Town Meeting does not mean more democracy or representation

This guest post was written by Ray La Raja, a professor of political science at UMass, and graduate student Wouter Van Erve.

Supporters of Town Meeting (TM) argue that it provides more democracy and better representation. We cast doubt on these claims, based on research and statistics in Amherst.

TM is more democratic only if you believe that the participation of 240 citizens counts more than participation of voters. Town Meeting government depresses turnout. In Amherst, average turnout from 2011 to 2016 has been 11 percent of registered voters (lower if one uses eligible voters). Across Massachusetts it is 21 percent for council governments. If Amherst switches to council elections, it can expect a similar turnout increase, which means roughly 3,800 additional voters would participate in elections. People participate more when they experience political campaigns, which engage them and encourage them to vote. Not so with discreet TM elections in which most voters have little or no familiarity with candidates.

An additional bonus of higher turnout is that the voters will be more reflective of the population. Our research shows that the small slice of residents who show up for TM elections mirrors the members of TM (older, homeowners, white, educated). One of us has written a dissertation that studies turnout in Massachusetts towns, which shows that electorates for councils are more representative of the entire town. Like any American election, it’s still not perfect, but much better than TM electorates.

In sum, the argument that Town Meeting creates more democratic participation is misleading. One could easily argue it creates less participation (especially for people who don’t look like Town Meeting members). We acknowledge there is a trade-off: do we want intense participation by the few? Or common participation by the many?

One key to choosing between this trade-off is to understand whether Town Meeting is more or less representative of the electorate that isn’t participating. At least TM might reflect the silent population that does not participate. Despite claims by supporters that TM is more representative of the town than other forms, our research suggests it is not. Political scientists like to distinguish between descriptive representation (officials looking like voters) and substantive representation (officials supporting policies preferred by voters). On both kinds of representation, TM has questionable outcomes.

We have addressed the highly skewed nature of descriptive representation in past commentary. We cite a few here: The median age of eligible voters in Amherst is 34; in TM it is 63. Home ownership among Amherst residents is 55%; in TM it is 91%. These differences can be meaningful because they often lead to substantively different policy choices. For example, TM members who voted against funding for building a new school tended to be older by several years and less likely to have children in the home.

Under a healthy democratic system, even if descriptive representation is not great, you should get decent substantive representation because your official should have an electoral incentive to represent your interests. This is especially true on salient issues like schools, which affect a large portion of the community. A majority of TM members voted against funding a new school even though a referendum showed a majority of voters approved it.

However, we are less concerned about the outcome of the school vote than whether TM members actually represented their precinct voters. On this score, we were troubled to see large discrepancies across precincts in how TM members voted and what voters in their districts preferred. In five out of 10 precincts the TM members voted substantially against voter preferences. The two most egregious cases involved Precincts 2 and 10. In Precinct 2, 49% of voters approved school funding, but a whopping 86% of TM members did so as well (a difference of 37 points). On the flip side, in Precinct 10, 53% of voters approved the funding, but only 19% of TM members approved it (a difference of 34 points).

You might ask, how well are TM members representing their constituents? It is certainly plausible to argue, like John F. Kennedy, that sometimes leaders need to make tough choices that put them at odds with their constituents. The leader, after all, has information and a broader perspective the voters may not. It would be a stretch, however, to say that TM members who went against their constituents were profiles in courage, even when acting on their understanding of the public interest. For this to be so, TM members who opposed the preferences of their constituents (assuming they knew what they were) had to fear losing the next election. That is not the reality of TM elections, since turnout is so low and the individual threat of losing a seat virtually nil.

As voters contemplate whether to support the charter reform, they should ask, compared to what? The strongest case that has been made for the TM status quo is that it is more participatory and more representative. This is a dubious claim. It is “participatory” if one is referring to the small slice of residents who are in TM and the 11% who vote for them. In reality, TM depresses broader participation, and the body itself is not representative of what Amherst voters look like. More critically, important policy decisions run the risk of not being representative of what residents want because there is no way to hold TM members accountable for decisions that reject voter preferences.

TM members should worry more about what voters think, including those they don’t know personally. That is unlikely to happen without genuine elections that come with a council system of government.

We believe a council government has many other benefits over TM, including greater transparency of public decision-making and a deliberative process that encourages compromise. Our purpose here was to illustrate that, properly understood, its participatory and representative functions are as good or better than Town Meeting.

Comments 5

  1. I’m reading this blog post with the current Amherst Bulletin (3/23) under my laptop. One of the top stories features the three residents competing for two open seats on the School Committee. One is a custodian and represents the younger, more diverse, low-income neighbors who are left out of the TM system. Interestingly, he opposed the school building project, and he opposes the charter, but his voice is one we need to hear––as are the voices of his two contenders, both moms and one of them a 36-year-old with a doctorate in urban policy/planning from MIT, a TM member who supports the charter. I feel very fortunate this morning to have both this blog and the Gazette story land in my lap, the one informing the other, and the other confirming the one. I hope I am not guilty of cherry-picking this input to fit my preconceived notions, but I do find the La Raja and Van Erve statistics very self-affirming as I prepare to vote for the Charter…and for the young and diverse candidates in our community. Thank you!

    1. No one can complain that there hasn’t been enough thoughtful writing for an undecided voter to consume in this campaign. I defy anyone to find a community that could generate this much interesting, provocative commentary from people who were or have been largely unpaid volunteers. We care about our town.

  2. This piece would have more power if it presented all the data, not just cherry picked numbers.

    Look at this statement: “A majority of TM members voted against funding a new school even though a referendum showed a majority of voters approved it.” Okay, so in the November vote 46% of voters voted in favor of the school project, 45% voted against it and 9% abstained. It barely passed–by about 130 votes of more than 13,000 cast. Then in fall Town Meeting: 46% voted against it and 45% voted for it. Does this vote by Town Meetings seem off-the-charts unrepresentative of the voters? No. And even if 46% had voted in support of the override, the vote would have failed — since Town Meeting had to by a 2/3rds majority. . And this then it failed to get 2/3rds vote among regular voters last March. There is no way to reconstruct any of these votes to get 2/3rds needed.

    Similarly, to get Amherst town voting down to 11%, you have to ignore the last 10 years of Amherst votes–which would get one to 17%. A March voting rate of 17% is not so bad considering the fact that most of Amherst voters are students voting almost exclusively during November elections. Voting rates in March elections is driven by contested seats-so the answer to this ‘problem’ is likely to have more candidates and contested seats, not changing the government.

    And how are council-only form of government more representative of the electorate given the abysmal numbers on women and people of color? What are the actual numbers on this? Well, Amherst Town Meeting has 56% women and the 9 council only governments are around 25%. Don’t the numbers of women count when looking at whether councils are representative?

    And what about incumbency? 91% of incumbent councilors win re-election. So once a councilor is in, how much does he have to worry about what the voters think?

  3. It’s really not all that complicated. 87% of Town Meeting members run unopposed. What is more democratic: 13 Council Representatives elected by all the voters, or 208 Town Meeting Members elected by one vote, their own?

  4. I appreciate my neighbor Ray La Raja and grad student Wouter Van Erve’s effort to talk about inclusion and democracy, because that’s my primary concern and interest in local politics as well. However, I have a few points in response.

    (1) Turnout in Town Meeting. Picking the set of 2011-2016 and ignoring 2010 and 2017 is interesting, because 2017 and 2010 were both high turnout years — 22% last year and 32% (the ten-year high) respectively.

    (2) Turnout in Council municipalities. Here’s a problem, because it turns out (pun intended) that turnout in MAYOR-Council municipalities is considerably higher than in Council/Manager municipalities. The 21% averages those numbers together, but it’s 18% for Councils without a mayor as the Charter proposes. 27% if we had a mayor.

    (3) Expecting additional turnout to the tune of 3800 voters? This is pure speculation. In fact, in the one town that recently switched from representative town meeting to a council, Randolph, which provides enough time for voting history, voting turnout DECREASED, going from 27% to 19% in its manager-Council form.

    (4) More fundamentally, this elides the confounding effects of the college age population, which routinely register in large numbers in Amherst to participate in higher-profile national or state elections, and participate in town elections in vanishingly small numbers. When you account for that effect, Amherst actually has pretty good turnout relative to neighboring towns.

    (5) Who’s representing whom. It’s absolutely true as the authors acknowledge in the third paragraph, that the members of TM reflect the voting population — for instance, in age. It’s surprising then to see in the sixth paragraph a confusing switch from actual voters to “eligible” voters, which includes the large number of Amherst residents who as students choose to register in their own home towns. Let’s be clear about when we’re discussing “eligible voters”, versus “registered voters”, versus “actual voters”, and why we use one category versus another. That has real implications in any locale, and particularly in a college town with a large number of short-term part-time residents.

    The home ownership numbers would exactly reflect the same age artifacts created by a large college student population. Any discussion that is hoping to shed light needs to carefully consider these differences.

    (6) On the school funding votes, the numbers are clear: In the November 2016 town-wide election, the vote for and against was split 50/50, as it was in Town Meeting a week later. In early 2017, there were two additional votes, and Town Meeting again reflected the population, with Town Meeting approval (57%) slightly ahead of the town (56%). I’m not sure why the turn to precinct-by-precinct data is relevant for this town-wide ballot, or when one would decide to look at precinct votes instead of town votes. Town Meeting members across town were contacted by passionate residents on both sides of the issue, across town, regardless of precinct.

    I’m also not sure why one wouldn’t pick the votes that are closest in time to compare to each other — November 2016 town-wide to November 2016 TM, for instance, or January/March 2017 votes. Most importantly, the fact that the two bodies have different criteria and different voting thresholds for approval seems highly relevant: The first town-wide vote was simply a vote about whether to permit a tax limit override, with a majority threshold, IF the spending was approved. All the other three votes were votes about whether to actually authorize funding, with a 2/3 super majority needed for approval. The last two votes, in Town Meeting and town-wide, were 57% and 56%. The town had the final say about whether to approve the spending, and town-wide, it did not.

    One final point. If we consider which bodies are “representative”, the 45% to 50% of town residents who opposed the project were represented by only one member of town-wide office. This striking homogeneity in viewpoint that stems from a lack of diversity and representativeness of the town’s politically connected classes, a group which has been difficult for people of color, renters, and students to break into. There is little reason to believe from any evidence produced in this article that that quasi-monopoly on power will loosen at any point. The Finance Committee, a town-wide committee appointed by the town moderator (an elected position) came a little bit closer to reflecting townwide views on this project — two abstentions and one vote against the project. But on a divisive issue with a set of clear town-wide votes for comparison, what body was closely representative of the town’s perspectives? Town Meeting, within a percent.

    (7) In terms of incumbency and accountability, it is an unfortunate reality in the US that incumbency is an enormously strong factor in whether or not someone will retain a contested seat. However, the authors make a conclusory statement about TM, without providing comparative data for councils. Council elections in Massachusetts return 90% of incumbents in contested elections. Are many of the Council elections contested? No—in fact 60% are uncontested.

    That said, however, “contested” is a metric that works well and understandably in a race for one seat. In a race where there are pools of representatives, you have to look a little deeper to see overall turnover. In fact, when seats are NOT contested, you can get more churn and fresh faces more frequently. Imagine for instance a precinct of 24 representatives, with 12 that stick around for a bunch of terms. The other 12 are shifting in and out as their interest and availability waxes and wanes, or new people come in. That’s basically what we have currently. If you look at contested seats using the metric of “more candidates than seats”, Amherst’s TM is doing pretty good: 35% uncontested seats (65% contested).

    And, by the way, when people DO want to run, as they did last year in the highly contested 2017 TM election, turnover happens. At least 8 people who voted against the building project were unseated, two of whom were people of color.

    (8) Which brings me to another topic, and perhaps the one that is most important. Throughout this piece, there is a lot of discussion of Town Meeting being less representative than the general population. There is zero discussion about the fact that Councils, Select Boards, and legislatures — all bodies with a high representative-to-population ratio — are by far less representative. In Amherst, we have not had a single person of color on the Select Board in 10 years. This is true of Select Boards around the state — white and male. How male? Really male. Women average 25% of Select Boards, as they do every other legislative body in Massachusetts and around the world. (Congress is worse, unsurprisingly.) What is the ONE EXCEPTION? Representative Town Meeting, which has 42% women across all RTMs in Massachusetts. If you just look at the larger RTMs, like Amherst has, it’s 45% women. So the low barriers to gaining access to Town Meeting seats provide an exceptionally open and inclusive environment for women. While data on racial/ethnic and class make-up is harder to get, we know that Amherst’s TM includes a significant number of low-income residents, renters, and people of color. We can contrast that with the Select Board, which has historically had almost all property-owners — no students at all that I can find.

    This exclusivity is particularly problematic in the Council model, which will require thousands of dollars to run for a seat ($6000 average for contested Council seats across Massachusetts), shutting out most people who are not connected political insiders.

    Does Town Meeting, like all forms of democratic participation, suffer from the effects of race, class, and gender? Yes. How can we solve these problems, reducing the barriers that keep so many people out of the governing, politically-connected classes? Will voting yes to the Charter, and moving to a system that has MORE of these problems, fix the problems we currently have that are endemic in society at large? No. What could work? Well, I worked with a group of people this fall interested in building on the strengths of Town Meeting — the accessibility and low barriers to participation — and we generated a group of candidates who are markedly more diverse across gender, race, and class, than the YES candidates, the current Town Meeting, or any other body in the area. Looking just at the YES candidates who publicly support the Charter: Only 45% women, two students, one renter, a property value about $50k more than the Amherst average —and zero people of color. The contrast with the candidates who are opposed to the Charter is stark: Of the 75 TM candidates identified who oppose the Charter, 63% are women, 14% people of color, 16% are students or college age, and an astonishing 26% are renters. The average property value among homeowners is $334k, not far from the average home value in Amherst of $323k.

    That’s what building on strengths looks like. This Charter will destroy a system that is strong in many respects, and able to be stronger, and replace it with a Council model that consolidates power in fewer hands — a model that fails across the nation in terms of any standard measure of accountability, accessibility, representativeness, and diversity. It is my hope that people who care about diversity will look at the numbers openly, and critically, and not be blinded by empty sloganeering or wedge issue politics.

    Laura Quilter
    Librarian, Attorney, Town Meeting member

Leave a Reply

Your email address will not be published. Required fields are marked *