The latest buzz in the Arizona Senate race is a poll from OH Predictive Insights purporting to show 15 percent of the voters supporting Libertarian candidate Marc Victor; the poll also has Mark Kelly up 13 points on Blake Masters, 46 percent to 33 percent. The poll surveyed 674 likely voters. Don’t fall for it.
Set aside the headline result: There is simply no way that Victor will get 15 percent of the vote. Third-party candidates rarely do; they almost always underperform their polling, because their support comes either from people who end up not voting, or from people who decide in the end to vote for one of the two contenders with a real shot. The exception is those third-party candidates who have formed a critical mass of support that shows them consistently in double digits.
Victor does not have that. Not every poll is even asking about him, but most of the public polls in Arizona are inconsistent with Victor’s drawing this kind of support. Look at the recent polls of the state. Arizona Republic/Suffolk had Victor at 1.8 percent, Data for Progress had Victor at 2 percent, Trafalgar had Victor at 2.7 percent. Among the polls that did not name Victor as an option, CBS/YouGov had “someone else” at 0 percent, CNN had “neither” at 3 percent and “other” at 1 percent, Fox News had “other” at 3 percent, Marist had “another party’s candidate” at 0 percent. For some reason, OH Predictive Insights has been the outlier, but even its prior polls showed Victor declining from 10 percent to 4 percent to 3 percent before rising to 6 percent and then, abruptly, 15 percent. Also, its polling sample is comparatively small: If you look only at the three polls in the RealClearPolitics average with a thousand or more respondents, Kelly’s lead over Masters is 3 percent (in CBS/YouGov), 5 percent (in Marist) or 2 percent (in Trafalgar), which averages out to a 3.3 percent Kelly lead, narrower than the average of 4.1 percent and far lower than 13 points.
I’m not buying it. This is a tight race.