Step 2a: Create Regions
Taking a printout of the step 1 map, I literally used pencil and paper to start grouping counties into regions.
I began by selecting the counties that had a sufficient population to add up to a whole number without the need to pull population from a surrounding county. These included Allegheny, Chester, Dauphin, Lancaster, and Philadelphia.
Next, I tried to group counties into regions that required 1 Senator without splitting up a county. Greene and Washington, for instance, quickly stood out as counties fitting this criterion.
The circling continued as it became apparent that some county splits would be unavoidable (the reason for the absolutely necessary clause). It took several tries to end up with population total within the correct range, but the efforts were finally successful.
You can actually try this out yourself. The Senate is a great place to start. Just print out the map from Step 1 and start circling! See what regions you create.
The green squares indicate that no county lines are split. Some Counties — like Allegheny and Philadelphia — may require splits within their borders,
The blue squares indicate one county in the group must be split. It is obvious which county must be split in the region because of population or keeping the district contiguous. For instance, size is the factor that caused Erie and Westmoreland (they were both too big, so the split would have to occur there). And the contiguous reason is behind the splits in Schuylkill, Monroe, and Luzerne.
The white squares indicate that there is more flexibility regarding how the split occurs. We might call them danger zones — where there is danger of ignoring the rules if we are not careful. There were three regions meeting that description.
Step 2b: Verify Population Within Allowed Range
The chart below shows the specifics of how each region added up.
| Average Region Size | % | # of seats | Appor-tion-ment | Total | County |
| 268,100 | 2.11% | 1 | 1.06 | 268,100 | Dauphin |
| 265,010 | 2.09% | 1 | 1.04 | 265,010 | Centre, Juniata, Mifflin, Snyder |
| 263,322 | 2.07% | 1 | 1.04 | 263,322 | Clinton, Lycoming, Potter, Tioga, Cameron, McKean |
| 261,647 | 2.06% | 1 | 1.03 | 261,647 | Beaver, Lawrence |
| 260,387 | 2.05% | 1 | 1.02 | 260,387 | Crawford, Mercer, Venango |
| 259,723 | 2.04% | 2 | 2.04 | 519,445 | Lancaster |
| 257,793 | 2.03% | 1 | 1.01 | 257,793 | Lackawanna, Susquehanna |
| 257,267 | 2.03% | 1 | 1.02 | 257,267 | Clearfield, Cambria, Elk |
| 257,262 | 2.03% | 3 | 3.04 | 771,785 | York, Adams, Cumberland |
| 256,345 | 2.02% | 1 | 1.01 | 256,345 | Perry, Huntingdon, Franklin, Fulton |
| 254,836 | 2.01% | 1 | 1.01 | 254,836 | Bradford, Sullivan, Wyoming, Luzerne (part) |
| 254,593 | 2.00% | 1 | 1.01 | 254,593 | Bedford, Blair, Somerset |
| 254,496 | 2.00% | 5 | 5.01 | 1,272,481 | Bucks, Northampton, Lehigh |
| 254,345 | 2.00% | 2 | 2.01 | 508,690 | Monroe, Pike, Wayne, Carbon, Luzerne (part) |
| 254,334 | 2.00% | 6 | 6.01 | 1,526,006 | Philadelphia |
| 254,223 | 2.00% | 1 | 1.00 | 254,223 | Montour, Northumberland, Columbia, Schuylkill (part), Union |
| 253,353 | 1.99% | 1 | 0.99 | 253,353 | Erie (part), Indiana, Clarion, Forest, Jefferson, Warren |
| 252,899 | 1.99% | 7 | 6.97 | 1,770,295 | Montgomery, Berks, Delaware |
| 252,803 | 1.99% | 1 | 0.99 | 252,803 | Armstrong, Butler |
| 252,671 | 1.99% | 1 | 0.99 | 252,671 | Lebanon, Schuylkill (part) |
| 250,888 | 1.98% | 2 | 1.98 | 501,775 | Westmoreland, Fayette |
| 250,812 | 1.97% | 1 | 0.99 | 250,812 | Erie (part) |
| 249,443 | 1.96% | 2 | 1.96 | 498,886 | Chester |
| 246,506 | 1.94% | 1 | 0.97 | 246,506 | Washington, Greene |
| 244,670 | 1.93% | 5 | 4.82 | 1,223,348 | Allegheny |
| 255,509 | 2.01% | 50 | 12,702,379 | Est. Overall Variant: 9.58% |
After several attempts, everything came out within range. Now I was ready for Step 3: Divide Oversize Regions.
[…] Now on to Step 2: Regions! […]