Comparison of Redistricting maps by the numbers.

I said earlier today I would work on a report about the introduced Dem map. In the meantime, the GOP map has been introduced as well. See below the jump for a side by side comparison.

First of all, let me explain how I am evaluating the maps.

Each map when introduced includes a spreadsheet of how the vote came in in each of the new proposed districts for 3 past races: the US Senate race in 2010 and in 2008 and the CU Regent at large race in 2010.

The thought behind using these 3 races is to use a big race from each of the “wave” years. 2010 saw a surge in GOP participation and 2008 saw a similar surge on the Dem side. I guess they also included to the CU Regent race because it is lower profile and therefore likely to be a better barometer of more purely partisan voting, where US Senate is likely to include more “voting for the person” mentality.

I would have picked a slightly different set of races for this purpose to try to get to a more neutral guage of party strenth in a district, but no one asked me so I will continue with the same premise they are using.

When you average together the percentage results of each of the 3 races, then subtract the GOP result from the Dem result (or vise versa to guage GOP strength) you are left with a number that is called the Democratic Performance Index or DPI. (Republicans would go the other way and get a RPI). For example a district with a DPI of -3.53 means in a race between two exactly equal candidates, it could be expected to produce the result of the Republican winning by 3.53 percentage points. Of course, nothing is ever that cut and dry, but it is a reasonably accurate measuring tool as long as it is applied consistently.

Here is the DPI for each of the districts on the maps introduced. I am including one for the current districts for the sole purpose of giving you a feeling for how DPI’s relate to actual districts.


Current: +45.06

SB268 (Dem map): +37.43

HB1319 (GOP Map): +39.39


Current: +30.7

SB268: +10.46

HB1319: +15.9


Current: +7.5

SB268: -13.67

HB1319: -4.33


Current: -4.1

SB268: -7.47

HB1319: -8.44


Current: -34.68

SB268: -22.66

HB1319: -28.04


Current: -22.26

SB268: +5.67

HB1319: -10.06


Current: +4.94

SB268: +4.64

HB1319: +3.97

So we can see the only place the two plans have a major conflict in terms of sheer numbers in CD6. Both plans have about the same degree of competiveness except that one district. In fact, the GOP has a more cometitive CD3 than the Dems.

7 Community Comments, Facebook Comments

  1. Just Anita says:

    At the very least you are consistent with trying to marginalize unaffiliated voters in the redistricting process – a group that is nearly 1/3 of registered voters in Colorado. The DPI assumes two voter affiliations instead of three with the third having it’s own theoretical DPI not used in your analysis. Your stats presume voters are static. Further, the “static” data being used is incomplete. The only point not flawed in your analysis is the statement, “Nothing is ever that cut and dry.”

    The Democrats have presented a map that includes compactness, common shared interests based on census population demographics and economic drivers as related to each other statewide, and national interests – reminder these are Congressional Districts – as a whole. As a bonus the Democrats map creates 7 competitive districts, because registered voters – Republicans, Democrats, and unaffiliaters – are disbursed effectively based on regional interests.

    Competitive is strictly a partisan term. Why aren’t you using tangible data – like census statistics and demographics – to qualify compactness and common interests within both maps? Instead you offer intangible data – DPI probability and statistics to prove/disprove competiveness; framing the discussion about redistricting into party controlled divides within the districts. Coincidence?

    The Republicans have not made any public statments qualifying their maps as having any connection to compactness and common shared interests with their map beyond their own limited scope of stacking the odds. And, splitting CD3 really rains on their parade based on the lines redrawn in the metro area.  

    • Dan Willis says:

      And that conversation was about partisan competiveness.

      That does not detract from any of the other issues you bring up, it is just they are a different part of the conversation.

      And actually the analysis does take into account the unaffiliated voter and is based on tangible date: vote outcomes. In most congressional races there are only two choices, a Democrat or a Republican. Using a DPI/RPI process measures, to the extent possible, the leanings of all of the voters, including the unaffiliated ones, to one side or the other. Of course there are many other factors that determine the final result of any given election, but for the purposes of redistricting, there has to be some mathematical basis of understanding as a foundation.

      And bottom line, for the legislators who are voting on the bills, partisanship is typically their prime motivator. That part is sad, but it is the way it is.

      • Ralphie says:

        It certainly blows Heath’s “competitive districts” talking point out of the water.

        I use a slightly different method to calculate DPI, which compares D and R performance in CU Regent at Large and other low-profile statewide elections.  That tells me how the unaffiliateds vote.

        But getting statewide precinct-level election results might be outside of your budget and/or your available time; to do it right, at the Census Block or even Block Group level, the data are simply unavailable.

        Still, a worthwhile and informative exercise.

  2. Go Raiders* says:

    Can you provide the DPI/RPI figures for your map?

    I am interested in seeing if yours is a balance between the two proposals.

    • Dan Willis says:

      My map was heavy D in CD’s 1 and 2 and R in CD’s 4 and 5 (I didn’t calculate the exact DPI but it would have been more than 15)

      For the competitive ones:

      CD3: -5.23

      CD6: -2.65

      CD7: +2.73

      I am no longer pushing for my map, and supporting HB1319 instead

Leave a Reply

Comment from your Facebook account

You may comment with your Colorado Pols account above (click here to register), or via Facebook below.