Redress redistricting: My gerrymandering reform proposal that will actually work

Lot of ideas have been proposed to reform gerrymandering.  They usually revolve around nonpartisan redistricting commissions, or in some cases purely algorithm-driven redistricting.  The main problem with these proposals is that they attempt to defeat human nature.

Nonpartisan commissions aren't completely unbiased and are subject to lobbying and human foibles.  The algorithm-based methods come closest but I doubt anyone would actually accept them.  Standards for fairness or compactness try to force mapmakers to do what they have every incentive to avoid doing. What we need is a redistricting method that works with human nature, not against it. My proposal, which I call redress redistricting, will do that.

How it works

I'll give a simplified outline first, then go into details after.

1. Districts are drawn as they already are.  By legislative bodies, nonpartisan commissions, whatever.  No attempt is made at this stage to "fix" the process or force them to do it fairly.

2. After every election that is held with the districts, the results are examined to determine if they were fair.  This is done by comparing the overall percentage of votes each party receives, to the number of seats it wins, determined by an objective formula or table.  If the formula or table indicates that those numbers are out of whack, the map is deemed a gerrymander and, while the results of the election are allowed to stand, the map is invalidated and must immediately be redrawn for the next election.

3. When the map is redrawn for the next election (the redress map), it is drawn exclusively by the party that was screwed by the previous map.

4. The cycle continues with the next election.  In fact steps 2-3 are done after EVERY election using the map, or its redress maps.

Why it would work

This method would work because it works with human nature instead of against it.  The incentive to create egregious gerrymanders is removed because if you draw one that's too egregious, your opponent gets to draw the next one.  Same with the redress map... if you draw your redress map too aggressively, you'll just give it back to the original drawers in the following election.

Parties would still be able to draw themselves an advantage because the formula or table wouldn't be super strict.  For example it wouldn't require that a party that gets 50% of the vote gets at least half the seats. But a party that gets, say 55% of the overall vote, should get at least half the seats, a party that gets 40% of the overall vote should get at least a third of the seats, etc.

The incentive of not giving your opponents the mapping power next time is strong enough that it's all that's needed.  The formula would be loose enough that the mapping power still means something so there is incentive to keep it, but strict enough that egregious gerrymanders would result in the other party getting the mapping power.

It also would encourage voter turnout and competition in safe districts.  Today if a district overwhelmingly favors one party, the other party often doesn't even run a candidate because effort and votes and those districts are wasted.  But since this method takes into account the overall vote, parties have incentive to run good candidates and compete hard in all districts, even ones they have no chance of winning.  Plus to counteract that, the party that safely dominates a district still has incentive to run hard.