In any political campaign, there is one ultimate measure of success: did your candidate get more votes than your opponent(s) ? To accomplish this goal, staffers/consultants are hired to perform various tasks like media, communications, strategy, polling, and the like.
Given that any campaign has a finite end date (i.e., Election Day), it is not possible or reasonable to attempt to reach 100% of the voters. For one thing, voters have their own way of looking at the world that may or may not mesh with your candidate. Voters also have varying levels of motivation when it comes to participating in elections.
Given that voters have varying (and in the opinion of the author, fairly predictable) levels of electoral participation, voter targeting (i.e., focusing on a predefined subset of the electorate) can and should be deployed to take advantage of this behavior.
While voter turnout can vary depending on the election (Presidential, statewide, special elections, tax elections, and the like), individual voter participation is consistent enough for us to place a voter in one of several participation “buckets.”
To test our belief that voter behavior is fairly predictable, we examined the electoral participation history of Louisiana voters between November 2008 and December 2012. From examining this history, we believe that voters can be classified as follows:
- Unlikely voters – 13% of the Louisiana electorate has never cast a ballot, and another 8% last voted before the 2008 Presidential election (there is actually a registered voter who last voted in 1966 !). This voter bloc of 616K, regardless of the election involved, is not worth the attention of a political campaign, with one possible exception: there are 26K voters statewide within this “bucket” who registered to vote after the books closed for the 2012 election. They may possibly be interested only in the 2016 Presidential (or the 2014/2015 statewide elections), but we have no way of knowing that until 2014/2015/2016;
- Occasional voters – Another 24% (or 684K) of the Louisiana electorate only shows up to vote for Presidential contest(s). In other words, they voted in 2008, 2012 (or both), but did not bother to cast a ballot in any statewide, local, or special election. These voters almost certainly would skip a local election, although they could be persuaded to vote in a “high wattage” statewide election for Governor or Senator;
- Likely voters – Another 20% (or 595K) of the Louisiana electorate only shows up to vote for Presidential and statewide contest(s), but NOT local/special/tax elections. Since these voters tend to fall in the middle of the participation spectrum, they certainly can be considered legitimate voters;
- Chronic voters – Another 31% (or 892K) of the Louisiana electorate has cast ballots in at least one Presidential, statewide, AND local elections between 2008 and 2012. These are the “prized voters”, since they have displayed an interest in different types of elections.
These four types of voters cover, in the aggregate, 96% of the Louisiana electorate (the voting behavior of the remaining 4% is not as easy to classify). This predictability makes voter targeting easy for a political campaign depending on the nature of the race being run.
Given that we believe that over 95% of voters display some predictability in their voting patterns, there is an obvious benefit to a political campaign: voter targeting based on their “election behavior” enables a campaign to best use its time and financial resources. To illustrate: if a campaign is being run in an election where only chronics are likely to show up, a campaign can essentially write off 70% of the electorate – and that’s even before we get into the ideological preferences of voting precincts/individual voters. It also means that “data people” (i.e., those who know how to query voter files/databases to find those likely to participate in a local/special, statewide, or Presidential election) now have a role in the campaign of equal importance to those involved in “front office” campaign work.