Karma: Gerrymandering Could Backfire

by Justin Meyer and Brendan Beery

There are many reasons why voters feel that their votes do not count. One of the most significant for the upcoming election is gerrymandering, a process where both Republican and Democratic state legislators have sliced up their states’ voting precincts in ways to favor their own parties. For example, Toledo and Cleveland are Democratic strongholds in an otherwise Republican Ohio. These cities are 115 miles apart, yet they share a congressional district. The Republican-controlled government in Ohio figured out a seemingly impossible way to join these two cities by stretching a narrow corridor between the two. Why would Republicans do this? Doesn’t it guarantee a Democratic victory in this district? It does, of course, but by not joining these urban centers with the more geographically reasonable adjacent suburbs, Republicans have maintained the electoral advantage in more districts than the one they intentionally sacrificed.

In a recent case (Gill v Whitford) before the U.S. Supreme Court, lawyers proposed a way to measure the extent to which legislative districts are gerrymandered. It’s called the “efficiency gap.” Here’s how it works. Imagine that the State of Oz has 4 election districts. The state is evenly divided — 200 Republicans live in Oz, as do 200 Democrats. But Republicans, having wrested control of the legislature in Oz, seek to redraw the election districts so that Republicans might maintain a durable and acute advantage for years to come.

The way to do that would be to pack as many Democrats as possible into as few districts as possible, creating a map that might look as formless as a cluster of inkblots, but achieves the goal of clumping left-leaning voters into concentrated geographical areas. Republicans might decide that they want 3 districts out of the 4 with safe Republican majorities (let’s say a 55-45 advantage) and that they’ll sacrifice one district and pack all the Democrats they can into that one district. It would look like this:

Screenshot (22)

Under this scenario, Republicans would likely win Districts 1 through 3 by 55-45 margins and sacrifice District 4 in a 65-35 Democratic landslide: 3 relatively safe districts and one throw-away. That would be good for control of 75% of the legislative seats in Oz, a state that is divided 50-50.

Assume an election where everyone voted predictably — Republicans for Republicans and Democrats for Democrats. The efficiency gap proposed in the Gill case would measure “wasted votes.” In District 1, for example, Republicans would have wasted 9 votes. That’s because they only needed 46 votes to win, but built in an extra 9 for a safe cushion. Democrats wasted 45 votes in District 1 because all those votes were cast in a district Democrats were unlikely to win — in a losing effort. The same goes for Districts 2 and 3: in each, Republicans wasted 9 votes and Democrats wasted 45.

In District 4, Republicans wasted 35 votes in a losing effort. But this sacrifice was nearly equally matched by the Democrats, who only needed 36 votes to win, but wasted another 29 votes achieving a 65-vote wipe-out.

Screenshot (23)

All in all, Democrats “wasted” 164 votes while Republicans wasted 62. Since there were 200 voters in each party statewide, that means Democrats wasted 82% of their votes while Republicans only wasted 31% of theirs.

In raw numbers, Democrats wasted 102 more votes than Republicans. The efficiency gap is calculated simply by dividing that number — 102 — by the total number of votes cast: 400.  So, the efficiency gap in Oz is just over 25% (102/400 = .255). (According to Nicholas Stephanopoulos, the scholar who proposed this methodology, any gap over 7% exceeds the historical norms in US elections and is evidence of improper gerrymandering.)

While this strategy works for the gerrymanderers in most elections, here we show why it may backfire during wave election years.

Pollsters are predicting a “blue wave” this year where the Democrats will win back the U.S. House of Representatives. The wave is in response to the increased anger liberal-leaning voters feel towards the Trump administration and corruption of the Republican-controlled government. This anger should inspire Democratic voters that tend to skip midterm elections to vote. Despite the palpable energy on the left, the magnitude of the wave has been questioned. This is because in the early 2010’s, when political lines were redrawn, the majority of states were controlled by Republicans, and they employed new technology to draw highly gerrymandered boundaries. The prediction of a weakened wave assumes that if gerrymandering will interfere with Democrats’ chances in a normal election year, it will have the same effect in a wave year. However, we argue that during a wave year there’s a chance that gerrymandering may backfire and cost Republicans more seats than they would have lost had they not gerrymandered.

To illustrate, consider the map of Oz. Like most regions, its population is spatially stratified with regard to political affiliation. Everyone living to the left of the Yellow Brick Road is a Democrat, and everyone to the right is Republican. The population size is the same on both sides, and during a typical year 40% of potential voters vote (typical midterm levels). To guarantee fair voter representation, each side of the river should be split in half, yielding 4 districts; 2 that always vote for Democrats and 2 for Republicans (Scenario 1). However, as previously described, Republicans have control of the state house and they choose to gerrymander the districts by making three districts with 55:45 Republicans to Democrats, and then one district with an overwhelming number of Democrats (Scenario 2). During a typical year the region will elect three Republicans and one Democratic representative.

Now consider a wave year like the one some predict this fall. If the left-side voter turnout increases by 25% (many special elections this year have seen an even larger gain) and the right side stays the same, then all of the districts will vote for a Democrat (Scenario 3). If the Republicans had not gerrymandered, then Republicans would have maintained half of the seats (Scenario 4). The cost of gerrymandering during wave elections could be greater than the benefit during normal elections.

Screenshot (21)

Scenario 3 is much like what occurred in Western Pennsylvania, where Democrat Conner Lamb won a special election in a highly gerrymandered district thought to be safe for Republicans. The assumption that gerrymandering will always help the party that gerrymandered is wrong. There are reasonable cases where gerrymandering can actually hurt the gerrymanderers.

Both voters and legislators can learn a lesson from this. First, voters: do not let gerrymandering discourage you from voting. In fact, your vote may have heightened impact because of gerrymandering. Your votes may not be “wasted” after all. Second, politicians: it may not be a good idea to gerrymander. Karma and math may conspire to punish your undemocratic impulses, pummeling you into political obscurity under a tidal wave of opposition.