#PropRep can kill the gerrymander. Here's how, specifically.
I have a plan that can end partisan gerrymandering. Once you’ve read it, I think you’ll agree that it’s something that’s politically and legally feasible to pass under a Democratic congress in 2021 (and be signed by a Democratic White House that year). I think you’ll also agree with me that the other existing plans would be harder, and are not likely to succeed that quickly.
If you’re reading this, you’re probably a bit of a wonk. People like you have important roles to play in my plan, especially now in the early days. We don’t all have to be unanimously pushing this one plan, but we do have to at least agree that it’s credible and avoid undermining it. I and others are already beginning to work to build this consensus in various ways, and this diary is one part of that effort. So I hope you read and recommend it.
If you’re already knowledgeable about gerrymandering and proportional representation, you can skip the introductory sections below and start reading at the section “The plan”, where I’ll begin presenting my novel proposal for resolving this problem once and for all.
The problem: partisan gerrymandering
Partisan gerrymandering is a huge problem, for multiple reasons:
- It helps Republicans and hurts Democrats on average. For instance, Decision Desk HQ thinks that if the House elections happened today, Democrats would win 54.2% of the vote but be favored to win in only 208 districts (which, combined with an expected 7 wild-card pickups, makes 215 total Dem seats). That is either 47.8% or 49.4%; a Republican gerrymandering advantage of 4.8%-6.4%. (Technically speaking, this advantage includes both intentional gerrymandering and natural demographic sorting, but the plan I’m proposing would fix both problems; more on that below.)
- It debases both parties by leading to safe seats and unaccountable incumbents. Essentially, incumbents only ever have to worry about the primary electorate, and often not even that. That leads to either radicalization or complacency/corruption (or in some cases, both). If you’re a progressive in a safe Democratic district, you might think that the radicalization works in your favor — ideally, you can push your representative left by credibly threatening a primary campaign. But overall, for various reasons, this gerrymandered playing field favors moneyed interests and/or the radical right far more often than it works for progressives.
- Gerrymandering stems from the first past the post (FPTP) voting method, which has much broader problems — promoting wasted votes and lack of representation; encouraging mudslinging, content-free campaigns; excluding minority opinions from the political debate; etc.
And the problem is getting worse. Computer-aided gerrymandering is both easier and more powerful than ever before. And although politics has always been a dirty game, Republicans, seeing long-term demographic trends turn against them, are fighting dirtier than they have in modern history (which is saying a lot).
Luckily, recognition of the problem posed by partisan gerrymandering is also growing. It’s always been seen by most voters as something ugly, but I remember that in earlier decades it was basically seen as an inevitable and tolerable evil. Now, it’s being recognized more and more as a crisis.
One indication of that growing recognition is the number of current legal cases about gerrymandering. This includes multiple cases in North Carolina, Virginia, Texas, and even Maryland (where Republicans are complaining about a Democratic gerrymander). But of course the biggest case of all is Gill v. Whitford, from Wisconsin, which has already had oral arguments before the Supreme Court and for which a decision is expected sometime before next July.
Possible solutions: nonpartisan redistricting or PropRep
There are basically two ways to fix gerrymandering: either draw better lines (nonpartisan redistricting), or make it so that the lines don’t matter (proportional representation, aka #PropRep). As progressive Democrats, we should clearly support anything that would help. But given the choice, PropRep is clearly a more complete solution.
(Quick note: I say “PropRep” rather than “Proportional representation” or “PR” because the former is too long and the latter is too ambiguous with things like Puerto Rico, public relations, or pull requests. Using syllable-based abbreviations rather than acronyms can feel a little strange at first but you get used to it. It’s a lot better for SEO and it’s actually more common than acronyms in many languages.)
The problem is, better line-drawing can help, but it can never really solve the whole problem. Consider the simple illustration of how gerrymandering works on the right (or above, on a phone). Because Democrats are a greater majority in urban centers than Republicans are in rural areas, the cleanest possible lines still lead to a Republican advantage. And no matter how you draw the lines, almost half of the voters will waste their votes and/or lack real representation. The problems of incentives towards polarization, lack of accountability, and low turnout might be slightly mitigated by drawing better lines, but they will never be eliminated that way.
On the other hand, under PropRep voting methods, gerrymandering simply… doesn’t exist. PropRep means a voting method designed so that seat share will always be roughly proportional to vote share. A party that gets a certain percentage of the vote should get that same percentage of the seats, give or take some small rounding error because you can’t get fractions of a seat.
Another way of saying this is that PropRep methods minimize wasted votes. In first past the post (FPTP), at least half of votes are wasted — either going to a losing candidate or just uselessly running up the score on a candidate who’s already passed 50%. In PropRep, wasted votes are limited to the rounding error I mentioned above; just a tiny fraction of the total. For instance, in a state with 9 House seats, PropRep could guarantee that only about 10% of votes are wasted; with 19 seats, that would be down to about 5%.
There are various flavors of PropRep, and I’ll discuss those distinctions below. But for now, I just want to emphasize that PropRep has the potential to completely solve the problem. No more gerrymandering, period. And in fact, it would help solve other problems with FPTP, and thus make our entire political culture healthier, at the same time.
The hurdles: Court case, initiative, and/or bill?
So there are ways to fix gerrymandering. How can we get these measures implemented? The four basic possibilities are court cases; state-level voter initiatives and referendums; state legislation; or congressional legislation.
Source: Propublica
I’ve already mentioned, above, a few of the multiple court cases relating to gerrymandering. Historically, the Supreme Court has been somewhat willing to rule against racial gerrymandering, but very reluctant to do so with partisan gerrymandering. It seems that it’s not so much that they believe partisan gerrymandering isn’t constitutionally problematic, as it is that they are worried that it’s hard to draw clear lines, legal tests that would distinguish egregious gerrymandering from the reasonable judgment calls of the legislative branch doing its job.
This distinction between unconstitutional racial gerrymandering and unrestricted partisan gerrymandering has led to some ridiculous spectacles. In North Carolina, when the original 2011 districts were overturned as a racial gerrymander and the Republican legislature was redrawing them, David Lewis, the Republican chair of the committee, openly bragged that “to the extent possible, the map drawers [would be instructed] to create a map which is perhaps likely to elect ten Republicans and three Democrats…. This would be a political gerrymander…. But that is not against the law.”
But that may well be about to change. In 2004, in Vieth v. Jubelirer, Justice Kennedy wrote that ““The failings of the many proposed standards for measuring the burden a gerrymander imposes on representational rights make our intervention improper… [but] if workable standards do emerge to measure these burdens ... courts should be prepared to order relief.” Now, in Gill v. Whitford, he’s the swing vote again, the problem is much worse, and the suggested standards are much clearer.
The Whitford plaintiffs have a clear suggested standard for determining whether a gerrymander is unconstitutional. If the district-drawing process was controlled by a single party; the resulting districts are biased towards that party beyond historical norms, as measured by an “efficiency gap” (partisan difference in wasted votes) of over 7%; that bias is reasonably robust over various potential scenarios; and if those who drew the districts cannot prove that the bias was necessary in order to achieve some other legitimate objective such as minority representation; then the gerrymander would be ruled unconstitutional. Each of these four standards is reasonably clear-cut, so that Kennedy would not have to fear a flurry of judgment calls making their way up to the Supreme Court. I think that, in those circumstances, he’s likely to swing to the side of the 4 liberal judges.
Still, even if the Supreme Court upholds the lower court in Whitford and strikes down the Wisconsin gerrymander, such legal remedies are not ideal. Even with clear legal standards, litigation can take years, so that Republicans could still get gerrymandered advantages for several elections before their maps were struck down. And wherever the line was drawn, they’d learn to dance along the edge of it; even if a 7% efficiency gap is ruled unconstitutional, a 6% gap in every Republican-controlled state would still be intolerably undemocratic.
Another way to respond to partisan gerrymandering is with voter initiatives and referendums. In California and Arizona, such initiatives have already led to establishing independent redistricting commissions. Similar efforts are underway in CO, MI, MO, OH, OR, SD, and UT; and in PA, though citizens cannot directly create a referendum, an effort is underway to get the legislature to enable one in order to amend the state constitution.
Still, as noted above, even the most impartially-drawn districts are still not as good at fixing the problem as PropRep. Furthermore, some of the states which need reform the most — such as FL, NC, and TX — are the very places where passing such initiatives is infeasible or outright impossible.
State legislation is another way gerrymandering could potentially be fixed, in theory. In practice, it’s hardly ever a key battleground; the states which most need such legislation are, by definition, the very ones where it could never pass.
Which leaves the final possibility: federal legislation. For instance, Rep Beyer (D-VA) has sponsored the Fair Representation Act, a bill which would impose both nonpartisan redistricting and PropRep. The specific PropRep method it would enact is called Single Transferable Voting (STV). This is based on the idea that if your vote is not needed by your first-choice candidate — either because they can’t win even with your vote, or because they could already win without your vote — then it should be transferred to another candidate to avoid wasting it. In order to enable these transfers, the ballot asks you to rank all possible candidates in the order you’d like your ballot to transfer in. Because in larger states ranking all candidates statewide could get almost impossibly tedious, STV restricts you to new, larger districts that would each elect (up to) 5 seats at once.
Though this bill obviously won’t pass in the current Republican-controlled congress, it’s an important symbolic first step towards a real solution for gerrymandering. And, as the bill’s preamble argues, federal legislation to control House elections is perfectly constitutional, under article I section 4, article IV section 4, and amendment XIV sections 1, 2, and 5. In fact, especially if the SCOTUS recognizes gerrymandering as unconstitutional in Gill vs. Whitford, amendment XIV could enable Congress to impose remedies for gerrymandering even in state legislative elections.
The plan: the PLACE voting act
Finally, it’s time to share what I promised you in the title: a concrete, realistic plan for ending partisan gerrymandering by 2022. This plan centers around a proposed bill, the PLACE Voting Act (PLACE Act for short). I’m going to explain this by contrast with the Fair Representation Act (FRA) mentioned above. In doing so, I’m not trying to minimize the historic importance of getting the FRA sponsored in the House; I’m just looking for the strategy that has the maximum chance of success going forward
First, something that almost goes without saying: Republicans will be trying their best to obstruct this plan at every step. That’s why the target date is 2022; because before 2021, the Republicans will probably hold the presidential veto pen.
Also, if Trump (or Pence or whoever else comes after) gets to create a 5-4 wingnut Supreme Court, that’s a separate problem that also needs resolving before then. Personally, I think that the Democrats should be ready for court-packing in that case; Gorsuch’s seat is stolen by McConnell’s unprecedented refusal to do his constitutional duty under Obama and at least hold hearings on Garland, and arguably Roberts and Alito’s seats are also the fruit of the poisonous Bush v. Gore decision. My plan below is clearly constitutional from an unbiased point of view; but with 5 wingnut justices, that wouldn’t be what mattered. So perhaps the SCOTUS would need fixing first, to ensure that the median justice is at least as non-hacky as Kennedy.
So, assuming that there’s a unified Democratic government in 2021 could they pass the FRA as currently worded? Probably not. It would extensively rework the system by which every Democratic representative would have won their seat, and make every incumbent justifiably worried that it would stand in the way of their re-election. As good as it might be in the abstract long-run for the party and country as a whole, if it’s a concrete short-run threat for each and every incumbent, it would be DOA even in a Democratic-controlled house.
Of the colored states, NY, MO, and PA would be exempt because they don’t have one-party control of the redistricting process. (Also, the NY advantage is not robust across vote variations.) So that’s 10 red states and 2 tiny blue ones.
This brings us to the first new idea of the PLACE act: impose the gerrymandering remedy only in the states that need it most. In other words, use standards similar to those proposed by the Whitford plaintiffs to test for the worst partisan gerrymanders, and leave the rest of the states alone. The PLACE act would only kick in if a state had one-party control of the redistricting process; if that was projected to lead to a clear, measurable bias in favor of that part, using a metric like the efficiency gap; and if that bias could be shown to be robust over reasonable variations in projected party vote totals.
That would mean that a Democratic congress would be imposing a new voting method on largely Republican states; especially those states where this change would most level the playing field in the Democrats’ favor. That’s much more feasible than asking Democrats to vote for the FRA and change the system that elected themselves.
So we want an act with a remedy that would only kick in in response to a Republican attempt at gerrymandering. Thus, we need a good remedy for gerrymandering that’s implementable in well under two years, even in a state where the local authorities are dragging their feet and sabotaging the process. Under those conditions, creating a new redistricting commission and waiting for it to finish its job, or procuring and testing new voting machines, are probably not workable.
Is there a way to fix gerrymandering that doesn’t require new machines or new districts? Yes: it’s called PLACE voting. The name stands for “Proportional, Locally-Accountable Candidate Endorsement voting”. There’s a sample ballot on the right; here’s a quick sketch it works:
- Like STV (the method in the FRA), PLACE voting is based on the idea of avoiding wasted votes by transferring those votes to other candidates. But instead of asking voters to determine the order of those transfers by ranking all the candidates — a task which, with literally dozens of candidates, can be burdensome to many voters — those transfers are made by combining the voters’ favorite candidates party and public pre-election endorsements with the votes of other same-party voters. So if I chose candidate Jessica Jones as my first choice, and the other Democrats she’d publicly endorsed were Peter Parker, Luke Cage, and Taylor Hebert, then if Jones were eliminated and Cage had the most votes of the other three, my vote would go next to Cage.
- Thus, in order to vote in PLACE, you’d only need to choose your favorite candidate. To do so, you could either pick them from a list of candidates running in your local district, or write-in another candidate from anywhere in the state. For the rare voters who’d rather their vote was wasted rather than let their chosen candidate help determine where it was transferred, there is a second box to check to the right of each candidate which chooses that candidate but prevents vote transfers. Also, for partisan voters who haven’t researched the candidates in other districts, there is a way to vote for a party as a whole.
- Immediately after the PLACE votes are first tallied, any candidate with less than 25% of the vote from their local district is eliminated (with their votes transferred as usual; and keeping at least one candidate from each district). This ensures that each district has an important say in who’s elected locally, and helps prevent excessive ideological splintering into tiny single-issue parties.
- Then, winners are found using a Single Transferable Vote (STV) process, with an additional guarantee that there will be exactly one winner from each district. This process eliminates the weakest candidates and transfers votes until there is a set of winners, each with an equal amount of votes, and the remnant of wasted votes is smaller than the amount each individual winner has. Thus, in a state with 9 or more seats, there would be under 10% wasted votes.
- To ensure that voters who didn’t win their district are still represented, each district is assigned as “extra territory” to one representative from each party except the one that won locally.
You can see more details of PLACE voting at the FAQ. PLACE has a number of advantages as a voting method, but the most important in this case is that it can use the same machines and districts as FPTP (the current system).
Alternate plans? “Fair districts”, STV, MMP, etc.?
Really, the plan above is based on two ideas rolled into one:
- A 2021 Democratic Congress should pass a gerrymandering fix that only applies to the places which need it most; that is, largely to Republican states.
- That fix should be based on PLACE voting.
Would any other gerrymandering fix work in place of PLACE? I’d like to hope so, because the more robust this plan is to small changes, the better it is. But on the other hand, other fixes would I think be both less likely to pass, and inferior if they did pass, than PLACE voting.
So, let’s look at some other ways gerrymandering could in principle be fixed, and their advantages and disadvantages.
Independent redistricting (“Fair Districts”):
Congress could require that states create nonpartisan redistricting commissions
Advantages:
- Doesn’t require any unfamiliar voting methods.
Disadvantages:
- There’s still a risk that the “nonpartisan” commission would somehow be captured by partisan interests.
- Even the best-drawn districts still lead to wasted votes, some tendency towards Republican bias, and many of the other problems of FPTP.
- Congress telling states how to choose commissioners runs into Constitutional issues that Congress requiring a certain voting method doesn’t.
- Fixing gerrymandered districts could take more than 2 years, so the bad districts would still get a chance to be used.
Single Transferable Vote (STV) with 5-member districts
This is the solution in the Fair Representation Act, using ranked ballots and 5-member districts. STV is used in countries like Ireland and Australia.
Advantages:
- This is the solution favored by FairVote, the largest US voting reform nonprofit. They’ve gotten a head start educating people about this voting method, which they class with IRV as a form of “Ranked Choice Voting”. Whatever solution we’re going for, it’s going to be important to get FairVote not to work against us, so having them on-board from the start is an important advantage.
- Like any PropRep method, wastes far fewer votes than FPTP, and is generally massively better.
Disadvantages (compared to PLACE):
- More work for voters to rank all candidates. And there’s no way for voters to know ahead of time at what rank their vote will be “used up”, so inevitably they either waste time ranking too many candidates or risk wasting their vote by ranking too few.
- Requires redistricting, slowing down the process of implementation.
- Requires new voting machines.
- Requires a centralized counting process and makes it impossible for ordinary citizens to check the outcome and prevent fraud.
- 5-member districts would still limit voters’ choice to find the candidate they truly like the best.
- STV can lead to excessive party splintering into tiny single-issue groups. Our current situation with just 2 parties isn’t healthy, but neither is it a good idea to become like Israel where the biggest parties hold under 25% of the legislature; in that situation, tiny parties can hold the country hostage.
- Doesn’t help voters organize and gain the power of collective action.
This last point is worth discussing further.
PLACE uses an STV-based system for vote transfers. So if you, as a voter, think that your favorite candidate will probably win a seat, PLACE and STV give you the same voting power. The difference is in who decides where your vote is transferred if your favorite candidate doesn’t win. In STV, each voter makes that decision by themselves, using a ranked ballot. In PLACE, that decision is made by combining your favorite candidate’s endorsements with the vote totals from the other voters in your party.
It might seem as if keeping the transfer decisions with each individual voter leaves more power with those voters. But in practice, it’s hard work to rank every candidate in terms of how well they agree with you on the issues you care about. Inevitably, once they get past their very favorite candidate, some STV voters will end up ranking based on fame more than issues.
But in PLACE, even a candidate who is likely to lose can still act as an agent for their voters. They can go to other candidates, especially those in their same party, and say: “my voters care about issue X; what can you promise them? Why should I endorse you?” In effect, PLACE unifies, and thus strengthens, voters’ bargaining power.
This same issue means that PLACE could be even better than STV (and thus far, far better than FPTP) at representing not just ideological minorities, but also ethnic or other demographic minorities. This could help boost turnout and spiral into a virtuous cycle.
Mixed Member Proportional (MMP) methods
This is the PropRep method used by countries like Germany and New Zealand. First, most seats are elected by district, using slightly fewer and larger districts than in FPTP; then, the leftover seats are allocated by party to balance the proportions. Various MMP systems use different methods to decide which candidates get the extra party seats; in the US, it would probably be whichever district candidates had come close to winning.
Advantages:
- Gets proportionality with a simple ballot.
- Has worked well and proved popular in other countries.
Disadvantages:
- Still requires redistricting, which could slow down implementation significantly
- Many representatives would have “safe seats” and be difficult to hold accountable even if they became badly corrupt.
- Polarization would continue because the hyper-partisan primary electorate still plays an outsized role.
- Creates two different kinds of representatives, which seems contrary to American egalitarian ideals.
- Gives far less power than PLACE does to ordinary voters to determine the ideological path of their party.
Other?
There are of course other PropRep proposals out there. It’s impossible to deal with all of them here. But here’s a few possibilities:
- Other widely-used methods like open list or closed list give too much power to party insiders rather than ordinary voters.
- Theoretical proposals like “Reweighted Range Voting” are fascinating for a geek like me, but probably too clever for the real world; I think they open too many possibilities for complicated voting strategies.
- “Asset voting” is a great idea, but not directly compatible with our current constitution; PLACE is about as close as we can get.
Possible objections and responses
Complex and untried?
It’s true that PLACE voting is a new proposal that hasn’t been used in other countries, and that it’s not easy to fully explain. But it’s based on well-understood ideas in voting theory such as vote transfers and partial delegation. And from a voter perspective, the choices it asks you to make are simple and straightforward.
Also, one big advantage of PLACE is that it’s easy to simulate what it would have done in past elections, using simple assumptions. Showing that it would have given reasonable, fair results in the past can help people trust that it will do so in the future.
Still, I don’t want to gloss over this issue. Though I don’t think it’s insurmountable, the “complex and untried” issue is certainly the biggest hurdle that PLACE voting faces.
Too much power to candidates?
When they first hear about PLACE, some people’s reaction is that they don’t want to give a candidate any power to decide what happens to their vote if that candidate loses. There are three simple responses to that:
- Fine, you don’t have to. If you object to delegating your vote, PLACE gives you the option to vote for any candidate without delegating. (See the boxes on the right in the sample ballot above.)
- Why don’t you want to? The idea of voting for a legislator is that they’ll use the power you give them to represent you in the legislature. If you trust them to do that, why wouldn’t you trust them to represent you in the voting process too?
- Why wouldn’t you strongly want to? Delegation consolidates voters’ power to their favorite candidate. Thus, before the election, candidates can act as agents for their voters, offering endorsements in return for useful promises. As with a union, this “collective bargaining” power actually increases the voice of minority voter groups.
Too much power to parties?
PLACE voting has an explicit role for parties in the vote-transfer process: a vote will always be transferred to any same-party candidate before it goes to any other-party candidate. Although that’s a pretty limited role for parties, some people think that it must be taking power from voters.
But ending gerrymandering is a huge win-win. By recovering the power that had been spent on wasted votes and stolen by gerrymandering, PLACE has plenty to go around. There can be a role for parties and still increased power for individual voters.
In particular, giving voters a guarantee that their vote will stay in the party if possible helps them make better-informed decisions with less need to worry about details. It might not fit exactly with what some highly-engaged voters would want, but it does a good job of serving the majority of voters who just want their voice to be heard.
Balance between major and minor parties
Some people view PLACE through a partisan lens. Is it good for the Democrats? Or, on the opposite side: is it just a plot by the Democrats to keep down the Greens?
PLACE voting strikes a balance between major and minor party power. It’s a lot better for minor parties than FPTP is, but still better for major parties than other PropRep methods like MMP can be. For instance, in a simulation of the 2017 elections in British Columbia, PLACE would have given the third-largest party, the Greens, 10 seats, while the actual FPTP system gave them 3, and MMP would have given them 14. I could go into more detail on this, but basically, I think that PLACE is a good compromise on this issue.
Other related issues
PLACE isn’t a panacea. Here are some issues it doesn’t address:
- The disproportionality, and unfair Republican advantage, in the Senate (statehood for DC and, if they want it, PR!)
- The Electoral College (National popular vote interstate compact!)
- Wasted votes and two-party domination in single-winner elections such as Mayor, Governor, or President. (Approval voting! Electology.org is the go-to organization for these issues.)
- The Supreme Court (This is a hard one, but I think Democrats should be considering court-packing.)
- Slow internet. Sorry, people, PLACE can’t do everything.
Immediate steps: what you can do
Share this diary
I’m an activist on this issue, but I’m also one of those people who doesn’t do Facebook, and probably wouldn’t be very good at it if I did. Unfortunately, as far as I can tell, not being on FB makes it a lot harder for diaries to reach the rec list. So please, if you think I’ve made good points here, share, retweet, like, etc.
Read more
I and others have written about this stuff on Medium, Twitter, Reddit, and a bit on Quora. There’s even a Slack channel for discussing this kind of issue.
Talk to relevant organizations
As I’ve mentioned above, if this plan is going to succeed, it’s absolutely critical that it have some amount of support from activist organizations like FairVote, the League(s) of Women Voters, Common Cause, the Brennan Center, as well as from think tanks and funders like the Center for American Progress, Pew Charitable Trusts, the McDougall Trust, etc. I’m not saying that these organizations all have to go all in on this idea, but at least they have to not be actively undermining it. If you have pull with any of these, please contact me. (You can send a message via dkos, or figure out my name from the links in the “read more” section above and email me at firstname dot lastname at gmail).
Spread the word
… and thanks for reading.
