News / Blog

The Data on Winning by Voter Canvassing

Every campaign guru will say a candidate needs to be knocking on doors to drum up support for your cause and help identify with voters, but does it really work? Furthermore, even if it does work, is it worth a candidate’s time? According to a paper by University of Alabama professor Dr. George Hawley released by Voter Gravity, it does make a big difference — actually, it can make THE difference. He writes:

Donald Green and Alan Gerber are the most renowned scholars of campaign techniques. In a 2000 study, they estimated that face-to-face voter mobilization increases voter turnout by 53 percent among those canvassed in a local election. These results are congruent with older studies, such as those conducted by Rosenstone and Hansen and Verba, Schlozman, and Brady In their analysis of all the major studies conducted on voter canvassing, Green and Gerber found that the overwhelming majority of all research on the subject indicates that voter canvassing boosts turnout. Based on their thorough examination of all the relevant research, they concluded that one additional vote is generated for every fourteen voters that canvassers contact. In a tight race, effective voter contact can make the difference between victory and defeat.

So, clearly it makes a noticeable, impactful difference. But does that mean that a candidate should spend time and money on canvassing instead of using volunteers and money in other areas? Not if part of your canvassing plan includes modern technology. Dr. Hawley explains that knocking on doors of voters who have already determined who they are going to vote for is, in fact, a waste of time. This is why the integration of voter identification and likely voting tendencies is so vital. He summarizes:

In a close race, effective canvassing can make the difference between victory and defeat. However, in a world of limited time, money, and volunteers, you need to target your canvassing efforts on those who can be persuaded to vote for your candidate. Because we are dealing with human beings, there is always a stochastic element – until we directly ask them, we do not know for sure if a person is planning to vote, for whom they plan to vote, or whether they can be persuaded. However, we now know enough about turnout and vote choice to make reasonable decisions regarding whom to target, and possess the technology to put that knowledge to work.

Please find Dr. Hawley’s full analysis and results here.