Monday, December 17, 2012

Presidential election results by congressional district.

Pic of 2008 Presidential election results by district. Courtesy of CQ Politics, 2009.

Cook Political Report editor David Wasserman recently tweeted the following:
Of course, what David is referring to is the "congressional district method" of awarding electoral college votes. Forty-eight of the fifty states award their electoral votes based on the winner of their state at-large. Maine and Nebraska award delegates according to the winner of each individual congressional district, awarding two bonus electoral votes to the statewide popular vote winner. Wasserman's tweet struck me as intriguing given that Romney lost the national popular vote by a fair margin of 3.7 points. How could this be? The most obvious answer is gerrymandering, the process by which political parties manipulate the geographic congressional boundaries for political gain. But the bizarre result still had me wondering how the gerrymandering process would have effected presidential results throughout history, in the event all 50 states adopted the Maine/Nebraska EV rule.

Below is a chart comparing the actual electoral vote to what the vote would have been had all states adopted the Nebraska/Maine method of apportioning electoral votes:


1960-2000 results provided by polidata, here. 2004-2008 results compiled using www.uselectionatlas.org.

As the fourth column of the chart indicates, the 1990 Census allowed Republicans to snatch away from Democrats their long held advantage in congressional district redrawing. The Republican gerrymander advantage becomes even more apparent following the 2000 and 2010 redrawings. The chart also indicates that only twice would the result by congressional district have effected the overall outcome of the race: in 2012 and 1960. Democrats won both of those races, but Republicans would have won under the Maine/Nebraska plan. In 1976, both Carter and President Ford would have won exactly 269 congressional votes apiece.

3 comments:

  1. How did you get the pre-1980 data from Polidata? I can't find that anywhere on their site.

    ReplyDelete
  2. Sorry for the late reply on this. The oldest data comes from this link: http://www.polidata.org/press/wpr1c20z.pdf Chart starts on page 8.

    ReplyDelete
  3. Great site! I am loving it!! Will come back again. I am taking your feeds also

    ReplyDelete