Today the Yes on 19 campaign released a new TV ad, featuring former San Jose Police Chief Joseph McNamara:
Recent survey data, which show support for Prop. 19 slipping, suggest it can use whatever boost ex-cops like McNamara can give it. In a survey (PDF) conducted October 10-17, the Public Policy Institute of California found 44 percent in favor of the initiative (an eight-point drop in support since September) and 49 percent against, with 7 percent undecided. A Los Angeles Times/USC poll conducted October 13-20 showed 39 percent in favor, 51 percent against.
By contrast, a campaign-commissioned poll (PDF) conducted October 13-14 found 54 percent support for the initiative in automated interviews, compared to only 36 percent in live interviews (for an average of 45 percent). This divergence jibes with poll maven Nate Silver's hypothesis that people may be reluctant to endorse marijuana legalization when they're asked about it by live humans. The initiative's backers note that SurveyUSA, which has consistently found more support than opposition for Prop. 19 (though the gap has narrowed recently), uses automated polling. Its latest results (from October 15-18) show 48 percent in favor, 44 percent against.
Last week Silver summarized the recent polling data:
Calculating a trendline from all surveys on the initiative suggests that about 46 percent of Californians plan to vote in favor of Proposition 19…but 47 percent plan to vote no. This reflects a reversal from before, as the ballot measure had led in most surveys prior to this month….
There are good reasons to think the polls could either be overestimating or underestimating Proposition 19's support. In spite of the recent trends against Proposition 19, therefore, I would be inclined to take the recent polling at face value, which suggests that the measure has about even odds of passing.
McNamara gives three reasons to support Prop. 19 in the December issue of Reason. An online Yes on 19 ad, like the spot featuring McNamara, plays up the "drug dealers don't card" angle.