Since all things being equal a Democrat should always beat a Republican in Massachusetts, each of the MassPoliticsProfs had varying levels of bemusement at the Democrats’ inability to field a decent candidate against Senator Scott Brown until his Republican senatorial colleagues gift-wrapped Elizabeth Warren for the Democrats. Talk about your random acts of kindness!
The things we know we know should favor the Democrats in the upcoming special senate election — right? And the known-unknowns may but need not favor a Democrat, and the unknown-unkowns – like 2010’s Black Swan event – may upset all expectations. I’ve gotten very interested in the things we don’t see coming.
A new “Features Symposium: Recap: Forecasting the 2012 Election” in PS: Political Science and Politics shows the utility of forecasting models in presidential elections but political scientist Carl Klarner did not have great success with senate elections: “US Senate elections appear to be influenced by race-specific factors that are difficult to include in forecasting models.”
One of the things we think we know is that if turnout is high (as it was in the presidential election year of 2012) then Democrats should win. Moreover because of the Democrats superior field organization capacity they should be able to influence turnout more than the Republicans can and thus have a clear advantage.
But the 2010 special election wasn’t a low turnout race – in fact the turnout was similar to turnout in the 2002, 2006 and 2010 gubernatorial elections.
My UMB colleagues Tom Ferguson and Jie Chen wrote a Roosevelt Institute Working Paper that concluded that while overall turnout in the special was in line with non-presidential general elections, turnout dropped in lower-income communities. Ferguson and Chen argue that the Obama administration’s passive response to high unemployment and default in protecting families during the mortgage foreclosure crisis damaged the Democrats with their core constituencies.
In a separate study political scientists Stephen Ansolabehere and Charles Stewart III noted a surge across the state for Scott Brown with a significant turnout drop in core Democratic areas – minority and low-income communities. They wrote that “The pattern is eye-popping—one is tempted to conclude that the results of the special election were primarily determined by the demobilization of the Democratic base: poor, minority voters in liberal parts of the state.”
Moreover, Ansolabehere and Stewart found that “the aggregated turnout patterns indicate clearly that Brown’s ‘ground game’ of delivering supporters to the polls was a first-class effort, far outstripping his competitor’s.”
So perhaps turnout and organizational advantage are not entirely the known-knowns we think they are. But what if minority and low-income Democrats had turned out as we would expect?
As a counterfactual Ansolabehere and Stewart calculated what would have happened if turnout in the poorest half of the state’s communities had matched that in the upper half - Brown would still have easily defeated Coakley.
Now what would have been the result if the 2010 special election turnout mirrored the turnout of the 2008 presidential election? “Assume that Coakley’s and Brown’s town-by-town turnouts are unchanged, but that each town’s turnout equaled the 2008 figures. Brown would have won just under 51 percent of the vote and Coakley just over 48 percent.”
Don’t forget until about ten days before Election Day very few of us outside of the Brown campaign saw him having any chance at all against the formidable Democratic Attorney General.
And we’ve hardly gotten to the known-unknowns or unknown-unknowns yet.