Forests and trees
Why do left-leaning independents in North Carolina’s bluest precincts turn out less than Democrats in their neighborhoods when independents statewide (including the Trumpiest counties) turn out more?
Perhaps because Democrats are policy liberals and campaign conservatives. Taking chances? Not when every damned election is “the most important of our lifetimes.” Throwing the long ball? They are more inclined to fall on it. “Leave it all on the field” remains aspirational.
What we’re getting at here is Democrats’ targeting of independents for turnout is too conservative. As Michah L. Sifry (The Connector) in July observed, “The Experience of Grassroots Leaders Working with the Democracy Party” report is sobering. Among the complaints (emphasis mine):
Most volunteer leaders see their state Democratic party’s efforts to organize outreach as “too little, too late.” One in four call their party unresponsive. A majority of respondents said the party does a terrible job targeting voters, saying that its lists are far too narrow.
That is what I find as well:
This is my focus right now. Independents (UNAffiliated voters in NC) are the largest bloc of registered voters in NC: 36% (2.6 million voters). But statewide they voted against Democrats here 58% of the time in the last two elections. Democrats cannot win without them, but their traditional tactics, as Pepper recognizes, focuses only on “the most frequent voters.” This tactic leaves many “removed from the political conversation” in what I’ve dubbed “No Voter’s Land.” These are voters campaigns are reluctant to contact (using the tactics of the last war, you might say) because computer scoring deems them not good bets.
Self-identified independents are also the largest tranche of voters in Arizona, with Republicans second and Democrats third. But we’ll come back to that. In North Carolina it’s 36% Independents, 33% Democrats and 30% Republicans.
A few more rough numbers. Independent registrants turn out statewide at 6.5% below Democrats. Horseshoes and hand grenades, if there are more independents than Democrats, and they vote 58% against Democrats, but turn out 6.5% less, we begin to see why results of statewide races, even with healthy Democratic turnout, are so narrow.
In North Carolina’s bluest precincts in its most urban counties, however, independents turn out at 12% less than Democrats. Why? Are the nonvoters less engaged than their voting independent neighbors? Or are Democrats not engaging them?
Well.
Can’t see the forest for the trees
Democrats’ principle targeting tool is NGP Van’s VoteBuilder database, another source of complaint among the campaignerati. Its very premise is microtargeting. Every voter gets assigned a support score based on a variety of factors. But when it comes to voter outreach targeting independent voters have three strikes (or more) against them.
Strike1: Not a registered Dem.
Strike2: Do they vote in R primaries (allowed in NC)? Looks like an R-leaner. Or they (the vast majority) don’t vote in Democratic primaries and are written off as potential R-leaners.
Strike3: Are they an irregular voter? Lowest priority for voter contact efforts.
When it comes to independent voters, VoteBuilder sees individual trees and misses the forest. The forests campaigns miss are those precincts where independents who turn out vote 60, 70, 80-90% with Democrats but turn out at 12% less than their Democrat neighbors, almost twice what it is in Trump-country counties.
I’ve calculated which precincts those are. VoteBuilder doesn’t. It’s not how it works, or at least it’s not how it’s used in the field. If we could coax those low-scoring independents in specific precincts off their couches, Democrats score. Birds of a feather. They’ll vote like their voting independent neighbors. If their turnout matched Democrats’ in those precincts (not likely), there are an additional 56k votes for Democrats. But if they simply matched independents’ statewide average turnout, we’re still talking an extra 25k.
Arizona Democrats could do the same since their voter rolls also have partisan registration. Likely, the pattern I see in NC is repeated there.
Well?