Medford City Charter: Why I support four district plus five at large councilors.
Background: What are we talking about?
First, let me orient people to the discussion. Many people in Medford have hoped that we would revise our city charter. The charter is like the constitution of a city, it determines things like how many city councilors you have, whether you have a city manager, etc. The Charter Study Committee (CSC) forwarded a draft of a potential new city charter to the Mayor, who edited it and forwarded her draft on to us. This is the appropriate process. The part of the process we are in now is where elected city councilors edit that draft (this part of the process is required by state law).
The City Council discussed the charter in three public Governance Committee meetings, then in a public Committee of the Whole meeting, and on Tuesday (today) it is coming to the city council meeting for final approval before going back to the Mayor. The version that the CSC drafted recommended eight ward city councilors and three at-large city councilors. The CSC recommended a different system for the School Committee: four district school committee members (where each district includes two wards) and three at-large. The Mayor did not change either of those. One of the edits the City Council made was to suggest that the city council should also have four district city councilors, and then have five at-large city councilors (at first it was the same as the School Committee, four districts and three at-large, but in order to increase the number of city councilors it was amended to five at-large).
Okay, now hopefully you understand what has happened. The argument that has become incredibly polarized is between whether we have eight ward councilors and three at large, or four district councilors and five at large. Now for my opinion.
Anna’s opinion: a more even balance between localized and at-large is best.
I’ve been interested in voting systems for decades. I studied the mathematics of voting systems while getting my undergrad degree in mathematics, and I studied inequality in urban planning grad school. I have been interested in how these intersect ever since. I’ve had basically the same opinion of this question for probably 25 years. It’s okay to change your opinion based on new information, but I happen to have thought a lot about this for decades and I haven’t changed mine. At a recent committee of the whole, one of the Charter Study Committee members reminded us that when I interviewed with the CSC over a year ago, I expressed reservations about ward representation. 8 wards and and 3 at-large is 73% ward councilors, and only 27% at-large councilors. I am in favor of a more balanced, closer to 50/50% split between ward representation and at-large representation.
People have said they don’t understand the reasons against a ward-based system. I think the pro-ward argument has been better articulated (many of these arguments are in the CSC final report), and so I’ll do my best to help people understand what I believe are the real problems with ward representation.
What science says about different voting systems
If you look at both national and international organizations that study voting systems, there is a consensus. Multi-seat systems are better than single seat, and within single seat elections first-past-the-post is one of the worst. Let me catch you up on what these terms mean. *
Single-seat vs multi-seat:
In a single seat election, people in a localized geographic area (usually called a district) only get to vote for one person (who is usually required to live in that district), and only one person represents those people. So ward representation is single-seat representation: anyone who lives in Ward 1 can only vote for the Ward 1 councilor, and there is only one winner in that ward. Other city councilors are not supposed to represent them; they are supposed to represent the constituents in their own ward.
In a multi-seat election, people in a localized geographic area (like a city or state) get to vote for multiple people, and multiple people represent them. At-large representation is multi-seat representation. Everyone in Medford currently votes for seven people, and the top seven vote-getters become city councilors. Every city councilor represents everyone in the city, and constituents have choices about whom to reach out to on the city council about their issues.
Multi-seat elections are associated with:
Legislatures that more accurately reflect voters’ political preferences
Districts contested by multiple parties and candidates
Governing by a coalition of parties rather than a single majority party
Greater gender parity in legislative office
Single-seat elections are associated with:
Uncontested districts and two-party systems
A lack of proportionality between votes cast across the country for a party and seats won by that party
Governing by single-party majorities
The election of fewer women to the legislature
Now on to the types of elections you can have in multi-seat and single-seat elections.
Proportional vs Winner-take-all
All single-seat elections are by definition winner-take-all elections.
But multi-seat elections can be proportional or they can be winner-take-all. Think of the Democratic and Republican presidential primaries. In the Democratic presidential primary, delegates are allocated proportionally to candidates based on the statewide vote. So if you win 40% of the vote, you get 40% of the delegates. That is proportional. In the Republican primary (most states), All of a state's delegates available in a contest are allocated to the candidate who receives the majority of votes in a state primary election or caucus. So if you you get 40% of the vote and your opponent gets 45% of the vote (and third parties get the remaining 15%), you get nothing and your opponent gets all the delegates. That is winner-take-all.
Proportional representation is associated with:
Greater likelihood that each constituent has at least one elected who agrees with them on policy
No gerrymandering is possible
Minority representation tends to improve under proportional systems by allowing groups to win representation in proportion to their numbers, regardless of where they live.
Proportional systems tend to be more competitive: with more seats in contention per district, more parties and their candidates are incentivized to compete.
Because multi-winner races create space for more than two parties, proportional representation tends to produce more fluid coalitions, which research finds helps to temper polarization.
Proportional representation systems tend to have less extremism in the general population, and
Tend to have elected officials who problem solve with people from other parties or platforms, and maintain democratic norms.
Winner-take-all is associated with:
Greater likelihood that a constituent’s policy beliefs are not represented by their elected official
Gerrymandering can only occur in winner-take-all systems
Winner-take-all elections uniquely disadvantage racial, ethnic, religious, and other political minorities, especially when they do not live in geographically concentrated areas and with district lines deliberately drawn around them.
Because winner-take-all elections make it easy for a single party to dominate in a district, they tend to depress political competition. As soon as a party can count on 55-60% of the vote, a district becomes “safe.” Except in a small number of swing districts, competition shifts to low-turnout primaries where candidates tend to be pulled to the extremes.
Winner-take-all systems tend to produce two-party systems, which are more likely to increase affective polarization (also called “negative partisanship”) — meaning voters from opposing parties don’t just disagree with one another, but come to reflexively distrust and dislike one another.
By definition, winner-take-all elections are high stakes. Marginal differences in support for either of two parties can mean total victory or total defeat. Politicians are often incentivized to do everything they can to beat their opponents, even at the expense of problem solving, good governance, or maintaining democratic norms.
Voters and politicians who lose in winner-take-all elections are less likely to trust democratic institutions, and more likely to resort to violence.
Protect Democracy says:
Researchers are especially concerned about the use of winner-take-all elections in highly polarized and diverse societies like the United States. As one global study of democratization concluded, …winner-take-all electoral systems “are ill-advised for countries with deep ethnic, regional, religious, or other emotional and polarizing divisions.”
I want to point out that negative partisanship is an indicator that strongly correlates with failing democracies. The rise of negative partisanship in America is a very troubling sign.
And now for the third distinction, different voting methods among single-seat districts. I’ll only look at three, since these three are the only ones Americans have any experience with.
First-past-the-post vs Runoff elections vs Ranked Choice Voting
In single seat elections in America, the one you most often see is called “first-past-the-post” (FPTP). Whoever gets the most votes wins. Even if you have five candidates and the highest vote getter only gets 25% of the vote.
A runoff election is where if you have five candidates running, and one of them gets more than 50% of the vote, they win. But if they get less than 50%, there is a runoff between the top two vote getters.
Ranked Choice Voting (RCV) is also called Instant Runoff Voting. Voters rank the candidates in order of preference. You are essentially asking the voter to tell you their vote and their runoff vote on the same ballot, which is why it is called “instant runoff voting.”
First-past-the-post is the most likely to lead to minority rule, it leads to two-party systems, it exacerbates the affect of “spoiler” candidates, and it leads to more polarized elections. Runoff elections help mitigate most of those, but have two major flaws: they are expensive, because now you have to run two elections, and they depress voter turnout, because far fewer people vote in the runoff. RCV is considered by all experts to be the best of these three systems.
Conclusion:
For the record, single-seat, first-past-the-post elections are known, according to experts/science/data, to be among the worst possible methods for getting people appropriately represented in politics. This bears repeating: Single-seat, FPTP elections, otherwise known as ward or district seats, are known, according to experts and academics, internationally and nationally, to be among the worst possible methods for political representation. **
Now, I’m going to guess that a lot of progressives heads are spinning around right now, because that’s not what they have heard in progressive circles. Progressives have been told that wards are better than at-large.
So why, in America, does research show that ward-based is a “better” system? Two main reasons. One is the distinction between identity based representation and policy based representation. The other is historical redlining.
What is representation?
In America we have a skewed understanding of what representation actually is, especially in the last 12 years (when identity politics has become more overt). People in America think that to be represented, you elect someone similar to you. To have women’s issues represented, you elect a woman. To have Black, or LGBTQ, or immigrant issues represented, you elect someone from one of those groups. If you want to be represented in your city, you elect someone who lives in your ward.
This is a fundamentally different understanding from what people in many other countries understand to be the essence of representation. I believe (and many people around the world also believe) that true representation is when you elect someone who is going to vote the way you want them to vote on policy, who will vote for the policies that genuinely help you and make your life better.
The argument for wards is based entirely on an identity politics view. This is why most of the time, in cities all across America, and I believe in Medford until very recently, it is typically Democrats/progressives who want more wards, and Republicans/pro-establishment people who want at-large representation.
Research on wards only ever looks for outcomes on identity. Ward representation research also exists in the reality of American cities that have had historic red-lining: forcing minorities to be concentrated in certain districts.
From an identity politics perspective, wards holds up well in research in cities that are highly geographically segregated — people of historically under-represented identities are in fact elected more often in these cities. From a policy perspective, especially a class policy perspective, wards don’t hold up so well, though I will say, I don’t think the research even asks or answers that question.
And when minorities are spread throughout the population, ward representation doesn’t even do a good job of bringing in different identities; in fact it does the opposite. Multi-seat representation does a better job of electing representatives from minority communities than single-district seats. This is important for progressives in Medford to understand, because historically Medford had more racial segregation in prior decades than we have now, so while ward representation might have been more helpful from an identity politics perspective 20 or 40 years ago, it won’t be as effective now.
How can we make our at large seats better representative of our residents? How can we get a better democracy?
Multi-seat proportional representation systems are the best. At-large is multi-seat, so how do we make it proportional? For a variety of reasons (America has a two-party system, MA and our region of MA is overwhelmingly from only one of those parties, our political parties do not have binding platforms, and our state mandates non-partisan municipal elections) we can’t write this into a charter. But we can make it happen. It would happen if candidates ran on a small number of political platforms. If we had 14 candidates running on two, three, or four political platforms, this would essentially be proportional representation.
For the last eight years much of my work has been centered around advocating for policy platform-based slates of candidates running in municipal elections across the country. Hopefully after reading this you understand why I spend my time doing this — I genuinely believe it will re-invigorate democracy in America. And just for the record, policy platform-based slates of candidates can still do well and get better representation in hybrid systems and even in all-district systems. Recently a slate won 10 out of 14 seats in an all-wards city of 275,000 people.
A couple of final points:
Most cities in MA have larger wards than we do. Somerville and Lowell have wards that are twice our size, in other words, equal to the size of the districts in the current city council proposal.
I find it quite frustrating that the dialog in the city council meetings has become “8+3 = good and righteous and virtuous, 4+3 or 4+5 = evil, power hungry, and dishonest!” This is what is meant by polarization, and America has grown more polarized, which is also not good for democracy. It would be healthier for our community to tone down the vitriolic rhetoric.
There is really not an enormous difference between the two, and this is exactly the kind of polarization that single-seat FPTP elections cause. I hope after this post you understand that I am voting for what I believe is in fact the best voting system we can put in the charter based on the wealth of information I have gathered over the last 30 years and what we are allowed to do in a non-Home Rule Petition charter.
* Most of my quotes are going to be from Fair Vote (https://fairvote.org/) and Protect Democracy (https://protectdemocracy.org/), but you can also look to orgs like the ACE Electoral Knowledge Network (https://aceproject.org) or the International Institute for Democracy and Electoral Assistance (https://www.idea.int/) and many others for other opinions and data.
** To help people understand a bit more viscerally, just look at the 2016, 2020 an 2024 presidential elections. These were 51%-49% elections, and it’s pretty clear that 49% of the country was incredibly unhappy for four years each time. So imagine that you live in a ward and you are one of the 30% or 45% or even 60% of people whose candidate did not win. Your one and only representative (if we were all wards) would not represent you. In the 8-3 model you would at least have three at-large city councilors, which is why having some at-large is important.
