Well, I gave you the barest explanation of what "my" system would be.
Here is the detail....
To make the math easier, lets say you have a state that has 10 congressional districts and 8M people. What you want is 800K per district--or close to it. Currently what is done is that they draw lines any which way they can to carve out enough safe seats for the two major parties while putting roughly 800K in each weird assed shape. Sometimes a district is as wide as a city block to accomplish such a feat. My system would do away with contiguous shaped districts and assign ZIP codes to each seat.
The way it would work is like so:
- Take all of the ZIP codes in the state and make a list numerically...00001, 00002, 00003 etc....top to bottom. Lets say you have about 800 ZIP codes.
- Divide the 800 ZIP codes into 10 groups using nothing but income of the residents of that ZIP code. Income is often the determining factor that goes along with what voters want from their elected officials. Less affluent folks likely would want development and security; more affluent folks want parks and capital improvements.
- Anyway, take the highest earners and put them into Group A. The next highest goes into Group B, then Group C...all the way down to Group J. Then re-organize the list by income into 10 groups.
- Then assign, at random, one ZIP code from the first group to each district until they're all gone. Then do the same to Group B, then C etc...
- What you end up with is nearly identical demographics where each seat will have a few affluent ZIP codes but mostly those in the middle or low end of the earning spectrum. There will be no more safe seats drawn to shoehorn blacks or Hispanics into some districts or insulate the wealthy.