Gerrymandering: In Arizona, the “Efficiency Gap” is Too Unstable a Measure

From the Arizona Capitol Times, Guest Opinion, August 23, 2018: Gerrymandering: In Arizona, the “Efficiency Gap” is Too Unstable a Measure

By Tony Sissons

Much in the news lately, the U. S. Supreme Court is having a difficult time determining how much is too much partisan gerrymandering.  When does a redistricting map favor a political party to an unfair degree? How do you measure unfairness?

How do you measure unfairness?

Just looking at a map, you really can’t be sure. Some might say it’s just like pornography – you know it when you see it. But, the accepted definition of a gerrymander is: a practice intended to establish a political advantage for a particular party or group by manipulating district boundaries.  Notice there is no mention of ‘shape’.

So, what’s a Court to do?  In 1986, the U. S. Supreme Court ruled that partisan gerrymandering is judiciable, but in the thirty-two years since, the Court has not struck down a single plan on that basis.  

In a movement away from geometry and geography, Nicholas O. Stephanopoulos, Assistant Professor of Law at the University of Chicago Law School, and Eric M. McGhee, Research Fellow at the Public Policy Institute of California, proposed a method that measures the relative success of the major political parties in a district plan by comparing the numbers of votes ‘wasted’ by each party.  The measure is called the Efficiency Gap, and is expressed in terms of the numerical difference between both party’s wasted votes, taken as a percentage of total votes cast.

In this calculation, all the votes for the losing candidate in a district are considered wasted, as are the votes for the winning candidate in excess of the votes needed to win.  Both categories of wasted votes for each party are summed for all of the districts in the plan, and then the difference between those party totals is taken as a percentage of the total number of votes cast by both parties.  The party wasting the least number of votes is considered to have the advantage (for that plan for that election).

The clause in parentheses triggers my concern.  I created a spreadsheet model to simplify measuring the Efficiency Gap for Arizona congressional districts since 2002.  Here are the Efficiency Gaps and who they favor:


  • 2002: 17.73% Republican

  • 2004: 16.78% Republican

  • 2006: 1.41% Republican

  • 2008: 4.30% Democrat

  • 2010: 6.70% Republican


  • 2012: 11.34% Democrat

  • 2014: 0.20% Republican

  • 2016: 4.54% Republican

The figures in the left column are two-party results on the congressional map created by the first Arizona Independent Redistricting Commission (AIRC). On that map, the Efficiency Gap ranges from a seventeen percent Republican advantage to a four percent Democratic advantage. On the same district shapes! Likewise, the districts drawn by the second AIRC swing from a four percent Republican advantage to an eleven percent Democratic advantage. Which Efficiency Gap figure is the correct one for each decade’s plan?

There is one other circumstance that casts an even greater cloud on the measure’s credibility.  The results from a single highly competitive contest can really mess up the measurement. Consider:  In 2014, Republican Martha McSally beat Democrat Ron Barber by 161 votes out of 219,247 votes cast between them.  On the basis of that result, the Efficiency Gap for Arizona’s nine congressional districts was a 0.2% advantage for Republicans.  But, if McSally had received 81 fewer votes and Barber 81 votes more, the Efficiency Gap on that same map would come in with a 15.5% Democratic advantage.

With such measurement instability, I don’t think the Efficiency Gap measurement can be considered a reliable indicator of the extent of partisan gerrymandering of Arizona election districts.

Tony Sissons is president and owner of Research Advisory Services, a Phoenix public policy and demographic research consulting firm.