Increasing political participation was the goal of a canvassing program implemented during the 2000 presidential election. The experimental evaluation was based on random voter mobilization experiments geared to those under 30 years old living near college campuses in New York, Colorado, and Oregon (Green & Gerber, 2001). Although this experiment focuses on an age-group generally older than the focus of this report, we believe that it is important to understand ways to increase political participation among a relatively young group; especially considering the low-level of interest that political candidates take in issue that are important to younger adults. The researchers found that phone-to-phone canvassing produced significant increases in voter turnout rates (five percentage points) in New York, but the program did not have any significant impacts in either Oregon or Colorado. However, when the authors aggregated the results across the sites, they found an overall significant treatment effect of a five percentage-point increase in voter turnout. It should be noted that this was a relatively conservative evaluation, because a successful contact included both actual conversations with the target youth or young adult, but also if a voice message was left on the answering machine.

The researchers also evaluated a face-to-face canvassing program in Oregon. They found that face-to-face canvassing produces a significantly greater rate (8.5 percentage-points) of voter turnout. Interestingly, through non-experimental analysis of others living in the participant's home, the researchers found that for every 100 treatment participants in the canvassing experiment, there was a total of 8.4 additional votes mobilized through what the researchers considered to be a "spill-over effect." Overall, the researchers conclude that 20 successful phone contacts translate into one additional vote and 12 successful face-to-face contacts translate into one additional vote. Although these results have great implications for political strategists, the results do not address the stability of political behavior. More long-term follow-ups are necessary to determine if the increase in voting behavior is a one-time effect or if it represents an actual change in young adults' long-term attitudes and behaviors.