Komitee Equal Shares in Participatory Budgeting

• Komitee Equal Shares is a dual-mode framework that uses both individual voter inputs and group-based evaluations to guide civic budget allocation.
• It employs virtual budgets and priceable allocation rules along with detailed voting receipts to ensure transparency and accountability.
• The framework’s deployment in Winterthur demonstrated its ability to fund a diverse set of projects while enhancing fairness and participant learning.

Searching arXiv for papers on Komitee Equal Shares and closely related Equal Shares work. arXiv search query: "Komitee Equal Shares Method of Equal Shares priceable participatory budgeting" Komitee Equal Shares is a priceable virtual-budget allocation framework for civic funding that extends the Method of Equal Shares by integrating two distinct decision signals within one proportional mechanism. In voter mode, participants cast point votes over proposals; in evaluator mode, small groups assess proposals against collectively defined impact fields. Both signals are translated into virtual spending power, and the resulting allocation is explained through project-level voting receipts. The framework was introduced and deployed in the 2025 Kultur Komitee in Winterthur, Switzerland, where it was presented as a way to combine fairness, learning, and participant-understandable outcomes within participatory budgeting and grant-making (Yang et al., 2 Oct 2025).

1. Conceptual basis and problem framing

Komitee Equal Shares was motivated by a mismatch between ordinary participatory budgeting or grant-making procedures and the roles participants often attribute to themselves in practice. The Winterthur setting was used to argue that residents do not act only as preference-reporting individuals; they also judge proposals in terms of public value, fairness, inclusion, and cultural priorities. Komitee Equal Shares addresses this by making a formal distinction between acting as a voter and acting as an evaluator, rather than collapsing both roles into a single vote stream (Yang et al., 2 Oct 2025).

This design places the framework within the broader Equal Shares lineage while shifting its emphasis. Standard Method of Equal Shares allocates equal virtual budgets and funds items only when supporters can jointly afford them; Komitee Equal Shares retains that proportional core but embeds it in a co-designed participatory process. The paper describes this as preserving the mathematical logic of a priceable allocation while surrounding it with deliberation, co-design of criteria, and explanation (Yang et al., 2 Oct 2025).

The immediate institutional predecessor was the Kultur Komitee 2024 workflow built around Human-in-the-loop MES. In that setting, participants first voted online, then explored allocations through an interactive slider that varied the budget delegated to MES, and finally discussed the algorithmic outcome in person. The 2024 process allowed vetoes or budget adjustments after MES selection; in practice no MES-selected project was vetoed, but one project, “Kultur-Agenda,” was reduced from 18,000 CHF to 9,000 CHF. That earlier workflow established MES as a practical component of Kultur Komitee before the 2025 framework internalized multiple input channels within a single priceable allocation (Yang et al., 7 Feb 2025).

2. Separation of voter mode and evaluator mode

The central conceptual contribution of Komitee Equal Shares is the separation of two decision modes. In voter mode, participants act as individuals. They receive an equal virtual budget share and distribute points across the projects they personally prefer. These point votes are transformed into normalized support weights, denoted by expressions such as $\hat u_i(p)$, so that each participant’s spending power is proportional to the support assigned to each project (Yang et al., 2 Oct 2025).

In evaluator mode, participants work in small groups focused on an impact field. Instead of registering only personal taste, these groups discuss proposals and allocate points according to how well projects serve the field. The group input is also translated into virtual spending power, but now the relevant virtual agent is the impact-field group rather than the individual alone. The paper emphasizes that the budget share of each field depends on how important that field was judged to be, so evaluator mode converts deliberative value judgments into weighted virtual budgets (Yang et al., 2 Oct 2025).

This explicit dual structure is intended to address a gap in social choice noted by the paper: participatory budgeting is often modeled purely as preference aggregation, whereas participants may also view themselves as evaluators of collective impact. Komitee Equal Shares therefore does not treat individual preference and collective value judgment as interchangeable signals. Instead, it preserves both and feeds them into one allocation rule (Yang et al., 2 Oct 2025).

3. Allocation rule, virtual budgets, and priceability

The background rule is the standard Method of Equal Shares. In the paper’s MES recap, the setting consists of voters $N=\{1,\ldots,n\}$, projects $P=\{p_1,\ldots,p_m\}$, costs $c_p>0$, and total budget $B$. Each voter starts with equal budget

$b_i=\frac{B}{n}\quad \text{for each } i\in N.$

A project $p$ is $q$-affordable if

$$c_p \;=\; \sum_{i \in N} \min \bigl( q \cdot u_i(p),\, b_i \bigr),$$

and MES selects the project with the lowest affordable price $q$, then updates budgets by

$N=\{1,\ldots,n\}$0

This formulation captures the equal-shares logic: each agent receives virtual currency, and projects are funded only when supporters can collectively pay for them (Yang et al., 2 Oct 2025).

Komitee Equal Shares extends this structure. The paper’s “Core KES procedure” takes as input projects $N=\{1,\ldots,n\}$1 with costs $N=\{1,\ldots,n\}$2, total budget $N=\{1,\ldots,n\}$3, individuals $N=\{1,\ldots,n\}$4, impact fields $N=\{1,\ldots,n\}$5 with weights $N=\{1,\ldots,n\}$6, and utilities $N=\{1,\ldots,n\}$7. Utilities are normalized as

$N=\{1,\ldots,n\}$8

For individuals, budgets are assigned as

$N=\{1,\ldots,n\}$9

where $P=\{p_1,\ldots,p_m\}$0 is the share of the total budget reserved for impact-field evaluation; impact fields receive virtual budgets according to their weights. The algorithm computes project support, builds an affordable set, selects an affordable project, and records the split of its cost through receipts. In the algorithm box, project support is written as

$P=\{p_1,\ldots,p_m\}$1

with affordable projects defined by

$P=\{p_1,\ldots,p_m\}$2

For the selected project $P=\{p_1,\ldots,p_m\}$3, the receipt update is

$P=\{p_1,\ldots,p_m\}$4

followed by

$P=\{p_1,\ldots,p_m\}$5

The paper presents this as an equal-shares-like allocation in which every funded project is fully covered by its backers and no agent spends more than its assigned virtual budget (Yang et al., 2 Oct 2025).

The framework is explicitly described as priceable. In the social-choice literature summarized by the paper, an allocation is priceable if agents hold virtual currency and spend it on supported projects so that exactly the funded projects are those that can be bought from available support. In standard approval-based committee voting, MES is the canonical priceable rule, and priceability is characterized through payment constraints that ensure selected candidates are exactly paid for while agents spend within fixed budgets (Brill et al., 2023). Komitee Equal Shares retains this priceable logic, but it applies it to a mixed set of virtual agents consisting of individuals and impact fields (Yang et al., 2 Oct 2025).

The paper also notes an important subtlety. Because impact fields are weighted unequally and because the evaluative criteria are socially co-designed, the resulting proportionality is not purely axiomatic in the way classical voter-only MES is. This places Komitee Equal Shares adjacent to, rather than identical with, the standard approval-based theory in which Equal Shares outputs priceable committees and priceability itself implies axioms such as FPJR (Kalayci et al., 21 Jan 2025).

4. Voting receipts and participant-facing explainability

A distinctive feature of Komitee Equal Shares is the use of voting receipts. The paper treats receipts not as a secondary visualization but as the interface through which priceability becomes legible. For each funded project, receipts display how much of the cost came from each individual and from each impact field, thereby turning the allocation into an audit trail that participants can inspect, reproduce, and contest (Yang et al., 2 Oct 2025).

This is a substantive reinterpretation of priceability. In the approval-based committee literature, priceability is usually an existence property of a payment system: every winner is fully paid, non-winners receive no payments, and unelected options cannot be afforded from remaining supporter budgets (Brill et al., 2023). In subsequent axiomatic work, Equal Shares is used as a rule that always outputs priceable committees, and priceability is then shown to imply FPJR (Kalayci et al., 21 Jan 2025). Komitee Equal Shares uses the same underlying notion for a different purpose. Rather than serving only as a fairness certificate in a theorem, priceability is operationalized as a participant-facing explanation mechanism (Yang et al., 2 Oct 2025).

The paper links this directly to transparency, traceability, and trust. Because a funded project can be described as having been “paid for” by identifiable supporters and impact fields, the allocation is presented as a transparent ledger rather than as a black-box ranking. During the Winterthur deployment, the outcome table was projected on screen, participants could browse the details, and the receipts explained how each project had been funded (Yang et al., 2 Oct 2025).

5. Winterthur deployment in 2025

The principal deployment took place in the 2025 Kultur Komitee in Winterthur, a cultural grant-making process run by the SKKG foundation. Residents were selected by lot to allocate roughly CHF 400,000 to cultural projects, and 38 residents ultimately served on the committee (Yang et al., 2 Oct 2025).

The process was intentionally hybrid. Before the workshop, participants completed online annotation homework: they reviewed roughly 30 project proposals, assigned each a four-point rating, and tagged projects with impact fields. At the workshop, they participated in two rounds of small-group deliberation, each organized around one field. In the afternoon, they cast individual votes by distributing 20 points across at least three projects they personally preferred. The resulting data were entered into a .pb file and processed through the Komitee Equal Shares tool in real time (Yang et al., 2 Oct 2025).

The committee jointly defined eight impact fields and assigned them weights. The final fields were Curiosity/Learning (19%), Community/Connection (18%), Public Space/Accessibility (15%), Emotional Impact (13%), Support for Newcomers (11%), Audience Creativity (10%), Nature/Environment (9%), and Tradition/History (6%). The committee also decided on the balance between individual voting and collective field-based evaluation, settling on a 50:50 split based on the distribution of participant preferences around parity (Yang et al., 2 Oct 2025).

The final KES portfolio funded 43 projects for CHF 378,901. The paper compares this with 35 projects in an individual-only greedy scenario and 33 projects in a group-only greedy scenario. It reports that the average project budget in the KES outcome was lower, which the authors interpret as evidence of support for a broader and more diverse set of projects, including smaller initiatives that might otherwise have been crowded out. One example was Velo volant, which received enough support only when individual and value-based inputs were combined (Yang et al., 2 Oct 2025).

Participant responses were mixed in ways that are analytically revealing. Many converged on the 50:50 split as fair, and survey responses indicated high satisfaction with participation opportunities. At the same time, some participants reported that the algorithm shaped their strategy, for example by discouraging support for projects already well backed in group deliberation, while others found the computation difficult to understand and wanted clearer explanations. The paper also reports that roughly half of participants described some form of learning or perspective change, including a broadened understanding of cultural work in Winterthur and of the diversity of the city’s cultural scene (Yang et al., 2 Oct 2025).

6. Position within Equal Shares research

Komitee Equal Shares belongs to a broader family of proportional allocation methods centered on the Method of Equal Shares. In approval-based committee voting, MES is a prominent example of a broader class of affordable committees and is the canonical priceable rule. In that literature, priceability and affordability are used to analyze representation, coverage, and utilitarian welfare after committee completion (Brill et al., 2023). In more recent axiomatic work, Equal Shares is also shown to satisfy FPJR by virtue of always outputting priceable committees (Kalayci et al., 21 Jan 2025). Komitee Equal Shares inherits the virtual-budget and priceability logic of this tradition but repurposes it for a setting in which citizens act both as voters and as evaluators (Yang et al., 2 Oct 2025).

The framework also sits within an expanding body of real-world MES applications. In Amsterdam participatory budgeting, MES was empirically compared with the greedy cost welfare rule on 35 instances and was reported to perform better on the studied notions of fairness, equality, and often representation, at the cost of a small reduction in total voter satisfaction (Nelissen, 2023). Kultur Komitee 2024 had already used a human-governed MES workflow to let participants choose how much budget to delegate to the algorithm before deliberative refinement (Yang et al., 7 Feb 2025). Komitee Equal Shares extends this practical line by making both individual preference and collective value judgment first-class inputs to one priceable allocation (Yang et al., 2 Oct 2025).

Several common misunderstandings are clarified by the 2025 paper. Komitee Equal Shares is not merely a standard MES run with post hoc explanation; the paper’s main design move is the explicit integration of voter mode and evaluator mode. It is also not a purely axiomatic extension of classical MES, because weighted impact fields and co-designed criteria introduce elements that do not exist in the voter-only model. Finally, it is not intended only for Winterthur. The paper explicitly presents it as generalizable to participatory grant-making and participatory budgeting more broadly, with a general recipe: co-design a small set of impact fields, let participants weigh them, gather both individual and group-based signals, run a priceable equal-shares-like allocation, and publish receipts and outcome tables (Yang et al., 2 Oct 2025).

In this sense, Komitee Equal Shares marks a shift in how Equal Shares is used. In the classical literature, Equal Shares is primarily a proportional rule for aggregating preferences under budget or committee constraints. In Komitee Equal Shares, the same priceable virtual-budget idea becomes a framework for combining preference aggregation, collective evaluation, and explanation. The proportional core remains, but it is embedded in a participatory design that treats fairness as both an allocation property and a communicative object (Yang et al., 2 Oct 2025).