Redistricting and the Abuse of Power

Every 10 years the Constitution requires a census to count heads so that Congressional seats can be properly divided between states and district lines can be drawn within states to more fairly reflect the last ten years worth of migration, immigration, births, deaths, etc. That's the idea, to more fairly reflect shifts in populations so that Congressional districts contain roughly equal numbers of citizens.

Unfortunately in practice the party that happens to be in power immediately following the census has the opportunity to abuse that power and lock in unfair advantages which they will hold over the next ten years -- perhaps even indefinitely, and, like those in power more generally, that lucky party often chooses abuse.

How the district lines are drawn matters, and even ostensibly fair-minded people might advocate for different boundaries which produce different political consequences.

Let's take a look at this more closely. In the images below each dot represents a voting block (precinct or group of precincts) that typically votes either Republican (red) or Democrat (blue).

Let's say that we need to form 5 voting districts from the fifty dots hence each district must contain 10 dots. Below, let's look at a few different maps that satisfy that criteria and consider the consequences of each.

The map above is "fair" in the sense that a 3 - 2 majority of seats could be reasonably won by either party, thereby accurately reflecting the overall partisan balance. It is also true of this map that 4 out of the five are "safe seats" in the sense that only Democrats could win districts 1 and 4, and only Republicans could win districts 3 and 5. Safe seats mean representatives elected from those districts face effectively no competition from the other party and their only threat is from their own party's primary. Safe seats encourage both partisan extremism, given that only the most partisan tend to vote in primaries, and loooooooooong tenured service in office for representatives once elected.

Another fair but very different map is depicted below:

Here we also have a fair map in that either party has a good chance of winning each seat and a fair chance at winning a majority of the total in any election. This map presents different advantages and different challenges. There are no safe seats here, meaning representatives face competition each cycle and are likely to at least make an attempt to reach out to moderates from the other party.

There are potentially three drawbacks to this map however:

Poor "contiguity." These long stretched out districts make it difficult for constituents to get to events held somewhere in the district far from home. Ostensibly, different parts of these long, skinny districts will have different economic and environmental issues making it hard for a representative to reflect the district's interests.
Churn. There is likely to be a great deal of turnover in office, with representatives gaining office and losing office again after a term or two. This is likely to lead to a diminution of experience in the Capitol and concern that when the legislators aren't experts due to lack of experience they are more dependent on staff, advisers and interest group lobbyists.
Thrash. One party could easily grab all 5 seats in one election and lose them all in the next. Huge swings in representation could lead to rapid and dramatic shifts in public policy, creating uncertainty and administrative challenges.

But Gerrymandering can also easily be applied to this map to give a lasting advantage to one party over the other as depicted below.

An unfairly Gerrymandered map giving long-lasting partisan advantage

Districts 1 and 3 are totally safe seats for the Democrats, and 5 is totally safe for the Republicans. But Districts 2 and 4 are almost totally safe districts for the Republicans as well. In all likelihood, regardless of the region's political dynamic and shifting enthusiasm, Republicans will always hold three of the five seats. It is hard to justify this type of biasing, but both parties have done this when given the opportunity. The problem now is we have gotten so much better at it.

Let's take a look at a few other maps with a more realistic voting distribution. The map at right represents a heavily Democratic urban area surrounded by heavily Republican suburbs or rural areas.

A more realistic distribution showing a region evenly divided politically.

Very densely populated cities and relatively sparsely populated rural areas are the norm in the US. It makes map drawing more challenging in that the economic interests of cities and rural areas are vastly different and crafting a map that is fair politically and represents the economic interest fairly demands districts of vastly different sizes and shapes and allows for much easier gaming of the system.

An unfair map, Gerrymandered to favor Democrats

The maps at right are examples of the ways that districts are drawn to favor either Democrats or Republicans depending on who had the majority when the last census was drawn.

Here we see a map drawn to favor Democrats and city economic interests. Democrats will win Districts 2, 3, and 4 seats every time. Republicans have two safe seats. The 5 representatives from the area have no reason to talk to voters or representatives from the other party.

An aggressively Gerrymandered map demonstrating "cracking and packing".

At right, we see an even more extreme case of Gerrymandering, this time creating a 4 to 1 advantage for Republicans in a region that is in fact split 50 - 50 politically. Here you see "cracking and packing" perfected. Here the Republican legislature draws a map to "pack" District 4 with Democrats, then "cracks" the rest of Democrats such that they are a minority in the other four districts.

Extreme cases like this happen. Virginia, a state that elected Democrats in both US Senate seats, went for Obama in 2008 and 2012, and Clinton in 2016 by 5%, has a 7 - 4 advantage for Republicans in the House of Representatives. The State Legislature in Virginia until 2018 had a 66 - 34 advantage for Republicans in the lower house (!) and a majority for the Republicans in the upper house as well. This extreme case of Gerrymandering was possible after the 2009 and 2010 Tea party wave elections gave Republicans control of both houses of the legislature and the governorship, granting them unchecked ability to lock in partisan advantage long-term. Extreme Gerrymandering can backfire in a real wave election cycle for the other party. In the map above it's not inconceivable that a true wave election could flip any of the Republican leaning districts because the majorities aren't prohibitively huge. 2017 was such a wave election in Virginia with a Democrat winning the governorship by 9%, yet the Republican gerrymandering scheme held even then! Democrats flipped many seats in the house, but fell one short (51 - 49) on a coin toss for a tied election in one district.

We live in a country where Congress has a 10% approval rating yet often see re-election of incumbent members of Congress over 95% of the time. More than 90% of the seats in the House of Representatives are "safe seats." That's not good for democracy. People ask: "When nearly two-thirds of the population support universal background checks and other make sense legislation, why doesn't Congress act?" The answer in part is that legislators don't have to listen to people in their district, even those in their own party. They have safe seats. In fact the personal advantage of having one's own safe seat even outweighs the partisan disadvantage from Gerrymandering today. The Democrat leadership in Congress has shown no interest in challenging Gerrymandering.

Of the two "fair" maps for this area depicted below and to the right, I prefer the bottom map with easily flipped seats.

A fair map with 2 safe seats for each party.

The top map has districts 1 and 4 as safe seats for Republicans and districts 2 and 3 as safe seats for the Democrats. 80% safe seats is preferable to 90%, but not by much.

The map below and right offers 5 contested seats. A more democratic mechanism than term limits is to make sure as many seats as possible are hotly contested. Waves happen for both parties, and they should have consequences. If a party sweeps in with a big enough majority they should have the power to make changes, and the accountability for the results of their work. I'm not so concerned about the experience issue as there is little evidence that lobbyists could have any more influence than they already do.

Moreover, I think citizens would be more informed and engaged in real, honest political discourse if far more of our elections were "winnable" by either side. Sadly, as is probably obvious, office holders like their safe seats and are far more likely to propose and adopt maps that protect the advantages in their own district.

This issue needs to be taken out of their hands. I'd like to believe the US Supreme Court will do the right thing with the case in front of them, but I don't trust them to do so.

There are ways to do redistricting in a fair manner, and organizations like One Virginia 2021 are working hard for fair redistricting here in VA. I prefer solutions involving mathematical algorithms. "Independent" commissions could work but the independence of these commissions might be for sale. I would settle for mathematical tests that could be applied to maps proposed by legislatures that would evaluate the plan for fairness and reject biased maps until the legislature puts forth something reasonable. There would still be more safe seats than I'd like to see, but it is better than what we have.

I'd love to hear your thoughts on how maps should be drawn and your reasoning behind those thoughts. What's the right balance between safe seats and competitive seats in regions where we have an option?