A Workable Proposal for Proportional Representation in Rhode Island’s Elections

I have, over the course of nearly a decade of writing about RI politics, advanced a number of reforms to voting in Rhode Island that are aimed at moving us towards fairer representation. I started, as many such reformers do, as a proponent of ranked-choice voting (or “instant runoff” as it was introduced to me). To me, it’s blazingly simple: rank your candidates in order of preference, and then eliminate the candidates with the fewest “1” votes, and divvy up their supporters 2nd choices (and so on until one candidate achieves a majority).

I’ve since grown more skeptical about the effectiveness of ranked-choice voting, especially when there’s only one winning candidate (a standard feature in most American elections above the municipal level). I’ve tried to push more for proportional representation (PR), with my favored choice being mixed-member proportional (which elects seats from districts, but then also elects a second set of seats to balance out the district results according to voters’ party preference).

There are simpler proportional systems, too. Some places just ignore districting altogether, and dole out seats according to votes for a party. But the most common feature in proportional electoral systems is some version of a party list (there are some other methods, but the other major method is a version of ranked-choice voting).

The party list is simply a list of candidates in the order that they will be selected based on how many seats a party gets. So if Party A wins 5 seats, Candidates 1 through 5 each get to hold office, but Candidates 6 and on have to sit around hoping the next election will be better. There are two versions of list PR systems: closed-list, where parties get to pick and rank the candidates; or open-list, where voters get to pick and/or rank the candidates. There are a lot of variations between various nations’ electoral systems, but this is broad strokes.

The problem for proportional systems in Rhode Island is that parties don’t really enter into it when it comes to voting (at least not since the elimination of the straight-ticket option). Oh sure, you may vote for the Democrat, but you are not actually voting for the Democratic Party. You are voting for Gina Raimondo or Jack Reed, an individual who the Democratic Party has authorized to use their ballot slot. In Rhode Island (and American elections more broadly) we vote for the candidate, not the party.

This has allowed the proliferation of things like the top-two nonpartisan “primary” (I put primary in quotes because it defeats the purpose of having a primary at all). Because we think in terms of individual candidates, not parties, it makes it easy to advance reforms that further undermine parties. It also makes it difficult to address problems like partisan gerrymandering, since we can hide behind individuals’ successes and failures (by votes the General Assembly in RI should be anywhere from 33% to 40% Republican, not 12%; even Providence should have one or two Republicans on the Council, not zero).

RI also has restrictive laws that prevent parties from forming, or if they do manage to form, from maintaining their ballot line for more than one or two cycles. These laws were created purposefully to ensure that no other parties but the Democrats and Republicans would get elected.

This legal and electoral antagonism towards parties makes it hard to dream about any sort of proportional system. How can you have party lists when we vote for individuals? How do you make a primary system work with proportional representation? If we shoot down modest reforms like ranked-choice voting for being too complex, how would we respond to list systems?

Luckily, smarter people than I think about this, and seem to have come up with a workable solution that makes open-list proportional representation work in America with minimal headaches for election officials and voters. I’m calling it the Santucci-Shugart Plan (EDIT: more accurately, it’s “a’ plan, since I introduce new elements), after the two political scientists who iterated it: Jack Santucci and Matthew Shugart (Santucci does a lot of research on proportional systems used in America in the past and Shugart actually has gone to other countries and helped develop their electoral systems so they’re quite experienced with this). While it’s intended for the United States House of Representatives, I think it would have more interesting implications locally.

EDIT: Jack Santucci points out to me that I missed key points of the proposal, such as preserving the partisan primary and leaving it so the list is open. This ends up being a closed list system, as opposed to Santucci’s open list system (Shugart notes that closed lists often generate opposition; I think this is just open enough at the front end). As Santucci points out, ironically, it incorporates much of a reform I dislike: Alaska’s recently-adopted nonpartisan top-four system. So despite me dubbing this a “Santucci-Shugart Plan”, there’s very little Santucci left in it!

Santucci-Shugart works like this: it’s 2022, it’s the primary. You go in to vote in the Democratic primary in East Providence, in House District 63. You vote for Governor, Lieutenant Governor, etc., and then you get to Senator. There aren’t just Democrats on your ballot. There are Republicans and Independents too, maybe a Green or a Libertarian. You’re a Democrat, so you pick your favorite Democrat. You go to Representative. Same thing, you see all the candidates, you vote for your favorite Democrat anyway (let’s say Katherine Kazarian). You move on. Mayoral election is back to just Democrats again.

So why is everyone on the same ballot for the General Assembly? Well, because this system is going to use lists. You vote for just one candidate, but per law, X number of candidates with the most votes are going to get to advance. This is going to create our effective maximum number of lists. If it’s four candidates going on, we can expect a maximum of four lists of candidates in the general election. So, the four candidates with the most votes get to advance to the general. If you set it at three, you’ll restrict it to three lists. If you go more, more lists. For the example’s sake, let’s say it’s set, by law, to four. And to save me some headaches, let’s just look at the House (a parallel process would play out much the same way in the Senate).

So you just voted for Katherine Kazarian who does get to go forward, but District 63 also advances Robert Britto (another Democrat), Christopher J. Holland (a Republican), and David Sullivan (an Independent). What now?

Well, all these Districts have been combined into groups of five across the state. It looks like this:

This is just an example using our current House districts. I’ve arbitrarily picked the groups and assigned names, so don’t read too much into it. They can always be altered, and the districts can be redrawn according to changes in population.

So what happens now? Well, District 63 is combined with Districts 64, 65, 66, and 67 in the North East Bay District. The old House districts (which we’ll now call nominating districts) have sent on four candidates each (well, except for District 65, which only had three people run) to this larger electoral district. Now, this is where the fun begins.

After the primary, but prior to the general, candidates from the nominating districts must now join together to form a list to compete in the electoral district. As Shugart lays it out, he would restrict lists so that you have to have a candidate from each nominating district, but in RI, we very quickly run into issues where we only have Democrats running, so I think it would be best to leave that up to the lists to decide. Maybe they want two candidates from high-turnout nominating districts, and will take a pass on having a candidate from a district that lopsidedly votes for another party.

The point is, these are all the sorts of calculations candidates will have to make. Placement on the list is another; that will determine the order they’re seated in. The first candidate on the list is the one likeliest to win a seat in the House. The last is the least likely. Voters will know this, so maybe you put a strong vote-getter at the top. Or maybe you put someone with a strong field operation there, so they’ll work hard. Or maybe the five candidates all agree that the top candidate should have the strongest policy chops, so they’ve picked that one over some newer candidates.

If candidates can’t agree how a list should be composed, or there aren’t enough candidates, or they don’t get on any lists, they’ll also go onto the ballot, just not on a list.

This list is what will appear on the general election ballot.

So let’s talk about who came out of the North East Bay District. Here’s all the candidates who made it through the primary (nominating districts are listed in parentheses):

With twelve candidates, the Democrats have more than enough candidates to form a list. If we restrict things to one party gets one list, then the Democrats are kind of screwed. They have so many strong candidates that list negotiations are going to be very tough. But in America, we elect people, not parties. So, thread the needle. Allow anyone to join a list regardless of affiliation, with the number of lists in an electoral district not to exceed the number of candidates elected from a nominating district (in our case, arbitrarily four). With twelve candidates, the Democrats will be able to form two lists of five.

This also helps the Republicans. They only got three candidates through, so that’s not enough to form a list made up completely of Republicans. But a Moderate and a Libertarian also got through, and those parties are (or in the case of the Moderates, were) close enough to the Republicans that they could plausibly all run on a joint list.

With a particular set of candidates in a particular electoral district, you could also imagine bipartisan lists, with both Democratic and Republican candidates. This might be likely in an area which elected a large number of new progressive Democratic

So it’s the 2022 general election: you go into vote. You vote for the federal offices, the statewide offices, and then you get your General Assembly races, you look at the state representative race and you see this:

I based this off the sample ballot PDFs distributed by the state Board of Elections, so this wouldn’t be too far from what it might look like. In the case of the first two lists, everyone is from the same party, so candidates don’t get their party affiliation listed under their name. The third list is multiparty, so each candidate gets their affiliation listed. I did this for its compactness, but you could imagine officials wanting uniformity in the names, so it could also just list affiliation for everyone. Underneath the lists you have candidates who are running outside of a list, but for voting purposes, they basically exist as a one-person list.

What will happen is that voters have just one vote, as they do now, which they can cast for a list or an unlisted individual candidate or for a write-in. Proportionate to how the votes are cast, lists will receive seats. So if a list wins 95% of the vote, they win all five seats. But more likely, competitive lists should be in the range of 20% to 75% (which would be very dominant) of the vote, shaking out to anywhere from one to four seats.

Though my lists are purely notional, you can create some rationale for why they are like they are. For the Labor & Progress Democrats, they’ve gone with a seniority system, so those with the longest time in office are highest on the list. The Jobs & Economy Democrats, meanwhile, have gone with a “who wants it more” order, with former representative Helio Melo mostly just there to lend his support or step up in case of an overwhelming victory. Meanwhile, the center-right Clean House List has had to make a difficult negotiation. Though its voters are going to be mostly Republicans or Republican supporters, to secure the participation of the Moderate and Libertarian candidates, they’ve had to give them the #2 and #3 slots, meaning even a surprising 60% victory is only going to deliver one seat for the Republicans.

You could imagine some other calculations. Rep. Amore is really a key player here; the only seriously credible candidate from District 65, he might’ve been tempted away from the Jobs & Economy List when Labor & Progress decided to give him first place (and instituted their seniority ranking to make it “fair”). Meanwhile, without Amore, the Jobs & Economy List has left the two other District 65 Democrats off their list in favor of bringing in East Providence Council President Robert Britto.

The unlisted candidates aren’t really out of the race either. All one of them needs to do to get elected is achieve 20% or more of the vote; it’ll be a tough lift, but it’s not impossible.

A real complication would be the exact formula you use to figure out how seats are allocated, but they basically boil down to whether or not they favor the lists with most votes for the last seat. I imagine RI would opt for a method that does favor such lists.

From an administrative standpoint, it’s the list negotiations that are maybe the hardest part of this. Since RI’s primaries happen in September and ballots need to be printed shortly thereafter, General Assembly members wouldn’t have very long to negotiate. This could be rectified by moving the primary to August, as has been argued for recently. Having a deadline to file lists creates a nice fire under the feet of candidates, as they have to negotiate fairly quickly or face the prospect of going it alone.

Those negotiations could get prickly. The fifth candidate on a list is very unlikely to get elected. In more traditional list systems, you tend to see these slots filled by celebrity candidates. But under Santucci-Shugart, candidates have to make it through a nominating round. It might be relatively hard to get your celebrity candidate through that round, especially if it was evident that’s what they were. On the other hand, if only four candidates run in a nominating district, it may not be such a problem. For politicians who genuinely want to represent their constituents, it might be extremely frustrating to have to take fifth place, and you could see them wring policy concessions out of their colleagues if they were in a place where the list wouldn’t be able to be submitted without them.

I don’t think that kind of drama would be that bad. It would give a lot for media to cover, and lead to a lot of interesting conversations about priorities. The advantage of this system is that it doesn’t punish Democrats for having a lot of candidates; they can form multiple lists. It gives space to Republicans to represent areas they otherwise might be locked out of (such as urban areas). It allows small parties and independents a much better chance to get into office. It both weakens parties a bit, but also allows them ways to maintain their strength. It doesn’t entirely break politicians from local representation either, while still implementing a more proportional system. Importantly, it doesn’t significantly alter the way voters vote; they still just fill in a single bubble. And it can be adapted to smaller governments, too. Providence could do three electoral districts electing five seats, for instance (which might see some Republicans or possibly a Green elected).

You can also see the advantage of the way it might break the nationalization of politics. You could see lists that share names across the electoral districts (though I imagine there would need to be some regulation of that). By allowing multiple lists of Democrats (or Republicans in places where they’re strong enough) to compete against each other, you could see state issues be more vigorously debated in the general election.

I think this proposal really offers the best chance towards a proportional system, one that looks like it can mostly be accomplished without drastic constitutional changes in RI (you would potentially only have to rewrite Article IV, Section 2 to allow for proportionality in elections and to allow groups of people).