Pith. sign in

REVIEW 4 major objections 4 minor 114 references

LLM resident agents reproduce real redevelopment negotiations closely enough to test housing policy before implementation.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-01 02:23 UTC pith:UEJFMTSS

load-bearing objection Useful platform paper whose qualitative validation is genuinely informative, but the abstract oversells quantitative fidelity. the 4 major comments →

arxiv 2607.25447 v1 pith:UEJFMTSS submitted 2026-07-28 cs.MA

CoRenew: A large language model agent-based policy simulation platform for multifamily residential redevelopment

Yudi Zhang , Yuming Lin , Li Tian , Yu Wang , Jianghao Yu This is my paper
classification cs.MA
keywords multifamily residential redevelopmentlarge language model agentsgenerative agent-based modelingnegotiation simulationpolicy evaluationtheory of planned behaviorurban renewalsemantic policy input
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

CoRenew is an open-source platform that replaces rule-based assumptions about resident behavior with large language model agents, letting each synthetic resident reason through the Theory of Planned Behavior while bargaining over redevelopment terms. The paper's central claim is that these agents reproduce real-world negotiation dynamics and survey-measured decision pathways well enough to support ex ante comparison of redevelopment policies. The authors validate this against 324 resident survey responses and a nine-month observed negotiation: the simulated structural model matched the survey-derived model in 100% of path directions and 94.7% of statistical-significance decisions, and the best LLM traced the observed negotiation trajectory with a DTW distance of 2.00. If true, this means policymakers can explore how subsidies, extension rules, and semantic institutional arrangements like escrow accounts reshape collective consent before committing public money. The platform also surfaces concrete policy insights—subsidy effects are nonlinear and partly captured by developers, and some equity risks accompany consensus-promoting policies—suggesting it can generate testable hypotheses for urban renewal.

Core claim

The central discovery is that LLM-based resident agents, structured by the Theory of Planned Behavior within a Markov game, can approximate both the process and the causal structure of real multifamily redevelopment negotiations. Across 19 comparable structural paths, the LLM-based model recovered all benchmark paths, agreed with the survey model on 100% of directional signs and 94.7% of statistical significance decisions, and reported a mean coefficient ratio of 1.11 to the survey estimates. The best-performing LLM also reproduced the observed negotiation trajectory—slow early convergence, delayed middle acceleration, late stabilization—while other LLMs converged too quickly. Applying CoRen

What carries the argument

The central mechanism is the LLM-based resident agent embedded in a finite-round Markov game, whose reasoning is prompted along the three Theory of Planned Behavior constructs—attitude, perceived behavioral control, and subjective norm—through a chain-of-thought scratchpad. A weighted representative-selection module (distance-decay probabilities, K-means clustering, and a low-income seat safeguard) compresses the synthetic population to a few voting representatives while computing utilities for every resident, enabling large-scale simulation. This combination lets the model treat policies as both numerical parameters and natural-language institutional arrangements, and it is what carries the

Load-bearing premise

Each selected representative's agreement decision is propagated to all residents in its cluster without error, so within-cluster opinion differences are assumed not to matter.

What would settle it

Re-run a full negotiation simulation for one mixed-income community using every resident as an agent rather than representatives; if the community agreement rate or policy ranking shifts materially, the representative-propagation shortcut biases the platform's conclusions.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • If the validation holds, CoRenew can serve as a low-cost ex ante testing ground for redevelopment policy, replacing or supplementing expensive real-world negotiation pilots.
  • The platform's noted nonlinear subsidy effects imply that modest subsidy increases may fail to reduce inequality, while higher subsidies yield diminishing utility gains for low-income residents—directly informing budget allocation.
  • Because part of the fiscal subsidy is absorbed into developer profit, the results imply that subsidy design must account for strategic bargaining between planners and developers.
  • The ability to assess semantic policies (escrow accounts, shared ownership, supervision mechanisms) extends policy evaluation beyond purely financial instruments.
  • The authors claim the architecture can be reconfigured to represent different institutional contexts, such as Singapore's en bloc majority-consent system or Finnish extension-based financing.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • A natural extension is to treat the representative-selection shortcut as a design variable and test whether community-level policy rankings change when cluster sizes shrink or within-cluster opinion diversity is explicitly modeled.
  • The paper's validation is single-case for negotiation dynamics; a direct test would be to run CoRenew on a second real case from a different institutional context and check whether the same LLM still reproduces the observed trajectory.
  • Because the authors find that heterogeneity facilitates agreement in high-appreciation settings but hinders it elsewhere, one could derive community-level policy portfolios and then verify these predictions via a randomized field experiment in similar neighborhoods.
  • The discrepancy in the Policy-to-perceived-behavioral-control path suggests that LLM agents under-represent how governance clarity empowers residents; fine-tuning prompts on institutional trust could close that gap.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper presents CoRenew, an open-source platform that simulates multi-round negotiations among residents, developers, and planners in multifamily residential redevelopment using LLM-based resident agents structured by the Theory of Planned Behavior. The platform generates synthetic populations from public data, selects weighted representatives, supports numerical and semantic policy inputs, and evaluates policies across multiple objectives. Validation is attempted at two levels: (i) comparison of simulated negotiation trajectories with a real nine-month case, reporting DTW distance, success rates, and rounds to consensus; and (ii) comparison of SEM path coefficients estimated from LLM agents' chain-of-thought with coefficients estimated from a 324-person survey. The paper reports 100% agreement in directional signs and 94.7% agreement in statistical significance for 19 paths, and DTW=2.00 for the best LLM (DeepSeek-V3.2). The platform is then applied to Huadu District, Guangzhou, to identify global optimal policies, community-specific designs, and effects of semantic policies.

Significance. If the validation were quantitative, CoRenew would be a useful transferable GABM tool for ex ante housing-policy evaluation, addressing two gaps: adaptive negotiation dynamics and semantic policy inputs. The platform's strengths include open-source availability, use of public data for synthetic populations, a clearly specified modular architecture, and validation against both independent survey data and a real observed negotiation. The manuscript also honestly disclaims 'precise predictor' status in the conclusion. However, the load-bearing quantitative claims are only partially supported: outcome-level validation (success rate, timing) is weak, and half of the SEM path coefficients differ significantly from survey estimates. These weaknesses directly affect the reliability of the policy rankings and subsidy-threshold conclusions that are the paper's main applied payoff.

major comments (4)
  1. [§4.1.1] The process-level DTW match (2.00) is presented as the main validation, but outcome-level quantities diverge from the observed case: simulated success rate is 84.6% vs 96.0%, and full agreement occurs in rounds 2.25–4.45 vs the observed 6.0. These are precisely the quantities used for policy comparison (agreement rates, subsidy expenditure, rounds to consensus). The claim 'closely matched the observed negotiation trajectory' is therefore only about shape, not levels. The paper should either calibrate the model to reproduce observed outcome levels or explicitly restrict the policy conclusions to directional/qualitative insights.
  2. [§4.1.2, Table 4] The abstract's '100% signs / 94.7% significance' hides substantial magnitude errors: 9 of 19 path coefficients differ significantly from the survey benchmark (e.g., Economic→PBC z=6.35; Policy→PBC z=−3.19). The single significance mismatch, Policy→PBC, is a policy variable central to the platform's purpose. Because the global and community-specific policy optimizations in §4.2 rely on relative effect sizes, the model is not yet validated for quantitative policy ranking. The authors should add sensitivity/calibration analysis showing which policy conclusions are robust to these coefficient differences, and temper the validation claim. The shared use of DeepSeek-V3.2 for both agent generation and chain-of-thought coding should also be explicitly discussed as a potential source of dependence.
  3. [§3.2, Eq. (4)] The representative-selection mechanism propagates each representative's binary agreement decision to all residents in its cluster without error. This is a strong assumption, especially for high-heterogeneity communities, which are central in the application. The weighting by cluster size can systematically bias agreement rates and, consequently, policy evaluations. The low-income seat safeguard does not address within-cluster opinion diversity. The authors should report sensitivity of key results to K and, ideally, validate the propagation assumption by comparing representative-based agreement rates with full-population decisions in at least one community.
  4. [§4.2.2 and §4.2.3] The headline results—the global optimum (30% extension cap, 12% subsidy cap), the community-specific optimal packages, and the nonlinear subsidy effects—are reported as point values without uncertainty quantification. Given LLM stochasticity, finite seeds (five in Phase 2), and the unvalidated outcome levels discussed above, the policy ranking could change across runs. The paper should provide variance estimates, pairwise policy-comparison tests, or at least a sensitivity analysis over seeds and representative-selection realizations before these results are used to support policy recommendations.
minor comments (4)
  1. [General] Many appendix cross-references appear as unresolved 'Appendix??' placeholders (e.g., §3.1, §3.2, §3.3.2, §3.7, §4.2.1). The appendices are needed to assess prompt templates, coding templates, data sources, and parameter settings. Please include them.
  2. [§3.6] Figure 3 is referenced as '3a'–'3d' but the caption is unhelpful ('Visualization interface'). Add a descriptive caption identifying each panel.
  3. [§4.1.1] It is unclear what 'success rate' means after the approval threshold—the observed 96.0% is defined for one real negotiation, so a precise definition of 'simulated success rate' and the number of independent runs should be stated in one place.
  4. [§3.3.5] The negotiation termination conditions are stated informally ('either the required approval threshold is reached or the maximum number of rounds is exhausted'). Please specify the default threshold and round counts explicitly and include them in the reproducibility package.

Circularity Check

1 steps flagged

No formal circularity: the platform is not trained or fitted on the validation targets, and the policy-optimization output is not derived from the survey/negotiation benchmarks. The only qualifying concern is that the TPB-path validation uses the same theoretical framing—and the same LLM family—for both generating and coding the agent reasoning that the structural paths are estimated from.

specific steps
  1. other [Section 3.3.2 (Resident agent), Section 3.7 (Model validation), Section 4.1.2 (Validation of TPB-based structural path estimates)]
    "resident prompting is structured using the Theory of Planned Behavior. Decision-relevant information is organized around three dimensions: attitude, perceived behavioral control, and subjective norm. ... To estimate the corresponding paths for the LLM-based agents, we coded residents’ chain-of-thought reasoning using a predefined coding template implemented with DeepSeek-V3.2. ... Across 19 comparable paths, the LLM-based SEM recovered all benchmark paths, achieved 100% accuracy in directional signs, and reached 94.7% accuracy in statistical significance."

    The LLM-side structural paths are not an independent behavioral output: the agent's chain-of-thought is already prompted to organize reasoning around the exact TPB constructs (attitude, PBC, subjective norm) that then serve as the latent outcomes in the SEM comparison, and the same LLM family used to generate the scratchpad is also used to code the scratchpad into TPB categories. Thus 'recovery' of TPB-mediated pathways is partly underwritten by the prompting/coding design rather than by a free-standing behavioral mechanism. The external-factor-to-construct signs and significances were not specified in the prompt, however, so the validation still carries meaningful external information; this is a confound rather than a full equation-level circular reduction.

full rationale

The central claim that CoRenew can compare policies ex ante is not circular in the formal sense. The survey-based SEM and the observed nine-month negotiation trajectory are external benchmarks; no parameter is fit to them, and no predicted quantity in the Huadu policy application is defined as a function of those benchmark values. The representative-selection weighting (Eq. 4) is an aggregation rule, not a fitted input masquerading as a prediction. Self-citations (e.g., Wang et al. 2026; Zhang et al. 2026) are used as background or as post-hoc explanations of simulation discrepancies, not as the load-bearing evidence for the validation. The one legitimate concern is the methodological overlap between the TPB-structured prompts used to generate resident reasoning, the TPB-based coding template used to convert that reasoning into path estimates, and the same LLM performing both roles; this can inflate apparent agreement with the survey SEM without making the claim equivalent by construction. The paper's own caveat that CoRenew is 'an exploratory model for policy comparison rather than a precise predictor' further limits the strength of the validation claim. Overall, no equation in the paper reduces a predicted result to its own input, so the circularity score is low.

Axiom & Free-Parameter Ledger

4 free parameters · 6 axioms · 0 invented entities

The central claim rests on the behavioral validity of LLM agents and the fidelity of the representative-selection pipeline. No new physical entities are introduced. The main free parameters are user-controlled settings rather than fitted constants; however, several ad hoc thresholds and user choices (K, intervention triggers, policy grid) shape results.

free parameters (4)
  • Number of representatives per community (K)
    User-defined in representative selection (Section 3.2); controls simulation cost and fidelity. No fitted value reported.
  • Planner intervention thresholds
    Rules like 'negotiation reaches the midpoint' and 'offers remain unchanged for two consecutive rounds' (Section 3.3.3) are chosen by hand and affect outcomes.
  • Policy search grid = extension caps 20–50%, subsidy caps 0–20%
    The search space in Section 4.2.2 is a case-study choice, not derived; different grids may change the 'optimal' policy.
  • Low-income definition = below area median income
    Used in representative-selection safeguard (Section 3.2); this binary cutoff is an ad hoc operationalization.
axioms (6)
  • domain assumption Theory of Planned Behavior (TPB) is a valid representation of resident decision-making
    Section 3.3.2 structures resident-agent prompting around attitude, perceived behavioral control, and subjective norm; if TPB omits a dominant driver, the simulation will be systematically misaligned.
  • domain assumption Residents are more likely to trust and be represented by people with similar socioeconomic profiles
    Section 3.2, citing Mansbridge 2009; this justifies the representative-selection probability kernel in Eq. (1).
  • domain assumption A Markov game with fixed turn order and rule-based planner/developer captures essential negotiation dynamics
    Section 3.3.1 models the process as a sequential game; planner and developer are rule-based, which limits adaptivity and may miss real-world strategic behavior.
  • domain assumption Synthetic residents generated from census aggregates behave like real residents
    Section 3.2 reconstructs households from public data; if the imputation misses key attributes, agent decisions may diverge from reality.
  • domain assumption LLM agents produce stable, human-like behavior across rounds and runs
    Section 3.3.2 relies on chain-of-thought prompting; the validation in Section 4.1.1 shows variance across models but not a stability analysis across seeds for the SEM comparison.
  • domain assumption The survey-based SEM is a valid ground-truth benchmark for behavioral realism
    Section 3.7 uses 324 residents' survey responses as the benchmark; if the survey instrument or SEM specification is biased, the 'recovery' of paths does not establish realism.

pith-pipeline@v1.3.0-alltime-deepseek · 15661 in / 14140 out tokens · 148446 ms · 2026-08-01T02:23:30.355970+00:00 · methodology

0 comments
read the original abstract

The difficulty of collective action remains a central challenge in the design of policies for multifamily residential redevelopment. Stakeholders continually adjust their decisions in response to evolving negotiation contexts and the reactions of others, meaning that when a policy intervenes and which stakeholders it targets can substantially reshape collective outcomes. Assessing these adaptive responses ex ante remains difficult because existing simulation models often rely on predefined behavioral rules. Here, we present CoRenew, an open-source platform that uses LLM-based agents to simulate negotiations among multiple stakeholders and evaluate the effects of alternative policy combinations. Integrating open source geographic and demographic data, the platform can generate synthetic residents, simulate negotiation dynamics under alternative policy settings and compares policy performance across competing objectives. It supports both numerical and semantic policy inputs and includes built-in tools for visualization and result export. We validate its behavioral realism against survey responses from 324 residents and a nine-month observed negotiation process from a real redevelopment case. With its modular and adaptable architecture, CoRenew can be used to assess policies across different institutional and cultural contexts.

Figures

Figures reproduced from arXiv: 2607.25447 by Jianghao Yu, Li Tian, Yudi Zhang, Yuming Lin, Yu Wang.

Figure 1
Figure 1. Figure 1: Two-level adaptive framework linking global policy design with local negotiation [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overall framework 12 [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Visualization interface of CoRenew. DeepSeek-V3.2, GLM, GPT-4.1, GPT-4o, and Gemini, and repeated each simulation 20 times to account for stochastic variation. Performance was evaluated at three levels: final outcome, negotiation process, and individual behavior. Second, we tested whether the simulated agents could reproduce the ef￾fects of economic, geographic, policy, and social factors on residents’ de￾… view at source ↗
Figure 4
Figure 4. Figure 4: Structural model for comparing decision pathways between LLM agents and [PITH_FULL_IMAGE:figures/full_fig_p021_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of LLM-simulated and observed consensus-building outcomes in the [PITH_FULL_IMAGE:figures/full_fig_p023_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Pareto frontiers of policy combinations under pairwise policy trade-offs. [PITH_FULL_IMAGE:figures/full_fig_p027_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Optimal policy combinations for different community types and their objective [PITH_FULL_IMAGE:figures/full_fig_p028_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Effects of natural-language policy interventions on agreement outcomes across [PITH_FULL_IMAGE:figures/full_fig_p030_8.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

114 extracted references · 1 linked inside Pith

  1. [1]

    Housing and Society , year =

    Badawy, Aya and Kingston, Richard and Pinto, Nuno , title =. Housing and Society , year =

  2. [2]

    Sustainability , year =

    Chen, Zhibiao and Bian, Ana and Wu, Zhongping , title =. Sustainability , year =

  3. [3]

    Proceedings of the ACM Web Conference 2026 , pages=

    PolicySim: An LLM-Based Agent Social Simulation Sandbox for Proactive Policy Optimization , author=. Proceedings of the ACM Web Conference 2026 , pages=

    2026

  4. [4]

    Complexity , volume=

    Policy Modeling and Applications: State-of-the-Art and Perspectives , author=. Complexity , volume=. 2019 , publisher=

    2019

  5. [5]

    Journal of Urban Planning and Development , volume=

    Effect of private externalities in urban housing renewal investment: Empirical assessment using a game-theory approach , author=. Journal of Urban Planning and Development , volume=. 2017 , publisher=

    2017

  6. [6]

    Political Research Exchange , volume=

    Complex system policy modelling approaches for policy advice--comparing systems thinking, system dynamics and agent-based modelling , author=. Political Research Exchange , volume=. 2024 , publisher=

    2024

  7. [7]

    Nature Computational Science , volume=

    Generative language models exhibit social identity biases , author=. Nature Computational Science , volume=. 2025 , publisher=

    2025

  8. [8]

    International Journal of Law in The Built Environment , volume =

    Christudason, Alice , title =. International Journal of Law in The Built Environment , volume =. 2009 , month =

    2009

  9. [9]

    and Randolph, Bill , title =

    Crommelin, Laura and Easthope, Hazel and Troy, L. and Randolph, Bill , title =. Australian Planner , volume =. 2020 , month =

    2020

  10. [10]

    Journal of Property, Planning and Environmental Law , year =

    Easthope, Hazel and Crommelin, Laura and Gillon, Charles and Pinnegar, Simon and Ruming, Kristian and Liu, Sha , title =. Journal of Property, Planning and Environmental Law , year =

  11. [11]

    and Ho, D

    Gao, L. and Ho, D. , title =. Habitat International , volume =. 2016 , month =

    2016

  12. [12]

    , title =

    Gao, Jinjian and Xu, Hanyong and Dao, L. , title =. 2024 , month =

    2024

  13. [13]

    Hauge, Å. L. and Thomsen, Judith and Löfström, Erica , title =. 2013 , month =

    2013

  14. [14]

    and Gao, Wei , title =

    Ho, D. and Gao, Wei , title =. Habitat International , volume =. 2013 , month =

    2013

  15. [15]

    , title =

    Kojima, T. , title =. 2013 , month =

    2013

  16. [16]

    , title =

    Liang, Xin and Yu, Tao and Hong, Jingke and Shen, G. , title =. 2019 , month =

    2019

  17. [17]

    Liu, Guiwen and Wei, Lizhen and Gu, Jianping and Zhou, Tao and Liu, Yong , title =

  18. [18]

    , title =

    Meno, F. , title =. Journal of Architecture and Planning (transactions of Aij) , volume =

  19. [19]

    , title =

    Park, Sang-hoon and Oh, E. , title =. Renewable & Sustainable Energy Reviews , volume =. 2011 , month =

    2011

  20. [20]

    and Tiilikainen, A

    Pennanen, K. and Tiilikainen, A. and Viitanen, Kauko , title =

  21. [21]

    and Falkenbach, Heidi and Arvola, A

    Puustinen, Tuulia and Pennanen, K. and Falkenbach, Heidi and Arvola, A. and Viitanen, Kauko , title =. Property Management , volume =. 2017 , month =

    2017

  22. [22]

    and Falkenbach, Heidi and Viitanen, Kauko , title =

    Puustinen, Tuulia and Pennanen, K. and Falkenbach, Heidi and Viitanen, Kauko , title =. 2018 , month =

    2018

  23. [23]

    and Baležentis, T

    Štreimikienė, D. and Baležentis, T. , title =. Energies , year =

  24. [24]

    Asian Journal of Comparative Law , year =

    Ti, Edward S W , title =. Asian Journal of Comparative Law , year =

  25. [25]

    and Easthope, Hazel and Randolph, Bill and Pinnegar, Simon , title =

    Troy, L. and Easthope, Hazel and Randolph, Bill and Pinnegar, Simon , title =. 2017 , month =

    2017

  26. [26]

    , title =

    Walters, M. , title =. International Journal of Social Economics , volume =. 2002 , month =

    2002

  27. [27]

    Sustainability , year =

    Wang, Li and Ren, Pan and Shan, Yongwei and Zhang, Guanqiao , title =. Sustainability , year =

  28. [28]

    , title =

    Winther, Tanja and Gurigard, K. , title =. 2016 , month =

    2016

  29. [29]

    International Real Estate Review , volume =

    Yamazaki, Fukuju and Sadayuki, Taisuke , title =. International Real Estate Review , volume =

  30. [30]

    , title =

    Yau, Y. , title =. Habitat International , volume =. 2014 , month =

    2014

  31. [31]

    Land , year =

    Yuan, Dinghuan and Li, Jiaxin and Li, Qiuxiang and Fu, Yang , title =. Land , year =

  32. [32]

    and Kim, Kabsung

    Yun, S. and Kim, Kabsung . , title =. Journal of Urban Regeneration and Renewal , year =

  33. [33]

    Korean Association for Housing Policy Studies , year =

    Yun, Young-ok and Lim, Mihwa , title =. Korean Association for Housing Policy Studies , year =

  34. [34]

    Zhou, Youmei and Chen, Jingjing and Zhang, Xiyu and Liu, Xiao and Lei, Hao , title =. 2022 IEEE Intl Conf on Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive Intelligence and Computing, Intl Conf on Cloud and Big Data Computing, Intl Conf on Cyber Science and Technology Congress (DASC/PiCom/CBDCom/CyberSciTech) , pages =. 2022 , month =

    2022

  35. [35]

    2023 , month =

    Zhou, Youmei and Lei, Hao and Zhang, Xiyu and Wang, Sha and Xu, Yingying and Li, Chao and Zhang, Jie , title =. 2023 , month =

    2023

  36. [36]

    Advances in Neural Information Processing Systems , volume=

    Cooperation, competition, and maliciousness: Llm-stakeholders interactive negotiation , author=. Advances in Neural Information Processing Systems , volume=

  37. [37]

    Organizational behavior and human decision processes , volume=

    The theory of planned behavior , author=. Organizational behavior and human decision processes , volume=. 1991 , publisher=

    1991

  38. [38]

    Nature Human Behaviour , volume=

    Playing repeated games with large language models , author=. Nature Human Behaviour , volume=. 2025 , publisher=

    2025

  39. [39]

    European journal of housing policy , volume=

    Large housing estates in Slovenia: a framework for renewal , author=. European journal of housing policy , volume=. 2001 , publisher=

    2001

  40. [40]

    Housing studies , volume=

    Why are housing cooperatives successful? Insights from Swiss affordable housing policy , author=. Housing studies , volume=. 2018 , publisher=

    2018

  41. [41]

    Journal of Public Economics , volume=

    A credible approach for measuring inframarginal participation in energy efficiency programs , author=. Journal of Public Economics , volume=. 2014 , publisher=

    2014

  42. [42]

    Journal of Comparative Economics , volume=

    Collective management of residential housing in Russia: The importance of being social , author=. Journal of Comparative Economics , volume=. 2014 , publisher=

    2014

  43. [43]

    Applied Energy , volume=

    Does social interaction have an impact on residents’ sustainable lifestyle decisions? A multi-agent stimulation based on regret and game theory , author=. Applied Energy , volume=. 2019 , publisher=

    2019

  44. [44]

    Journal of Housing Economics , volume=

    Redevelopment values in multi-family properties: Evidence from en bloc sales in Singapore , author=. Journal of Housing Economics , volume=. 2023 , publisher=

    2023

  45. [45]

    Urban Studies , volume=

    Collective action dilemmas in condominium management , author=. Urban Studies , volume=. 2013 , publisher=

    2013

  46. [46]

    Urban studies , volume=

    Decisions to renovate: Identifying key determinants in Central and Eastern European post-socialist countries , author=. Urban studies , volume=. 2013 , publisher=

    2013

  47. [47]

    2025 , type =

    Jiang, Qingyue , title =. 2025 , type =

    2025

  48. [48]

    Journal of Affordable Housing & Community Development Law , pages=

    More than money: What is shared in shared equity homeownership? , author=. Journal of Affordable Housing & Community Development Law , pages=. 2010 , publisher=

    2010

  49. [49]

    Energy and Buildings , volume=

    Factors impacting energy-efficient retrofits in the residential sector: The effectiveness of the Slovenian subsidy program , author=. Energy and Buildings , volume=. 2020 , publisher=

    2020

  50. [50]

    Housing Studies , volume=

    Governing the compact city: the challenges of apartment living in Sydney, Australia , author=. Housing Studies , volume=. 2009 , publisher=

    2009

  51. [51]

    Environment and Planning A , volume=

    Urban renewal and strata scheme termination: Balancing communal management and individual property rights , author=. Environment and Planning A , volume=. 2013 , publisher=

    2013

  52. [52]

    International Journal of Housing Policy , pages=

    Effectiveness and equity of housing renovation subsidies: universal versus targeted approaches in Barcelona , author=. International Journal of Housing Policy , pages=. 2025 , publisher=

    2025

  53. [53]

    Nature Cities , volume=

    Large language models in urban planning , author=. Nature Cities , volume=. 2025 , publisher=

    2025

  54. [54]

    Habitat International , volume=

    Does owner heterogeneity matter in the management of multi-owned housing? , author=. Habitat International , volume=. 2016 , publisher=

    2016

  55. [55]

    Humanities and Social Sciences Communications , volume=

    Large language models empowered agent-based modeling and simulation: A survey and perspectives , author=. Humanities and Social Sciences Communications , volume=. 2024 , publisher=

    2024

  56. [56]

    System Dynamics Review , volume=

    Generative agent-based modeling: an introduction and tutorial , author=. System Dynamics Review , volume=. 2024 , publisher=

    2024

  57. [57]

    American Sociological Review , pages=

    Collective action and group heterogeneity: voluntary provision versus selective incentives , author=. American Sociological Review , pages=. 1993 , publisher=

    1993

  58. [58]

    IEEE Transactions on Fuzzy Systems , volume=

    An automatic consensus reaching approach with preference adjustment willingness for group decision-making , author=. IEEE Transactions on Fuzzy Systems , volume=. 2023 , publisher=

    2023

  59. [59]

    Environment and Planning B: Planning and Design , volume=

    A strategic analysis of urban renewal in Taipei City using game theory , author=. Environment and Planning B: Planning and Design , volume=. 2014 , publisher=

    2014

  60. [60]

    Planning Theory & Practice , volume=

    Negotiation processes in urban redevelopment projects: Dealing with conflicts by balancing integrative and distributive approaches , author=. Planning Theory & Practice , volume=. 2015 , publisher=

    2015

  61. [61]

    Habitat International , volume=

    Residents’ behavioral intention to participate in neighborhood micro-renewal based on an extended theory of planned behavior: A case study in Shanghai, China , author=. Habitat International , volume=. 2022 , publisher=

    2022

  62. [62]

    Buildings , volume=

    Financial Modelling of Transition to Escrow Schemes in Urban Residential Construction: A Case Study of Tashkent City , author=. Buildings , volume=. 2025 , publisher=

    2025

  63. [63]

    1997 , publisher=

    Evolutionary game theory , author=. 1997 , publisher=

    1997

  64. [64]

    Urban Policy and Research , volume=

    Understanding property rights in the contemporary urban commons , author=. Urban Policy and Research , volume=. 2014 , publisher=

    2014

  65. [65]

    Energy and Buildings , volume=

    Application of cumulative prospect theory in understanding energy retrofit decision: A study of homeowners in the Netherlands , author=. Energy and Buildings , volume=. 2022 , publisher=

    2022

  66. [66]

    Computers, Environment and Urban Systems , volume=

    Evolution of residents' cooperative behavior in neighborhood renewal: An agent-based computational approach , author=. Computers, Environment and Urban Systems , volume=. 2023 , publisher=

    2023

  67. [67]

    Cities , volume=

    An agent-based model of residents' group decision making in neighborhood regeneration: Assessing the impacts of consensus-reaching approaches , author=. Cities , volume=. 2025 , publisher=

    2025

  68. [68]

    Policy design and practice , volume=

    Policy failure and the policy-implementation gap: can policy support programs help? , author=. Policy design and practice , volume=. 2019 , publisher=

    2019

  69. [69]

    Scientific Data , volume=

    100-m resolution Age-Stratified Population Estimation from the 2020 China Census by Township (ASPECT) , author=. Scientific Data , volume=. 2025 , publisher=

    2020

  70. [70]

    Housing studies , volume=

    Development pathways of large housing estates in post-socialist cities: An international comparison , author=. Housing studies , volume=. 2012 , publisher=

    2012

  71. [71]

    Science Progress , volume=

    To join the rebuild or not? An exploration of the factors influencing the public's intention to participate in urban renewal , author=. Science Progress , volume=. 2022 , publisher=

    2022

  72. [72]

    Emerging Media , volume=

    Agentic AI for sustainable development: Leveraging large language model-enhanced agent-based modeling for complex policy strategies , author=. Emerging Media , volume=. 2025 , publisher=

    2025

  73. [73]

    International journal of law in the built environment , volume=

    Legal challenges in ensuring regular maintenance and repairs of owner-occupied apartment blocks , author=. International journal of law in the built environment , volume=. 2010 , publisher=

    2010

  74. [74]

    2024 , institution=

    Automated social science: Language models as scientist and subjects , author=. 2024 , institution=

    2024

  75. [75]

    selection model

    A “selection model” of political representation , author=. Journal of Political Philosophy , volume=. 2009 , publisher=

    2009

  76. [76]

    Journal of Cleaner Production , volume=

    Agent-based modelling to simulate farmers’ sustainable decisions: Farmers’ interaction and resulting green consciousness evolution , author=. Journal of Cleaner Production , volume=. 2022 , publisher=

    2022

  77. [77]

    arXiv preprint arXiv:2506.12110 , year=

    EconGym: A Scalable AI Testbed with Diverse Economic Tasks , author=. arXiv preprint arXiv:2506.12110 , year=

    Pith/arXiv arXiv

  78. [78]

    Urban studies , volume=

    Social capital and neighbourhoods that work , author=. Urban studies , volume=. 2005 , publisher=

    2005

  79. [79]

    Land Use Policy , volume=

    Organizational model of municipal housing stock management in the contracting system--A case study of Poland , author=. Land Use Policy , volume=. 2022 , publisher=

    2022

  80. [80]

    Annual Review of Psychology , volume=

    Embracing complexity: A review of negotiation research , author=. Annual Review of Psychology , volume=. 2023 , publisher=

    2023

Showing first 80 references.