PLACE voting explained
Update: Thanks to Matt Yglesias & Vox for linking this article from their piece on Proportional Representation. If you want a more comprehensive look at #PLACEvoting, you can check out the PLACE FAQ; if you want a light-hearted intro, there’s the Caveman Dialogue.
According to Decision Desk HQ, if US House elections were held today, Democrats would get 54% of the vote but just 47% of the seats. That’s an enourmous gap. It’s mostly due to partisan gerrymandering, though demographic clustering also contributes.
PLACE voting is a proportional representation (PR) voting method — that is, a solution to gerrymandering that also handles demographic clustering. It’s designed to bring the advantages of PR — including not just a closing of that gap, but also increased turnout, improved minority representation and gender balance, reduction in mudslinging, and even a boost to grassroots organizing—without giving up any of the advantages of the current voting method—including simple ballots, local representation, clear and direct chain of accountability to constituents, an ability to vote out even entrenched party insiders if a scandal happens, and an incentive for “big tent” parties with a comprehensive platform rather than just single-issue splinter parties.
Like first past the post (the usual name for our choose-one plurality method when it’s used to elect a legislature), PLACE voting uses districts, and ensures one winner per district. But it does so in a way that ensures proportionality, wastes fewer votes, and represents minority groups while still encouraging “big tent” parties.
- Before the election, candidates can officially endorse each other. From the perspective of a given candidate, that divides the other candidates into 4 tiers: same party endorsed, same party not endorsed, other party endorsed, and other party not endorsed. (Independent candidates could use the same number tiers without considering party.) If they endorse members of the same party, they must endorse at least 2; this slightly discourages fake candidates who are running purely as vote-funnels.
- In the election, voters either choose a candidate from their district, or write in one from another district. Write-ins are done using error-resistant codes from a list available in each voting booth. Voters may also simply vote for a party.
- Each ballot is transformed into a preference order, using the endorsements of the chosen candidate, combined with the initial tallies. Within each tier of endorsement, preferences are ordered by initial vote tally. Say you voted for candidate A, who had endorsed B, C, D within their party; was in the same party as E, F, G; and had endorsed H, I, J outside their party. Say also that the initial tallies were in order of English letter frequency (more votes for the letters with lower scrabble scores). Your filled-in preferences would be A, D>C>B, E>F>G, I>H>J.
- All but the strongest candidates in each district are eliminated. Any candidate with under 25% of the local votes is out, and those votes are transferred. (In the unlikely case that no candidate reached 25%, the strongest one would survive.) Usually that leaves 2 per district, but occasionally it will be 1 or 3. The overall effect of this step is to ensure that each district is happy with its main candidate and to discourage smaller splinter parties.
- Votes from the eliminated candidates are transferred, and the elimination/transfer process continues until all winners are chosen. This uses a well-studied process known as “Single Transferrable Vote” (STV) which ensures that each winner gets a full quota of votes with no more than 1 quota wasted overall. So if there were 9 districts in a state, each winner would have at least 10% of the statewide vote (90% of the average district), with less than 10% wasted. STV is slightly modified here to ensure that there’s exactly one winner per district.
- Each winning party splits any districts that they did not win as “extra territory” for their winning candidates. Thus even if you are in the minority in your district and the local winner is from a party you hate, you will be a constituent from the “extra territory” for some winner from a party you support more, and they will listen to your petitions.
This method has several advantages.
- Ballot format is basically familiar, and quite simple. The only possibly new aspect is that each party, in addition to the local candidate, has a party line with a write-in slot.
- By the same token, it’s easy to simulate the result of past elections, assuming that voters would have voted similarly to the actual election. This can reassure both voters and politicians that the results are reasonable — basically the same as FPTP, only more proportional.Note that since gerrymandering tends to favor Republicans overall, a more-proportional result will also be more Democratic.
- Candidates with a particular connection to a particular community or set of issues could pull in votes from across the state. This would encourage higher turnout and increase minority representation. Even if a particular candidate like that got not enough (or significantly more than enough) votes to win, any votes which didn’t go to elect them would be transferred to other candidates they’d endorsed. Thus they could use those endorsements to make sure that their votes continued to support candidates sympathetic to their community or interest group. Thus, the voting method itself would help minority groups organize and strengthen the voice of their leaders.
- Representation is still local, and this method is compatible with all existing US law (including the 1967 federal statute requiring single-member districts).
- Candidates are individually accountable to voters, and voters in practice have the power to throw out even well-ensconced party insiders.
These advantages above are all on top of the basic advantages of any proportional method: no gerrymandering and fewer wasted votes.
PLACE voting is a very new proposal. The basic ideas have been around for a long time (mostly about a hundred years), but PLACE-like combinations of those ideas have been around for only about 5 years, and PLACE itself is just months old. So it’s worth mentioning a few older proportional representation methods too.
- In mixed member proportional (MMP) methods like Germany’s, only some winners are chosen by district, while the rest are chosen nationally to even out the party proportions. This is a great method, but it would be hard to make it work with our federal system (different states of different sizes), candidates are less personally-accountable than in PLACE, and voters might not like having two different kinds of representative.
- In single transferrable vote (STV), super-districts of around 5 seats each are drawn up, and voters rank the candidates in their district in preference order. It has good proportionality and accountability, but ballots and voting are very complex, and vote-tallying must be centralized (which decreases security).
These methods would both be good if they could pass. But since they would represent a bigger change from FPTP than PLACE would, they’d be harder to pass. In particular, incumbent Democrats would probably be nervous about losing their seat because of the change, something that simulations show would not be likely under PLACE.
Still, it’s worth mentioning that Rep Beyer (D-VA) has introduced a bill to require all states to use STV for US House elections. This is obviously not going to pass in a Republican congress, but it’s good to start the conversation. I think that he should look at proposing PLACE next year.
Here are some other articles about PLACE voting and voting reform activism: