There is a special election today in Connecticut State Representative District #121 in Stratford, to fill the vacancy caused by the passing of longtime State Rep. Terry Backer (D), who had served this area since 1992. The Democratic Party nominated current Stratford councilman Joe Gresko, while Republicans nominated former Stratford Board of Education member Susan Barksdale. The current partisan composition of the Connecticut House of Representatives is 86 Democrats and 64 Republicans.
As DailyKos points out, this district voted 70% for Obama in 2012, and they regularly re-elected the late Rep. Backer by even larger margins. But my quick analysis says that this may not be the easy Democratic win that it initially appears to be.
To start I broke down voter turnout at each of the district’s four main polling places, using data from last November’s 2015 Municipal Elections:
However, a voter turnout rate of 29% would be unheard of for a state legislative special election. Special elections historically see turnout rates of between 10 and 20%, so I adjusted the overall total turnout rate down to 16%, while keeping the relative share from each precinct the same. That produced a forecast of total votes cast by polling place:
The problem for Democrats is that neither President Obama, nor late Rep. Backer, is on the ballot. So, not only is turnout likely to be low, there is no prominent incumbent who benefits from name recognition and the like. Therefore, I forecast what today’s vote might look like if I applied the party voting percentages by polling place from the 2010 Governor’s race (where Democrat Dan Malloy narrowly defeated Tom Foley) to the projected number of votes cast:
This illustrates the impact of lower turnout. Governor Malloy carried this district with 61% of the vote in 2010, but even when I keep the ratios and percentages the same, and just lower the number of votes cast, it become a much closer race.
I became curious, then, as to how much of a swing in partisan voting would it take for the Republicans to win the seat under these assumptions? The answer, as it turns out, is about 7%:
To create a 50-50 race, all I did was adjust the partisan vote percentages across the board. And while that seems like a lot, remember that I’ve kept the relative turnout constant between different voting locations. In all likelihood, that 7% change could come about by having a GOP-friendly polling place, like Ms. Barksdale’s home area around Lordship school, have a slightly higher turnout combined with vote totals that slightly favor her. Or any combination of factors.
I will be curious to see how close the actual voter turnout percentages and totals are to these projected numbers. And I do think that a 7% swing would be pretty significant. But this illustrates how easily a district that the President carried with 70% of the vote four years ago can elect a Republican State House member today.