Today, I have three part-time jobs. I have not ever held a tenure-track position, but I have come close. Here are some reflections on that career trajectory.

1. Like it or not, we are all embedded in a party system. People have partisan attitudes. Just as these affect how non-political scientists process information, they may affect how political scientists process what they believe are the implications of your work.

2. We are at a point where attitudes about electoral reform work similarly. I recall a job talk (early 2023) at which I presented my core findings on the repeal of the single transferable vote.  Nobody objected to the data or its interpretation. Some objected to its implications for the “good people” (I quote) promoting an entirely different form of RCV in the city where I was interviewing.

3. As the above anecdote suggests, your audience’s opinion, if it is strong, may be based on a relatively blunt understanding of the underlying phenomenon. People must understand a thing before they can engage arguments about its etiology constructively. Don’t pick a complicated thing.

4. Local politics may matter. Let’s say you are in a place where a powerful politician has come to understand his/her success as tied to the thing you study. That will give him/her a direct interest in what you say about it and/or the research questions you ask about it. I am fairly sure that a Dean once nixed hiring me because he feared a mayor who recently had gone from opposing to supporting RCV. I cannot rule out that others saw the work as unlikely to attract external dollars.

5. Stay away from practitioners. We often hear about “need” for “engagement” between academics and practitioners. In my experience, this mostly comes down to trying to control research questions and interpretations of facts. Not complying will get a target put onto your back. Also, the “bigger” the venue, the bigger and/or more numerous the target(s) can become.

6. Avoid distractions. Your job is to teach and especially publish journal articles. Ignore or at least discount advice to seek media attention and/or practitioner engagement. Even if your institution says it wants these things, your advancement odds probably have more to do with perceived flight risk, which is a function of publications.

7. Don’t share ideas with people unless you trust them fully. They will use the information to their perceived benefit, which is fine unless it also puts a cost on you. I made this mistake specifically in the context of seeking a senior colleague’s approval.

I know that the above is negative in tone, but I lived it. All responsibility is mine. I wish I knew in the past what I know now, but I also could have made different choices at critical junctures. My goal has been to show what some of those junctures can look like.

The political system is not in good shape. All of us are flailing to some extent as a result. Be careful, and good luck.

A zombie vote is one that cannot count because the chosen candidate has left the race. RCV’s promise here is to reduce zombie votes and therefore boost the nominee’s legitimacy.

Zombie votes came up in response to my speculation that the DNC had permitted states to use RCV as a party-diversification measure. (The phrase I used was “to make more candidates viable.”) At the time, the ranked ballot was widely seen (and marketed) as a way to reduce entry costs for non-traditional candidates. This claim was so widely believed that it became a foil for one of four “RCV” articles I know to have entered top-three journals of political science.* (Maybe there are others.)

Alas, readers commented that the goal was instead to manage zombie votes. This seems reasonable. Several 2018 congressional primaries were noteworthy for having been highly fragmented, inexperienced candidates have been entering Congress at greater rates since 2016, and the Sanders faction may have been perceived as uniquely cohesive.

As an aside, IVN reported today on zombie votes in the Wisconsin Democratic gubernatorial primary (for which RCV is not used).

Close readers may be forming an impression that RCV and candidature are unrelated on average. I am led to agree, but the race might not be average.

If one is worried about zombie votes, the following questions arise in my mind.

Does the legitimacy of a presidential nominee flow from their vote share? I don’t know, and this is not my area, but I suspect there are bigger factors.

If the answer is yes, does the ranked ballot encourage more candidature than fragmentation alone? Two sources I know say yes but only in rare elections (like the first or with some shock to the respective party system).

If we go with yes on the other two questions, we enter a discussion of exhaustion. I don’t think we know enough about what causes that in the U.S.

I am a fan of symmetry, so here’s a riff on the Republican process. On that side, delegate allocation also follows closed-list rules but without proportional representation. Fans of this system might say it produces a nominee with fewer transaction costs (time, bargaining, etc). Detractors might say the benefit is more legitimacy. 🤷‍♂️

Have a good weekend. I hope it’s nice outside where you are.

*I changed the wording of this sentence after publication from “became the basis for” because I did not want to imply the article supports the claim.

My aim in this post is to sketch some potential reasons from the perspective of a political scientist. They are not exhaustive. Rather, the post is meant to push the conversation in an analytic direction. Supporters of this class of reforms have attributed the move to personalities, and its opponents have tended to be quiet. I’ve argued for roughly a decade that debates thus structured do little to advance understanding of an issue that, gradually, is being forced onto analysts of U.S. politics: whether to keep or change American electoral institutions and, if so, exactly how.

My working argument is that the potential ban reflects understanding and acknowledgment of recent changes in the party system. It also may reflect deepened understanding of how RCV accelerates political fragmentation.

RCV as used in Democratic presidential primaries is best understood as a device to make more candidates “viable.” This contrasts with the usual form, used in New York City and Alaska and other locales, which promises to make more candidates “viable” and find a majority winner at the same time. The crucial difference is the national Democratic Party’s use of a proportional delegate-allocation rule. RCV in New York City, etc., is the so-called “instant runoff” version for selecting single winners. In Democratic presidential primaries, it is instead “bottoms-up” — candidates are eliminated until all remaining have 15 percent or more of primary votes. The resulting vote shares then determine how many convention delegates each candidate gets: proportional allocation.

RCV in Democratic presidential primaries thus combines two fragmenting devices: proportional representation at the nominating convention, used since 1984, and the ranked-choice ballot, used in five states since 2020.

Contra to what some might think, electoral systems do not cause fragmentation immediately or mechanically. The coalitions that impose them — reform coalitions, in political science terms — often reflect high degrees of internal cohesion at the outset. Rules we might consider fragmenting (like proportional representation) often get adopted for other reasons like neutralizing gerrymandering or making sure that an expected coalition partner can win sufficient seats — or delegates, as in the case of the Democrats’ national convention.

Over time, however, the coalition that imposed the rules can change. This is the main reason why RCV in myriad forms was abandoned by U.S. cities during the Progresssive and New Deal periods. It is also why the most famous critic of proportional representation saw it as a road to chaos in Weimar Germany. Research literally shows the dissolution of the so-called Weimar coalition in the mid-1920s over issues of redistribution. Analysts looking back at the situation from the mid-1930s therefore saw more years with “fragmentation” than “effective” government and blamed the rules.

Why would some states’ Democratic parties have wanted to make more presidential candidates viable? I have not been in the rooms where such matters are discussed, but the timing suggests we should look back to 2016 for answers. In that race, Bernie Sanders nearly defeated Hillary Clinton for the nomination. The contest also became notable for party elites’ efforts to enforce message discipline and, as some have noted, paint the Sanders camp as racially insensitive. RCV meanwhile had become a cause célèbre among leftists, some of whom also believed open primaries would let them court white social conservatives otherwise on their way out of the party.

Several political scientists have documented the attendant realignment. This work includes a recent book-length analysis of party politics during the Trump years; analysis of vote choice by education, income, and race; and a research note in which I showed both parties becoming more internally divided on class issues from 2012-16.

There have been two further changes in the party system since 2020. One is the partial departure of non-black, non-white voters from the Democratic tent. These people, according to political scientists John Sides and Michael Tesler, report more conservative views on race and racially tinged policy issues than those who stayed.

The other change, of course, has been a bona fide socialist challenge to the Democratic Party establishment. We saw the first salvo last summer and fall in New York City. As I write this, several more are playing out in the tabulation of results for U.S. House primary elections (without RCV).

The Big Apple’s run with RCV-in-primaries also may have reinforced a perception that it lets initially weak candidates stay in a race for longer than they would otherwise. Witness Zohran Mamdani’s early standing in the polls. Further, there is no guarantee that such candidates’ supporters will coalesce around other options. The best available analysis of RCV in the 2020 presidential primaries gives prima facie evidence that Sanders supporters were less likely to rank choices.

In sum, the Democratic Party’s leadership has more reason to worry about its coalition than it did in 2020, when it permitted several state parties to experiment with a reform whose effects might have seemed innocuous.

Fragmentation is becoming a real problem for the American political system. Whether due to leadership failure or the stakes of the issues involved, normal channels of party competition and lawmaking so far have not been able to produce settlements on several hot-button issues. One result has been to turn to executive power. Another has been haphazard “experimentation” with allegedly novel electoral systems. In this context, it may be helpful step back from old reform commitments in favor of a systemic perspective on events like the RCV “ban” we may see this weekend.

Years ago, before graduate school, I studied to join and worked in a business called “democracy promotion.” Its purpose was to bring democracy to formerly authoritarian regimes and, in places where that already had been done, advance its “consolidation.” Two mantras from those years now stick out in my mind:

“Healer, heal thyself.”

“There is no perfect system.”

Those who used the first were trying to solve a legitimacy problem. It’s hard to persuade others to do things one won’t do themselves. At the time, the main concerns were gerrymandering and elite polarization. While neither yet had metastasized to current levels, they apparently were problematic enough to serve as a source of national embarrassment.

The second mantra — no perfect system — referred to the tradeoffs one encounters when designing political institutions. These should be familiar to anyone who reads my little missives. My experience with the phrase was through practitioner-oriented essays and slide decks. In these, it was customary to present a list (or table) of “pros” and “cons” that might flow from various institutional choices. I don’t know whether these lists shaped decisions because I never actually sat down with the deputies of a foreign government to tell them how to change a constitution or electoral law. My job was just to keep track of the changes from an office.

It is hard these days not to be a little pessimistic. The healer now faces real domestic pressures to heal itself. The lists of “pros” and “cons” have become both more numerous and important. It feels like we’re flailing.

What next?

You can use the data to generate figures like these. They did not win friends because they raised questions about the reigning explanation for STV’s repeal (racism, anti-communism).

There were two other academic projects to have produced similar data. The most recent was Barber et al. (1995). The other is much older and the subject of a working paper. Neither led to a public archive.

I hope the data will be useful to junior scholars, undergraduates, and others entering this research area. Good luck!

I want to suggest that cities like New York and L.A. instead consider increasing the sizes of their councils. Doing so might take some pressure off of Democratic primaries — or what passes for primaries in cities with nonpartisan elections.

Consider the matter from the DSA/WFP perspective. Picking off an establishment Democrat in a low-turnout, first-round election is easier than running as a third-party candidate in a general election.

Political institutions in cities like New York and Los Angeles have fallen out of step with their underlying party systems. This is producing acrimony all around: efforts on the right to paint these politicians as enemies, efforts on the left to paint electoral reform as a conservative plot. It’s classically winner-take-all (as much as I dislike the term for describing an electoral system).

Nasty first-round elections (and/or primaries) are a sign of demand for representation. City-council expansion is the inclusive way to confront it. The alternative is to continue moving in the direction of nonpartisan elections. This doesn’t solve the problem and, for cities like L.A., isn’t even possible because elections already are nonpartisan.

Proportional representation also is worth considering. Here is a new piece in the City Journal Substack about how and why that might work well in New York City.

Separate from the simulations, the dynamics of the Democratic primaries speak to the potential value of having party factions compete as separate entities. PR’s rival reforms seek to manage fragmentation by funneling voters’ preferences through some form of two-round election.[23] These proposals take SSDs as a point of departure because many believe that they deter third-party entry.[24] Instead, the key intervention is to make it easier for voters not affiliated with a party to affect the candidate it sends to a general election (i.e., primary reform). This assumes that a party’s regulars also are its “extremists.” However, the Working Families Party won 23 council seats by winning the respective Democratic primaries. Data from the NYC Campaign Finance Board (CFB) suggest this may have been due to new-voter mobilization rather than the outsized influence of party regulars. Working Families saw a 157% increase in registrations, followed by 78% for Democrats, 53% for Republicans, and 32% for Conservatives. New registrations concentrated among voters under 30 (64.8%). Further, CFB notes that single-day registration before the deadline beat 2021’s record fivefold.[25] This surge occurred in a context of closed party primaries, which are widely thought to impede participation by voters not tied to either major party.[26] “Moderation” via primary reform therefore may require voter mobilization to counteract whatever is its target.

The usual case against PR is that it might break up existing political parties. The New York City party system is already fragmented.[27] When we treat candidates’ combinations of party endorsements as parties in their own right, the effective number of electoral parties is 3.3. Note that this situation obtained under the current “first-past-the-post” system. It also obtained in a system that should be expected to produce 2.7 parties without any electoral reform.

Have a look!

There’s another issue to flag about the relationship between these two versions of PR in the rhetoric around reform. I’m concerned that advocates for the second version of PR will draw on the intuitions behind the first version of PR to gain support for the second version. In other words, support for the first version of PR will become a Trojan Horse for the second version.

Two points are worth making:

1. The U.S. House of Representatives already is expected to produce effectively 2.8 legislative parties and 4.6 electoral parties. Where are they?

2. What if the second rationale were precluding serious consideration of the first?

I otherwise agree with Prof. Pildes’ analysis of PR adoption prospects:

But there’s also the political economy mountain that seems especially insurmountable with this version. I think most reformers envision unified Democratic control of Congress and the White House as required for any change of this sort. But is the Democratic Party, having managed to attain that level of control, likely to agree to fragment itself into three different parties, along with however many parties the Republicans would fragment into?

For these and other reasons, I recommend minority-party representation (MPR) as an anti-gerrymandering device.

This week’s news has been about the potential for new Republican district maps to reduce the number of Black members of the U.S. House of Representatives. MPR can be designed to address that problem too.

One solution is to use closed (or maybe flexible) party lists to allocate the “at-large” seats. Here’s an example of how those work.

Let’s say Party X ends up entitled to three at-large seats. It has fielded five at-large candidates. There are many ways to decide which three candidates get the seats. A “closed-list” procedure does not let the general election alter the order of the list. Therefore, as long as the target candidates are at the top of the list, a closed one will get them into office.

There would be other details to discuss.

My purpose in writing is not to advocate. It is to show that there are options.

Here is an alternative called minority-party representation (MPR). I call it that because its main effect would be to give more seats to the largest underrepresented party in each state. Start with some modest addition to the overall size of the House.* Apportion these seats to states in proportion to their populations. These seats then become ‘at large’ in any state that gets them. At each election, allocate these seats to parties in declining order of their deviations from seats/votes proportionally, accounting for their seat numbers at each stage of the process.

Thus described, MPR would reduce the effects of gerrymandering and ‘natural’ packing/cracking under the current SSD [single-seat-district] regime.

*No such addition is necessary if members are willing to give up their single-seat districts, or if others in their party are willing to make that decision for them.

What if the lesson used an example students care about? What if it were interactive?

As you see above, a website now lets users explore the 2025 New York City Council results under different electoral systems. It also covers measures used in comparative politics to analyze the result of an election. The numbers below came from the existing electoral system. 3.3 “effective” parties is a lot!

Check out The Bigger Apple below for more information on the site and the 2025 results in general.

The Bigger Apple

Explore New York City’s Emerging Three-Party System

New York City’s recent council elections suggest a de facto three-party system, driven by cross-endorsements and growing competition on the left and right, that is straining rules built for two parties…

Read more

4 months ago · 11 likes · Jack Santucci

Some years ago, I joined forces with Matthew Shugart and Mike Latner on a blueprint redesign of our federal-level electoral institutions. The piece was a major undertaking. In it, we proposed:

1. Open-list proportional representation for U.S. House elections. The thinking was that (a) STV does not reliably produce coalitions that can organize a legislative chamber in the U.S. case and (b) MMP would require an appreciable increase in the size of the House. (An excellent new paper explores that scenario, in fact.) Further, we speculated that the ranked-ballot feature of STV might harm descriptive representation relative to open-list PR. The reason is that people might not rank things they don’t like. So we got to OLPR by process of elimination.

2. Cross-endorsement ballot fusion for the Senate and Presidential elections. At the time, the stated reason was to let the coalition that emerged in advance of a House election duplicate itself in the other single-seat contests. We were skeptical that ranked-choice voting would reliably produce such outcomes. Note that this proposal does not imply altering the Electoral College. In fact, historical examples make me think fusion could work well. The key thing to be avoided is a contingent election in the House of Representatives, and my knowledge of single-seat electoral systems says fusion would do it better than anything else on offer.

There was a tension in our paper between multiparty democracy, on the one hand, and, on the other, an observation that the current institutions were not already generating it. To wit, the seat-product model of the effective number of parties leads us to expect a U.S. House with “effectively” 2.8 parties represented. Part of the argument for multiparty democracy was that under-represented groups had gotten a “raw deal” from the Democratic Party. I don’t know if this is true, it’s not my place to say, and I had reservations about that section. More interesting to me is what I just said about multiparty democracy not already existing in the House. I strongly suggest meditating on that empirical fact.

Why the two proposals never caught on remains a mystery. Hypotheses I’ve encountered include:

• Widespread misunderstanding about PR’s effects due to baggage in the literature, sometimes reinforced by modern-day reformers’ public-facing arguments.

• Opposition from the Congressional Black Caucus, perhaps based on the same (and maybe even the relevant part of our paper).

• The sheer difficulty of “socializing PR” on the Hill, whatever that might mean.

• That the proposal isn’t a real “Senate solution” due to the persistence of staggered elections. This objection came mostly from backers of the 2024 RCV ballot initiatives.

I write this because many in my network seem to be frustrated, flailing, and angry as a result. I know what’s that like.

The good news is that we have one possible blueprint here.

In a later post, I will consider the role of bipartisanship vis-à-vis scholarship on how electoral reform happens in practice. Teaser: I think it will be needed, I don’t think it has to result in King’s “negative peace,” and that’s why I’ve been comfortable working on similar problems at the local level.

Here is a picture that stops my face from coming up in the timeline preview:

I do not know much about the Colorado initiative process. All three measures appear to be awaiting legislative hearings (whatever that might mean). All three measures come with as-filed, amended, and final versions. What I say below is based on my interpretations of the final versions.

I will attempt to describe each measure in terms of the classification scheme used in my book on the topic: ballot format, district magnitude, allocation rules, nomination rules, and assembly size. There are interesting provisions about filling vacancies, but I will set these aside for now. Ditto for provisions about ballot order and district competitiveness.

Measure 262

This aims to provide for open-list proportional representation (OLPR) in districts of five seats. There would be seven such districts for senate elections and 13 for the lower chamber. Upper-chamber elections would be staggered across but not within districts. Each party would be permitted run a slate of five in such a district, and there would be a primary permitting the voter to choose up to three candidates to represent that slate. There also would be provision for slates of “unaffiliated” candidates. Minor parties would be able to close their primaries to unaffiliated voters. Seat allocation would be by the D’Hondt method (AKA Jefferson method). Any party with at least one sixth of votes in a district would be entitled to at least one seat.

Measure 263

This also aims to provide for OLPR. It is similar to the above except that it provides for joint lists (qua “joint slates”), replaces primaries with nomination-by-party-convention, and requires the nominating process to be by OLPR as well.

Measure 264

This measure aims to provide for mixed-member PR. For upper chamber, it would provide for 25 single-seat districts and 10 statewide compensation seats. Staggered elections for this body would affect five compensation seats at a time. For lower chamber, there would be 45 single-seat districts and 20 statewide compensation seats. This would be a two-vote system such that a voter could split their ticket. The measure would set a threshold of five percent for compensation seats and allow the general assembly to adjust that rule. Nominations to the compensation seats would be by party convention. However, there are several uses of the phrase “filled by assembly,” and I don’t know exactly what this means. As for district seats, I interpret the text to provide for nominating primaries. Compensation seats would be allocated by D’Hondt based on the party-vote shares, letting any party keep seats it had won via district-based election. For the upper chamber’s staggered elections, parties’ seats from the previous election would count toward its seat share for the ongoing election.

That’s all for now. Again, this was a first cut, and it was quick. Let me know if I’ve made mistakes.

Here is a throwaway picture to prevent Substack from pasting my face behind social-media previews. The GPT prompt was: “What picture would you draw if I asked for an image that evokes mixed-member proportional representation?”

• Twice called a “fascist” in the last two weeks, once for refusing to talk politics in a bar.

• Once told I “look like a Democrat” because of the way I was dressed (jeans, 2013 Marmot soft shell, flannel shirt, mild stubble, curly hair).

• Once called a “conservative,” presumably for my “moderate” views on one particular policy area.

I am none of the above. I am a pro-parties independent. Let’s parse that phrase.

“Pro-parties” means I share a view, widely held in political science, that “modern democracy is unthinkable save in terms of parties.” Democracy involves collective action on the scale of a nation-state. I am not sure what devices other than political parties might facilitate that. One option, I guess, would be the Two Minute Hate to which we all have access on our phones. How has that been going? Should we double down?

“Independent” means I do not feel sufficiently positive toward either major party to claim it as “mine.” I feel this way because I do not trust either major party as a collective actor to behave ethically when in control of the national government. Things may differ at other levels; it’s a case-by-case situation. I came to feel that way after 20+ years of voting for Democratic Party candidates. The straw that broke the camel’s back is something we can discuss over coffee, maybe. It’s my personal straw; you do you.

There are other reasons I now call myself an independent. Some are beliefs I formed after reading and reflecting on the urban-politics literature several years ago. Also, I’ve been studying how modern partisanship works more closely than I did in the past. That has given me even more reasons to reject partisanship.

Here are a few words on the reasons I got by studying partisanship. They help me explain some of the interactions I described at the top of the post.

I think many partisans are walking around trying to determine others’ partisanship to find out if those people might be enemies. Enemy. An enemy is a person in the business of trying kill an opponent. Japan became an enemy on December 7, 1941. It stopped being one four years later.

To make inferences about friend/enemy status, one might rely on visible and auditory markers (dress, hair style, soft-spoken or not, etc). Then they might administer what now gets called a “purity test.” This involves asking questions about another person’s views on contentious issues. Lately, it seems too easy to fail such a test. One might ask why we’re at a point where the tests are being given in the first place.

Some get frustrated when I deprive them of the chance to infer my partisanship. This may be even more frustrating than “knowing” the other person is an “enemy.” What I’ve just written will not scratch that itch. That such an interaction has become frustrating is interesting on its own.

I don’t want to participate, and maybe I’m even doing the right thing by bowing out.

How should we measure “pure” or “true” independents? For years, the respective item required a respondent to volunteer that answer. Recent surveys have moved toward presenting it explicitly. Those that do produce estimates of pure independents that are much larger than in past surveys. We present evidence of this phenomenon across multiple surveys and ask: Are self-administered surveys overcounting independents, or are traditional live-interviewer surveys undercounting independents? We answer that question by comparing live-interview and self-administered samples from the 2012 and 2016 American National Election Studies, by undertaking tests to rule out mode effects (including an experiment), and by seeing which question wording correlates more strongly with measures of latent ideology, vote choice, and ratings of the parties. Our findings suggest that surveys that include an explicit response option, allowing Americans to self-identify easily as “(pure) independent,” offer a more precise measurement of the concept of party identification. This has implications for the study of independents, as well as for discussions about polarization and party-system dealignment.

This is a part of a series of projects on the state of the U.S. party system. It was a joy to work with Josh on this project over the past few years. The paper is open-access thanks to the generosity of the Provost at UMASS-Lowell.

Gamble built the house after landing in the area by boat. He held many slaves and grew sugar for the New Orleans market. Then demand collapsed. Stories of his tenure suggest he was a bad guy, paranoid about slaves stealing his food, and with little concern for other than making money.

The Confederate army then used the house until the end of the Civil War.

The next inhabitant was Judah Benjamin, who served under Jefferson Davis as both Attorney General and Secretary of War. He fled from Virginia to London via the plantation, studied, and became a commercial lawyer.

Here are photos of two monuments on the property. Seeing one of them was jarring. It made me recall a southern friend being surprised some years back by a “yankee statue” in the middle of a New England town we had been visiting. These monuments are a generation newer than the ones often seen in New England.

The place is architecturally interesting and points to challenges of life in what had been a subtropical jungle. It is made of a brittle concrete with lots of seashells in it. There had been some dirt floors, but the park service filled them to protect other parts of the property. No wood in the structure is original.

You can get a good meal across the street at Popi’s. This is a Greek-leaning diner with a big menu.  It caters to many palates.

The ruin of a sugar mill used to be toward the beach where Gamble landed, now across the street. My dad remembers seeing it years ago. A local told us it’s gone.

My next memory of redistricting involves writing a report on California. I do not remember what we had been trying to sell. However, I remember that the 1990 census had produced a “sweetheart gerrymander” (incumbent protection) whereas the 2000 exercise had been more about partisan advantage.

News of gerrymandering kept piling up. First came Tom DeLay’s mid-decade exercise in Texas. Then came the gerrymander that reversed control of the U.S. House in 2012. Now we have California Proposition 50, which asked voters explicitly whether they want to do it for partisan advantage.

The current situation has produced a mutual-disarmament case for reform. One flavor is proportional representation (PR). Another is independent redistricting. Rather than play games in states, the argument goes, wouldn’t the major parties be better off tying their hands?

I am weary of mutual-disarmament cases for redistricting reform. I have tried to conjure paths by which such change could happen. The best run up against the Senate filibuster, which is stable because it permits legislators to buck their parties in hiding.

The filibuster thus imagined means that redistricting reform must pass through states. A large literature asks whether this reform can overcome partisan advantage. It reminds me of an old debate in international development about whether an “independent electoral commission” really can be independent of the politicians that create it. That debate concerned unitary states, not federal ones, and so maybe the United States holds promise of letting regional difference find expression through maps.

Two things make me pessimistic about state-based reform. One is California Proposition 50. We still need to see who voted for it and why, but I suspect their eyes were on Washington. Another is the structure of public opinion across states. I recently went looking for the sorts of difference that might make bipartisan reform coalitions possible. Consumers of that paper told me I was barking up a dead tree.

Earlier this year, I co-wrote a report on how to go about electoral reform in big cities. The work started in late 2024. Just before, an urban-politics think tank asked me to talk about my book’s implications for this policy area. I suspect the organization was trying to decide how to approach several upcoming ballot initiatives on “open primaries” with instant runoff, and perhaps what to do on urban governance more generally. Some months later, the report came out, and I recently shared it with my contact. “Seen this?” I asked. “No! But ewwww [insert outlet],” they said.

The paper might as well have been written for a liberal outlet. Its core question was how to get majority rule without sacrificing minority representation. It also dealt with more obviously liberal themes: disparate impact of confusing ballots, sensitivity to desire in some corners to preserve single-seat districts, and the burning issue of housing supply.

The word on the street is that there will be a big push for adoption of proportional representation (PR) at the state and local levels. In that context, I periodically share my report on BlueSky. Few pay attention, and sometimes a reader will retract their “like.” I suspect the reaction would be different if the proverbial book had a different cover.

Some do read the paper, and they tell me privately that they like it.

We hear a lot about the dearth of conservatives in academia. I think this frame misses the mark for two related reasons. One is that the meanings of “liberal” and “conservative” change routinely. The other is that this sort of group identification has become central to how people process information.

One doesn’t have to be “conservative” to see that cancel culture is anti-science. Any number of perspectives can get us there: a Socratic view of pedagogy, research on information processing (see above), and canonical models of opinion formation. The good news is that we find cancel culture exactly where we’d expect: among those not socialized into using scientific rules of evidence. The bad news is that we find it among scientists too.

[Proportional representation] advocates have long believed their preferred systems lead to compromise. Portland has an opportunity to show the nation that this is true. Failure could hobble a reform movement whose ideas are worth considering.

I support PR in the abstract. It has desirable properties, depending on what it’s replacing. Chief among them is its ability to deliver majority rule and minority representation at the same time. Majority rule comes mainly by blunting gerrymandering and/or “unpacking” or “un-cracking” constituencies that cannot elect their representatives otherwise. Minority representation comes by not using the other major way to unpack/un-crack: multiple non-transferable vote where the number of votes a person may cast is equal to the number of seats in a district.

Some nonetheless interpreted my article in either of two ways:

• That STV in Portland has “failed.”

• Or that representation (for socialists) its a bad idea.

Neither is what I argued or believe. What I do believe is that:

• Now is a time for compromise-oriented leadership in politics.

• Institutional determinism lets one avoid the hard question of how to get such leadership.

Politics is the compound effect of individuals’ choices. That you and I are here to read and write this, respectively, is proof. Structures like PR matter insofar as they make certain choices more or less likely, but that’s it.

Have a look!

For those reasons, I wrote a simple explainer that avoids both topics. It also touches on the relationship between PR and “fringe” parties, why PR gets adopted in practice, the widespread view that PR means “voting for parties and not candidates,” and how to start thinking about the constitutionality of multi-state electoral districts.

The current situation in American national politics has me thinking about similarities to and differences from New Zealand in the 1980s-90s. Similarities include frequent alternation in House control and reasons for both major parties to question single-seat districts. Obvious differences include population size and parliamentary vs. presidential regime type. Those differences are worth analyzing in subsequent conversation. Much already has been written on them.

A related consideration is whether it is possible to “go back” to having parties made of discernibly regional factions. One of my ongoing projects suggests probably not, consistent with the behavioral literature on “nationalization.”

Have a look, and thanks!