North Carolina adopts non-partisan redistricting rules

[Foolacious]

6. I use software to draw congressional districts, as a hobby.

I've developed my own software but more recently use a publicly-available system. Here's something I've noted:

If I do NOT take account of partisan (Democratic/Republican) data, but simply create compact districts of equal size that follow existing precinct and, to the extent possible, county boundaries, then the maps I make are almost invariably "auto-gerrymandered" in favor of Republicans, by which I mean that they receive a higher proportion of safe and likely seats than their proportion of the state population.

One reason for this is that Democrats tend to live in more densely populated areas, so they end up in a few districts that are 70%, 80%, even 90% Democratic, while Republicans end up in a larger number of districts that are 55% to 60% Republican.

But even if I do try to redistribute precincts into districts with an eye to correcting this partisan imbalance, it can still be difficult to achieve good proportionality. This is because the precincts in more densely populated areas are frequently five or ten times as populous as rural ones and frequently "interlock" with each other (where one precinct contains "outposts" of another precinct) so they tend to get automatically lumped together no matter what. Now, why are precincts in Democratic areas larger, and why are they interlocked? I can see no good reason for it, but it seems to happen in many if not most states.