In Defense of Gerrymandering

Gerrymandering gets a bad wrap these days. At first blush, it makes a lot of sense. Political parties attempt to bake in a political advantage when district boundaries are drawn. There’s something that seems inherently wrong with that.

Republicans and Democrats alike do this. In Democratic controlled states like Maryland (before governor Hogan) and California, partisans drew lines that diluted current Republican households across a number of districts or they packed as many Republican voters into a single district as possible. The idea is to win a lot of Democratic seats comfortably without wasting too many votes, and to have Republicans win fewer seats by running their totals up in a few districts but wasting many votes that could theoretically be used in other districts had the lines been drawn differently.

In Republican-controlled states, gerrymandering is usually more extreme. By a quirk of how the two parties voters are currently dispersed geographically, it is much easier to “pack and crack” democratic voters. Democratic voters disproportionately live in urban areas close to one another, whereas Republican households tend to be more spread out between suburbs and rural areas.

At the federal level Democrats are further disadvantaged. A number of US states are predominantly rural or suburban with few large cities. As a result, there are more Republican-held states than Democratic ones, and there are more Republican examples of gerrymandering. This makes gaining control of the House of Representatives more difficult for Democrats. Gerrymandering helps, but does not completely, explain why Democrats need to win the popular vote by between 4-7% to gain a majority in the House.

One common view is that Democrats will always be disadvantaged as long as gerrymandering is allowed. In this view, the asymmetric advantage that gerrymandering provides Republicans makes gerrymandering inherently undemocratic and unconstitutional.

This argument has an underlying premise that parties’ makeups are locked in and that a disadvantage today is a disadvantage tomorrow. That simply is not true. Recent elections have seen shifts in the two parties’ make-up. Suburban and more college-educated voters have moved into the Democratic party. Women, a group practically impossible to gerrymander, have also moved in the Democrats direction. Republicans likewise have seen a rise in white, non-college educated, working-class voters. If political parties can change their coalition, even if slower than desired, this significantly dampens the impact that gerrymandering creates.

Second, Democrats and Republicans should not have a right to prevent their current, ‘optimal’, or ‘ideal’ political coalitions from being electorally disadvantaged. Just because it is currently easier for Republicans to get rural non-college voters to vote GOP, does not mean that courts should mandate that those voters should be spread in such a way to maximize or disadvantage their electoral power. Among all the potential political fault lines and coalitions that can be created, a few actually exist. Why should we spend time elevating and solidifying these coalitions when making districts but not take into consideration other potential groupings?

Gerrymandering can act as a change agents in parties, keeping them from being sclerotic. Parties can change the types of people they appeal to by reforming their party platform or by allowing non-conforming members to run in districts where their party platform is not a winning argument. It used to be this way. New England Republicans were more liberal than Southern Democrats. Now instead of putting politicians that reflect a district, people want to litigate the district’s existence.

Gerrymandering arguments are often made in bad faith. Galen Druke, of fivethirtyeight, produced and hosted a podcast titled “The Gerrymandering Project”. In it he interviewed a number of partisans. In one instance a woman in Madison was upset because her Democrat-leaning district was split into two districts that Republicans consistently won. She felt like her vote was being suppressed because her preferences were no longer winning the day. But when her vote represented 60% of the will of the voters before gerrymandering, she won all the time. This lady wasn’t upset that she wasn’t winning 60% of the time. She wanted to win every time. Most people who are mad at gerrymandering simply wish that their political preferences were disproportionately over represented instead of underrepresented.

A large popular vote has downsides too. There’s a reason we have districts. One reason is to ensure that the strong opinions of one district (or state) does not override the wishes and desires of other districts or states. Policies have to be generally popular across the state or country instead of just being intensely popular in denser areas. In this set-up, parties have to appeal to a more diverse group of people, at least geographically, which is important for keeping a state or country together.

Gerrymandering is certainly bad for the party being gerrymandered in the near-term. When parties fail to put forth nominees that mirror the desires of the district, they lose. The nationalization of parties makes this strategy harder for parties but not impossible. Parties have a number of options. They can try to convert voters, expand the party’s tent, or nominate better candidates that reflect the district. The result is a healthier and more fluid politics with (hopefully) less partisanship and greater zones of policy compromise.

The parties are currently accepting an electoral detente in most districts and do not accept any heterodoxy. This is not gerrymandering’s fault. Fixing gerrymandering just allows this status quo to continue.

All that said, gerrymandering is bad. Just not as bad as people think.

Partisanship Elevates Extremism

Nate Silver, over at fivethirtyeight, has mentioned numerous times that, after holding a number of factors constant, more moderate candidates perform better than candidates with extreme positions.  The mechanism works in two parts.  First, moderates are more likely to win over a larger share of the ‘getable’ independents that genuinely sit in the middle and vote for democrats and republicans.

The second reason is a little less intuitive.  Extreme candidates on the right and left certainly motivate the wings of their party.  They bring, who would otherwise be unmotivated to vote, off the sidelines and into the political arena for their side.  However, there is evidence that extremists disproportionately motivate the other side’s voters.  So, while members on the extreme can motivate extra voters on their side, those numbers are swamped by the other side’s newly motivated voters.

In an age of weak parties and strong partisanship, does this still hold?  When any democratic nominee will be characterized by republicans (read: Trump) as an identity-politics wielding socialist, do these two factors actually hold?  Perhaps.  But there is a concern that Trump’s ability to (mis)characterize moderate contenders may actually dampen the both practical electability advantages moderates have over extremists within the democratic party. 

In a world where moderates and moderate independents refuse to cross party lines due to strong partisanship and one in which the political strategy of the day is to ramp up the extremes on both sides of the aisle, the argument and advantages for a moderate candidate seem to be non-existent.   When either side’s moderates won’t budge and the other side’s extreme is already motivated, we might as well pick the candidate that motivates the most people.  This person is often rather extreme.

Independents and moderate in both parties need to be willing to reward centrism and moderation—no matter the party.  Otherwise these voters risk rewarding the very types of politicians and policies they most dislike.