Dan Payne’s analysis of the Presidential race in today’s Boston Globe illustrates why he was a bad fit for the Deval Patrick campaign, which he left soon before Deval blew the lid off the 2006 Democratic gubernatorial primaries. Payne repeatedly cites, while also chastising himself for citing, poll numbers without any serious analysis of the correlation between pre-election polling and final results.
To analogize between a political campaign and a military one, tactics like endorsements and advertising are like long-range bombing: all they do is “soften up” the populace and provide the potential for votes. But warplanes and artillery do not capture territory: for that you need “boots on the ground,” which in the electoral context means real people making phone calls and canvassing door-to-door to convert general support into real votes.
Political analysts like Dan Payne are biased towards covering the “air war” because it’s sexy and easy to see. But a more accurate way to interpret poll data is to weight them by the presence of ground troops. Sudden shifts in popular numbers are unlikely to show up in real votes without a large and well-organized volunteer base to realize those gains. (The analysis is somewhat different when the numbers are static, in which case the leader will win, all else being equal). As I’d written earlier, Obama pulled the organizational structure out of Massachusetts, and Deval Patrick’s supporters just couldn’t cover the ground fast enough to capitalize on the sudden shift in public sentiment.
