Many news outlets (NY Times, Wall Street Journal) are talking about gerrymandering, but they all basically say the same thing:
- Gerrymandering's been going on forever
- Republicans are the primary culprits and Democrats the victims
- Democrats are using it effectively this year to negate the Republicans' historic advantage
They're also all leaving out many important points:
- Democrats had a huge advantage from gerrymandering for 50 years.
- Gerrymandering's effect in 2018 and 2020 was almost nil.
- Democrats gerrymandered in 2010-2020 as well, and the most gerrymandered state is California.
- Democrats passed numerous commissions in the 2010s to reduce gerrymandering, but those are basically ignored.
Why Is This Important?
I imagine for most readers who aren't following this every day, their eyes glaze over as soon as they start reading the word gerrymander. It's one of those words that is easy to forget its exact meaning because it's rarely used, its meaning can't be intuited from the term itself, and is only used in the context of boring political stories. While you can easily look up the definition somewhere and get an explainer, simply put, it signifies the process for determining who gets to vote for each of the 435 congressional seats, which matters to which political party controls Congress. Politicians in each state get to figure out who will vote for each of their state's representatives to Congress. This process is known as redistricting because they draw the lines for the districts, and who falls into what district. Politicians can draw these lines in a way so certain voters fall within certain districts and thereby determine whether the district has more Democrats or Republicans. If they do it well, they can distort the number of seats that each party will win in their state.
Republican Gerrymandering Isolated to 3 Election Cycles Out of Last 25
Taking a historical perspective, you can see that Democrats held the advantage in seats continuously from 1946 to 1994. With some years having an advantage of more than 40 seats. The average from 1958 to 1992 was 28. Then in the 1990s, after the Republican revolution, Republicans began punching above their weight, winning more seats than votes from 1996-2006, with a switch during the Democratic wave of 2008. Then, because Republicans were so successful at all levels in 2010, they finally had the chance to gerrymander districts for themselves, on a scale they were never able to before. You can see the effect of that in 2012-2016, where they averaged a 19 seat advantage for three cycles. Then in 2018, when Democrats retook Congress, that advantage shrank to just 1, then 2 in 2020.
The point of all this is that the effect of gerrymandered districts has historically advantaged Democrats, and at a magnitude much higher than Republicans' edge in the 2010s, yet it's only now that we hear about the scourge of gerrymandering, and always without the historical context. That context is important to understand when judging the scale of the problem and its import.
Gerrymandering Effect in 2018 Very Small
Using the election result data available from the FEC, I ran an analysis of the votes casts versus the seats won. If seats were distributed proportionally to votes, there would have been 2 additional seats awarded to Democrats, so nationally, there is a small effect. There are 8 states where Democrats have the advantage and 17 states where Republicans have the advantage, so there are about twice as many Republican states. This shouldn't be a surprise as Republicans tend to have a broader geographic appeal, holding majorities in more states than Democrats which is counterbalanced by the fact that Democratic states tend to have more people.
Looking state by state, California, has the largest discrepancy in votes vs. seats, and Democrats have a 10 seat gerrymandering advantage. They should won 36 seats, given the vote totals, but they won 46. The next highest was Texas, which gave Republicans an extra 4 seats beyond what their vote totals would merit. Adding up all the states individually, Republican gerrymandered states net them an additional 25 seats, and Democratic states net Democrats 20 seats. This doesn't match the +2 national Republican advantage because combining the votes across all states, including the 26 that show no advantage, produces a different outcome. What readers should take away from this is that Democrats still gerrymandered a great deal, but just not as much as Republicans.
States with gerrymandering commissions
Lastly, the other context that these articles consistently omit, is that in the 2010s, as Democrats realized they were newly on the losing side of the gerrymandering effect, they led a movement to eliminate gerrymandering altogether. Several states passed redistricting commissions to ensure district lines were drawn fairly and objectively, without favoring either party. According to ballotpedia, eight states have a commission to manage the redistricting process (California, Washington, Idaho, Colorado, Arizona, Michigan, Hawaii, and New Jersey). The National Conference of State Legislatures says the total is ten, Virginia having added one in 2020, and it also includes Montana. These are independent commissions which draw up Congressional district maps.
In 4 states (Maine, New Mexico, New York, and Utah), there is an Advisory Commission, which draw up maps that are submitted to the legislature for approval in a conventional vote. Notably, this is less helpful, as the legislatures can easily, and on a partisan basis, reject the independent map, and adopt its own.
Lastly, in 3 states (Connecticut, Indiana, and Ohio), there is a back-up commission if the legislature can't agree.
Of these states, which purportedly want to reduce gerrymandering, New York's final map gets an F score from the Princeton Gerrymandering Project. In 2018, New York's map provided Democrats a +2 seat advantage, and the new map will give Democrats three more seats. California, as mentioned previously, already has a pretty strong Democratic advantage already, and it's expected to either maintain the current composition, or give Democrats an extra seat.
The importance of including this information is so that readers and citizens understand that Democrats aren't as committed to fairly drawn districts as they say and also, the existence of a commission doesn't guarantee fairly drawn districts (which should make everyone wonder whether a federally mandated solution to this will only reward the more craven).
Background Links
Princeton Gerrymandering Project - They conduct pretty good quantitative analysis on the maps and effects.
National Conference of State Legislatures - Redistricting - more detail about the commissions
Notes on Methods and Sources
- I did analysis using different sources and at different times, so numbers don't always match up. Primarily, the state analysis and national analysis were done using different sources.
- National analysis used data provided through Wikipedia (their sources are always listed).
- The basic methodology was not strictly a gerrymandering analysis but was a comparison of the composition of votes by party to the composition of seats. In a randomly distributed state, with fairly drawn maps, this should not lead to a difference in seats of more than 1.
It should also be noted that there are many complications to conducting this analysis.
- Four races had only one candidate so the votes weren't tallied or entered.
- Taking the national composition and comparing to seats ignores that there are several states with only one seat so can't be gerrymandered.
- Because the voters are not distributed uniformly by party throughout the country, there can be discrepancies between the vote composition and seat composition due entirely to geographical distribution.
- Results can change from election to election based on turnout. When one party turns out in higher numbers, it gives the illusion of a temporary gerrymander.
No comments:
Post a Comment