried them over the top, resulting in a much smaller GOP edge in the State House of 58-52 (as opposed to the current 63 to 49 edge).

If there was any one candidate in particular who took the lessons outlined in the first edition of Get Out the Vote, it was Texas Republican Governor Rick Perry. This article in the Texas Observer is an excellent piece on his campaign’s efforts to develop an extremely effective turnout machine over the past four years, and I recommend that you read it in full. As much as one might disagree with Perry’s priorities and secessionist tendencies, his reelection after thumping his opponents in the Republican primary in June and besting Democrat Bill White in November 2010 is nothing short of amazing after being declared dead by most observers in late 2009. While Perry spent heavily on the airwaves and direct mail, his campaign also invested handsomely in developing a very strong door-to-door campaign game.

The lessons from Get Out the Vote are even more important for races with low turnout that require careful voter targeting. A school board election in Michigan, for example, will, at best, get 20% turnout, with turnout around 15% much more likely. When you run television ads and other paid media, only 20% of the voting population will really be interested. Likewise, direct mail has a horrible record of increasing voter turnout for small turnout campaigns. In the end, through careful use of voter files, a campaign should stick to door-to-door contacting, volunteer phone banking, and run a vigorous absentee voter contact operation. In the potential recall elections in Wisconsin, it is especially important to reach out to voters who typically do not show up at special elections. While I’m not certain that the Democrats will be able to get the needed number signatures to recall the eight Wisconsin State Senators, but if the recall elections do occur, I seriously hope that the state Democratic Party will read Get Out the Vote before wasting much needed funds on television advertising and direct mail and instead resort to old fashion door-to-door campaign. Also, you might want to call someone who does political number crunching and mapping. Just a thought.

No comments: