Studies and control group experiments show that political ads have small effects and sometimes move no voters at all. Reinventing Political Advertising reviews measurement of the small impacts of political ads on television, the internet and in the mail.

Drawing on testing data, the book shows how these small effects can be increased by new approaches to both targeting and creative design. One example is a series of studies showing that well informed voters are hardly moved by advertising at all but low information voters are a much better target. In one experiment, voters in the treatment and control group were asked, in addition to their candidate choice, which party controlled congress. Those voters who did not know were thirty-one times more likely to change their candidate support than those who knew,. Despite these findings political advertisers place more than half of all television ads on news programming.

When 90% of voters are voting straight party tickets, why are almost 100% of political ads about candidates and none about parties. Breakthrough mail experiments show that providing voters with impartial information about candidate votes on issues, with no endorsement at all, works better than ads telling voters how to vote.

Hal Malchow reviews these and other issues and recommends an agenda to help Democrats improve the effectiveness of political ads and win more elections.