The McMahon Surge is Real

In the last two weeks, polling in the Connecticut senate race has shown Republican Linda McMahon leading Democrat Chris Murphy by 3 point, 49% to 46%. The polls, by Rasmussen and Quinnipiac, were shocking to most observers. Connecticut is a blue state in a blue region. Could Connecticut voters really be leaning towards selecting a Republican as a United States senator? It seemed like an historic shift, but the polling firms involved have decent track records, and so it seems that we may indeed see senator McMahon sworn in next January.

Today, there is further proof that the surge in voters for Republican McMahon is real. Public Policy Polling is out with a poll of the race that shows Democrat Murphy ahead by 48 to 44 percent. You may wonder how a poll showing Murphy in the lead is proof of a surge to McMahon, so let me explain. First of all, Public Policy Polling has a track record for its polls always overperforming for the Democrat. The differences in some states are as large as seven percent better results for the Democrats in a Public Policy Polling effort compared to the average of all other polls. That means that a 48 to 44 margin for Murphy probably translate to at best a tie for Murphy in a poll by an unbiased pollster, and more likely, a McMahon lead. Second, Public Policy polled registered voters for this poll rather than likely voters. In general and particularly in Connecticut, registered voters poll about two or three percent more Democrat than likely voters. In other words, an unbiased poll of likely voters would show McMahon in the lead by a margin much like the one in the other two polls. Third comes the clincher. I realize that many folks will look at this matter with cynicism. Did Public Policy Polling actually favor Murphy, they will say? Here is what you need to know. One month ago, Public Policy Polling did a poll prior to the McMahon surge; they poll gave results for likely voters. Then, McMahon moves ahead of Murphy, and the pollster switches back to registered voters. Public Policy Polling made a switch in methodology that it knew would add two or three points to Murphy's numbers and subtract a like amount from McMahon's. Clearly, the pollster determined that without the switch to registered voters, McMahon would have a third poll showing her in the lead.