In Vieth v. Jubelirer (2004), the Supreme Court decided not to intervene in the redrawing of Pennsylvania's Congressional districts. Racially discriminatory gerrymandering has been struck down by judges many times, but the Court could not find a formula for evaluating partisan gerrymandering - disenfranchising voters of a political party.
The 2010 census then determined new Congressional districts. Though a slim majority (50.28%) of Pennsylvania's popular vote was for Democratic candidates, Republicans won thirteen seats to Democrats' five. The parties continued to win the same districts in 2014 and 2016 (mapped below).
In the same case, a concurring (but optimistic) opinion was written by frequent swing-voter Justice Anthony Kennedy:
"Technology is both a threat and a promise… On the one hand, if courts refuse to entertain any claims of partisan gerrymandering, the temptation to use partisan favoritism in districting in an unconstitutional manner will grow. On the other hand, new technologies may produce new methods of analysis that make more evident the precise nature of the burdens gerrymanders impose on the representational rights of voters and parties. That would facilitate court efforts to identify and remedy the burdens, with judicial intervention limited by the derived standards.
If suitable standards with which to measure the burden a gerrymander imposes on representational rights did emerge, hindsight would show that the Court prematurely abandoned the field."
I like to think of this as Justice Kennedy's "hackathon pitch."
Since then, there's been an ongoing discussion about what might be a good test to show unconstitutionally unfair districts. There needs to be an objective mathematical formula, and a solid constitutional argument, for changing the system. We can expect this to be a major issue as we get closer to the 2020 Census.
Let's look at some of the best methods:
Partisan symmetry doesn't expect perfectly proportional representation in Congress. It tests whether if the parties were to switch statewide vote totals, they would also switch seat totals.
In an ideal system, the parties would have equal representation at 50% and mirror each others achievements as their vote total rose. Because we have a first past the post system, small increases add seats faster than proportional representation. Using the same data as the map at the top of the page, we can predict that Democrats would need around 56% of the vote to win just 7 of Pennsylvania's 15 contested Congressional elections.
| Dem Popular Vote | Seats Won |
|---|
Also called wasted votes - this method counts up votes of a recent election with over-representation in 'packed' districts and lost votes in 'cracked' districts, to show how a popular-vote-majority for one party is being diluted. This expects a model of hyper-proportionality (a party with a 2% lead would have 4% more seats, a 5% lead meaning 10% more seats, etc).
Click or tap a district to see it up-close:
Use the dropdown menu to select a district:
By this metric, the Democrats lost more votes than the Republicans in all but two of the contested districts. Even districts that they won in Philadelphia count against them, because the Democratic candidates received so much more than 50% of the vote.
When most people think of gerrymandering, they point to unusually-shaped districts as examples of shady redistricting gone amok. Multiple states require their districts to be geometrically compact, by comparing the edges or area of the shape to a simplified ideal. You could compare the shape to a circle of the same area, or the smallest circle which encloses the district, or a convex polygon.
Many unusually-shaped districts, such as Illinois's 4th District, are created as majority-minority districts to better represent our population in Congress. These were created and continue to be protected by the Voting Rights Act (it's a separate section from the rules struck down in 2013 by Shelby County v. Holder).
The general consensus is that compactness could be a helpful standard in redistricting, but it can't be the only factor that's considered. You can read more about how compactness can still cause 'unintentional gerrymandering' [PDF] (because so many Democratic voters live in densely-populated areas) and how purely computer-drawn lines can be discriminatory.
A computer can generate many thousands of redistricting maps, and show that the state's chosen map is multiple standard deviations more partisan when compared to its alternatives. You can see examples of how this could be applied in Wisconsin and in national maps.
Up until the 1960s, many states would leave their boundaries stagnant to protect incumbents and empower rural voters. In a case where one district’s population outnumbered another almost 10:1, the Supreme Court began to intervene. Since then, even a 1% difference between the most and least-populous districts has been ruled unconstitutional under “one person, one vote”. Theoretically this guarantees proportional representation within each state (remember that Wyoming’s one district and a district in California have considerably different populations).
After that decision, districts had to be redrawn after every census. This meant more frequent changes, but politicians found it difficult to design manipulative equal-population districts, until computers got involved.
In studies of redistricting after the 1990 and 2000 census, partisan gerrymandering was not seen as a significant factor - research in Drawing the Lines points to a problem with both partisan and bipartisan legislatures, who would use gerrymandering to protect incumbents. The 2010 census and 2012 election were a major shift:
From Gerrymandering in America:
…the consensus of the academic literature was that redistricting had only minor political effects. Incidentally, we are not challenging the consensus of this literature — there really does not appear to be much partisan bias in this time period. Rather, what we are arguing is that… the world changed, and it became possible to take partisan gerrymandering to its limits.
But it hasn't always been this way. And it doesn't have to be.
Cases against partisan
gerrymandering in
Wisconsin, Maryland, and North Carolina are moving forward and may set one of these methods as the standard.
"The Supreme Court will examine partisan gerrymandering in 2017. That could change the voting map." - Washington Post, Jan 31
The source code for this page is open source. It uses D3, Walkway.js, and boundaries from Ken Schwencke's national repo. Vote totals via the New York Times and Wikipedia.
Recommended reading: Gerrymandering in America and Drawing the Lines: Constraints on Partisan Gerrymandering in U.S. Politics.