Trial Adoption with a Review Date, Not Standing Majority Rule or a Two-Thirds Threshold: New House Rules in Housing Cooperatives and the Members Who Voted Against Them

Abstract

We cluster-randomized 54 housing cooperatives, with 3,412 member households, to one of three ways of adopting new house rules for 18 months. In "trial" cooperatives (18), a rule passed by simple majority took effect for six months, with signs agreed at the vote that would prompt changes; at a review meeting, members then voted to keep, amend, or drop it. In "standing-majority" cooperatives (18), a rule passed by simple majority stayed in force until repealed, and in "two-thirds" cooperatives (18) the same applied but a rule needed two-thirds of votes cast. Trial cooperatives adopted 46 of 61 proposed rules (75.4%), against 42 of 58 (72.4%) and 35 of 57 (61.4%), after a mean 23 minutes of discussion per proposal before the vote, against 28 and 29. Members who had voted against an adopted rule rated the cooperative as taking their view seriously at 4.9/7 in trial cooperatives, against 3.9 and 4.2, and 58.5% of them attended at least half of general meetings in months 13 to 18, against 46.3% and 50.2%. Rule-related complaints in months 13 to 18 ran at 6.1 per 100 households in trial cooperatives, against 7.4 and 6.9. Trial cooperatives spent more meeting time on rules, and members who had supported adopted rules rated themselves more frustrated at re-arguing settled questions.

1. Introduction

Housing cooperatives make many of their own rules: quiet hours, the use of a shared garden, where bicycles go, whether a dog can live in a flat. Each rule is decided by members who will live next to each other afterward, and a member who loses a vote does not leave the building. If the losing side stops coming to meetings, the cooperative is governed by fewer of its members.

Two common answers pull in different directions. A simple majority settles questions quickly but can leave a large minority living under a rule they opposed, for as long as it stands. A higher threshold protects minorities but blocks more proposals, some of which most members wanted. Both treat a vote as permanent, which means members try to foresee, before voting, every way a rule might go wrong.

A third possibility lowers the cost of being wrong instead. A rule adopted for a trial period, with an agreed review, can be changed without a fresh repeal campaign. Members may need less certainty before voting, and the people who voted against it know there will be a point at which their concerns are heard again. We compared this arrangement with standing majority rule and with a two-thirds threshold.

2. Method (Summary)

Fifty-four housing cooperatives took part, with between 24 and 140 member households each (3,412 in total). Cooperatives were randomly assigned, stratified by size, to one of three adoption rules for 18 months: trial (18 cooperatives, 1,152 households), standing-majority (18, 1,121), or two-thirds (18, 1,139). The assigned rule applied to new house rules only; changes to the cooperatives' constitutions followed their existing procedures. As a condition of taking part, all cooperatives recorded votes on house rules by named ballot, which allowed members who voted against an adopted rule to be identified. Dissenters were defined by the original adoption vote only; votes at trial reviews were not used.

The process ran as follows. First, a member or the board proposed a rule and circulated it at least two weeks before a general meeting. Then, at the meeting, members discussed it and voted. In trial cooperatives, a rule passed by simple majority took effect for six months. At the same vote, members agreed on two or three signs that would prompt a change, such as a number of complaints or a pattern of use, which the board then recorded. Six months later, at a review item on the general-meeting agenda, the board reported on those signs and members voted to keep, amend, or drop the rule; a kept rule then stood until repealed. In standing-majority cooperatives, a rule passed by simple majority stood until repealed. In two-thirds cooperatives, a rule needed two-thirds of votes cast and then stood until repealed. In all three groups, any member could propose a repeal or amendment through the usual process.

Proposals, votes, amendments, and meeting attendance came from cooperative minutes and attendance registers. Discussion time per proposal and time spent on house rules at each general meeting came from timed minutes taken by a trained member at each cooperative. Rule-related complaints came from each cooperative's complaint log; two researchers, blind to condition, coded whether each complaint concerned a matter covered by a rule adopted during the study. In month 18, members completed a survey. Those who had voted against an adopted rule (1,098 members; 833 respondents, 316 in trial, 296 in standing-majority, and 221 in two-thirds cooperatives) rated whether "the cooperative took my view seriously even when the vote went against me" (1 = strongly disagree, 7 = strongly agree). Those who had voted for an adopted rule rated how frustrating they found "having to re-argue questions I thought were settled" (1 = not at all, 7 = very).

3. Findings

All 54 cooperatives completed the study.

Trial cooperatives received 61 rule proposals and adopted 46 (75.4%). Standing-majority cooperatives received 58 and adopted 42 (72.4%), and two-thirds cooperatives received 57 and adopted 35 (61.4%). The trial adoption rate was 23% higher than the two-thirds rate. Discussion before the vote took a mean 23 minutes per proposal in trial cooperatives, against 28 in standing-majority cooperatives and 29 in two-thirds cooperatives.

By month 18, 12 of 46 trial rules (26.1%) had been changed: 9 amended and 3 dropped, 10 of the 12 at the six-month review. In standing-majority cooperatives, 9 of 42 adopted rules (21.4%) had been amended or repealed, and in two-thirds cooperatives 8 of 35 (22.9%).

Members who voted against an adopted rule numbered 412 in trial cooperatives, 389 in standing-majority cooperatives, and 297 in two-thirds cooperatives. Their month-18 ratings of whether their view had been taken seriously were 4.9/7, 3.9/7, and 4.2/7. Trial ratings were 26% higher than standing-majority ratings and 17% higher than two-thirds ratings. In months 13 to 18, 241 of the 412 (58.5%) attended at least half of general meetings, against 180 of 389 (46.3%) and 149 of 297 (50.2%).

Complaints about matters covered by rules adopted during the study ran at 6.1 per 100 households in months 13 to 18 in trial cooperatives, against 7.4 in standing-majority cooperatives (21% higher) and 6.9 in two-thirds cooperatives (13% higher).

General meetings in trial cooperatives spent a mean 37 minutes on house rules, against 29 in standing-majority and 31 in two-thirds cooperatives. Members who had voted for an adopted rule rated their frustration at re-arguing settled questions at 4.1/7 in trial cooperatives, against 3.3/7 and 3.4/7.

4. Discussion

The simplest explanation for the dissenters' ratings is that trial cooperatives dropped more of the rules they disliked, so fewer of them were living under a rule they had opposed. That does not hold up. By month 18, trial cooperatives had amended or dropped about a quarter of their adopted rules, only slightly more than the other two groups, and most of the changes were amendments rather than repeals.

What changed was the members' position at the point of losing. In a trial cooperative, a member who voted against a rule knew that in six months the board would report against signs the meeting had agreed, and that there would be a vote on whether the rule had worked. Their concerns were scheduled for a hearing, whatever the result of that hearing. That seems to have been enough for more of them to keep attending, and complaints about matters covered by the new rules were lower too, although the complaint logs cannot show who was complaining or about whom.

The shorter discussion before each vote fits the same reading. It is consistent with members needing less certainty before voting when they knew the rule could be revised after six months of experience, though we did not record what was argued. A two-thirds threshold did not buy the same respect, although it was designed to protect minorities. It blocked more rules, and minorities who lost had no scheduled point at which their concerns would be heard again; reopening a rule meant starting a repeal campaign.

Members who had won paid for this. They had to defend a rule a second time at a review meeting, and meetings ran longer. Trial adoption asks the majority to treat a win as provisional so that the minority stays involved, and their frustration ratings show that this cost them something.

5. Limitations and Future Directions

Named ballots were a condition of taking part, and members of cooperatives willing to accept them may differ from members elsewhere; the ballot itself may have changed how people voted. Fifty-four cooperatives is a small number of clusters, and a few large cooperatives contributed many of the dissenters. Trial rules adopted after month 12 had not reached their review by month 18, so 39 of the 46 were reviewed during the study; each rule was reviewed only once, and we cannot say whether a cooperative holding many reviews over several years would tire of them. Trial cooperatives also had more adopted rules in force, which would tend to raise their complaint rate, since complaints were not counted per rule. Discussion times came from minutes taken by members who knew their cooperative's condition. Members could have voted against more than one rule, and the rating asked about their experience overall. Complaint rates depend on members' willingness to complain as well as on compliance. We did not record dissenters' attendance before the study, so with 18 cooperatives per group, differences that existed beforehand cannot be ruled out. Under a two-thirds threshold, members who voted against an adopted rule were at most a third of voters, a smaller minority than under majority rule, which complicates the comparison of their ratings. A follow-up will test a single review at 12 months instead of six, to see whether the benefit to dissenters survives a later review.

Editorial Note

This report underwent CfAHC's Rapid Review Protocol, including AI-assisted literature cross-checking and a two-reviewer human sign-off.