Ins and Outs of Moving a District


The term, “moving a district”, is a bit misleading. State Senate and House Districts don’t actually move. The population moves and districts just go chasing after them.

Think of the changing number of congressional districts within a state. We don’t say that they moved a seat from Pennsylvania to another state. We simply view it as Pennsylvania losing a congressional seat and other states gaining ones. There is no moving involved.

In a time not so long ago, we had no statewide district numbering system. It was all done within the county. Every 10 years, each county was reassigned a certain number of districts based on the new population. This meant counties just gained or lost seats, district numbers did not actually move.

But then states started to run afoul of the equal population requirements. 1968 was the dawn of a new era in district numbering. After the statewide district number assignment, those numbers became a fixture in certain counties and regions. Unintentionally, it now fosters the perception that districts move. The accompanying political hype enforces the idea that “moving a district” is a big deal.

My system for reapportionment avoided the issue by not assigning district numbers until after all the boundaries had been drawn. Then I sort of overlaid the 2001 district boundaries and compared them with the new boundaries to see where the district numbers fell. Some areas had more numbers than districts, others not enough.

Naturally, it all balanced out by collecting the extra district numbers from the areas that lost population and reassigning them to the areas that had gained population. My findings showed three instances of this in the Senate and six in the House.

But are there impartial clues that would indicate which areas will lose or gain a district, before redrawing the map?

Impartial Indicators

Looking back, it seems like there are two possible means for identifying where the population moved out and where it moved in. The result will reveal the likeliest places for district gains and losses.

The House and Method 1

With 203 House district boundaries, the simplest and quickest way is to compare House county apportionment with the number of representatives living within that county. This assumes that the house district is mostly in the county where the representative lives, which is often true.

County House Reps Difference
Berks 6.58 5 +1.58
Chester 7.97 7 +0.97
York 6.95 6 +0.95
Butler 2.94 2 +0.94
Wayne 0.84 0 +0.84
Montgomery 12.78 12 +0.78
Cumberland 3.76 3 +0.76
Mifflin 0.75 0 +0.75
Union 0.72 0 +0.72
Monroe 2.71 2 +0.71
Clearfield 1.30 2 -0.70
Cambria 2.30 3 -0.70
Luzerne 5.13 6 -0.87
Beaver 2.73 4 -1.27
Allegheny 19.55 21 -1.45
Crawford 1.42 3 -1.58
Philadelphia 24.39 26 -1.61


Anything outside of a +/- .7 difference will likely either lose/gain a district. In the gain department, it might also mean that the Representative lives in a neighboring, smaller county – as is the case in the Juniata and Mifflin area, for instance.

This also works for the City of Pittsburgh (the only other city really large enough to use an apportionment gage). The city apportionment is 4.9 but is currently has 8 Representatives. The suburbs (which form the rest of Allegheny County) have an apportionment of 14.6 but are only represented by 13. What does this indicate? That the house district lost in Allegheny County will likely be in the city of Pittsburgh, while another two districts from the city will shift into the suburbs.

The Senate and Method 1

Because Senate districts are larger in size, the first method only works for the larger counties –particularly Allegheny and Philadelphia.

Counties Apportionment Senators Equals
Allegheny 4.83 -6 -1.17
Philadelphia 6.01 -7 -0.99
Berks 1.62 -1 +0.62
Monroe 0.67 0 +0.67
York 1.71 -1 +0.71
Chester 1.96 -1 +0.96


The Senate and Method 2

Another method works better for identifying population gains and loss related to Senate Districts. It divides the Senate districts into regions – like Southwest, Northwest, South Central, Southeast, and Northeast. Then we take the number of Senators in that region; multiply it by the target district size, and minus the total from the total population in those regions.

If a region comes up over 50% shy of the population needed to compose a district, that region will likely lose a seat. If it is over 50% above, then it will likely be gaining a district.

The Census Bureau already has a list of the 2001 district boundaries with the 2010 census data inside. Using this chart and a 2001 map, we put the districts into the different regions. Each ended up containing 7 districts. Here are the population results:

Region Over/Under Percentage
South East +117541 +46.27%
North East +138226 +54.41%
South Central +145279 +57.19%
South West -176186 -69.35%
North West -137293 -54.04%


127,024 persons = half of the target district size. Based on the above results, Northwest and Southwest will each lose one district while the South Central and Northeast will each gain one district.

From our first test, we already know that the Southwest district will be coming from Allegheny Couth. We also know that there will be a shift in the Southeast evidenced by the findings in Philadelphia.

Because the counties in the Northeast are so small, it is impossible to identify which county will move. Looking at apportionment numbers in the South Central, York seems a likely candidate. In the Northeast, the apportionment numbers lean toward either Monroe or Berks County but it is not conclusive.


Taking all of these numbers into account, the evidence points to the following conclusions:

  • 3 counties will lose a Senate seat (Allegheny, Philadelphia, and one in the Northwest).
  • 3 counties will gain a Senate seat (Chester, York and likely Berks or Monroe).
  • 6 counties will lose a House seat (Allegheny, Beaver, Cambria/Clearfield, Crawford, Luzerne, and Philadelphia).
  • 6 counties will gain a House seat (Berks, Butler, Chester, Cumberland, Monroe, and York).

Commission plan results:

  • 1 county lost a Senate seat (Allegheny).
  • 1 county gained a Senate seat (Monroe).
  • 4 counties lost a House seat (Allegheny, Crawford, Lackawanna, and Philadelphia – Allegheny lost two seats).
  • 5 counties gained a House seat (Berks, Chester, Lehigh, Monroe, and York).

Holt plan results:

  • 3 counties lost a Senate seat (Allegheny, Philadelphia, and one in the Northwest – Venango).
  • 3 counties gained a Senate seat (Chester, Adams/York and Berks or Monroe – it could be done either way with the same number of divisions).
  • 6 counties lost a House seat (Allegheny, Beaver, Cambria/Clearfield, Crawford, Luzerne, and Philadelphia).
  • 5 counties gained a House seat (Berks, Chester, Cumberland, Monroe, and York – Berks got two because one shared a border with the Lancaster County population growth).

As the impartial system shows, it is the population that is really doing the moving. District numbers just tag along for the ride.



  1. Hi Amanda,
    I just wanted to thank you for the excellent work you’ve done creating a PA redistricting map that meet the PA constitutional requirements, makes sense and is apolitical. I’ve never in my life understood anything less than why the commissioners can’t jut accept your map and move on (no I’m not from Mars I do understand they can’t accept it because it doesn’t favor the Republican’s in office and running for office). They are wasting so much tax payer money continuing the whole 3 ring circus over this issue.

    I’ve written the commission several times on this and called some of them too but they couldn’t care less what us citizens want.

    Thanks again for showing how well a Pa legislative district map can be drawn when politics are removed from the process. I salute you.

  2. Dear Amanda –

    Just wanted to thank you for putting in so much work into fighting these obviously biased divisions against not just the people of the Lehigh Valley but all of Pennsylvania. Like I put it in a letter to the editor, it was as if a bunch toddlers went jack’o’lantern carving at the state map with a chainsaw.

    I ran for US Congress in the Lehigh Valley as an independent in 2010 and competed against both parties. Under the new redistricting lines, it is much less possible to defeat the incumbents – half the old district (15) went to the GOP and the other half (17) to a Dem.

    Gerrymandering is election-rigging, pure and simple. I had thought that the Supreme Court would pass over the PA constitution’s pretty obvious “Unless absolutely necessary no county, city, incorporated town, borough, township or ward shall be divided in forming either a senatorial or representative district.” but your efforts happily proved me wrong.

    Your efforts give me more hope for the future,

Leave a Reply to Jake Cancel reply