Blog Post 1.7 "Does Campaigning Work?"

Campaign activities have virtually no effect on those that they attempt to reach according to Kalla and Brockman. They analyzed the 2016 and 2015 elections and reached this conclusion. They estimate that one in eight hundred was persuaded. Campaign actions that occurred some time before the election were more effective than right before elections. When voters are less set on their choice of candidate, such as during the primaries and ballot-initiative phases, campaign activities could have a greater effect.

Canvassing was effective when at least two months remained before a primary but not a general election. Therefore, Brockman and Kalla think that campaigns should try to boost voter turnout instead of trying to persuade and focus their efforts on primaries. For example, in Oregon, information about pro-choice Senate candidate Merkley was sent to pro-choice voters to correct a misconception about Merkley's views and contrast him with his opponent. The result was a three percent increase in Merkley's share of the vote. In another case, research on flyers in North Carolina indicated that flyers had a greater impact on African-American voters, so the distribution of flyers was optimized to reach the most receptive audience. 

Campaigns will often spend about sixty dollars per person they attempt to register and can be ineffective when a candidate is from a party that does not have a large number of existing party members. Persuasion can have a double effect of taking away a vote from one party and giving it to another, essentially affecting two votes instead of one.