PNAS 123/37 pp. e2600126123-e2600126123 2026-09-09 Original Plausible nonsense and deliberative reasoning: Benchmarking LLMs against human judgment
Francesco Veri, Gustavo Umbelino
Abstract
Large Language Models (LLMs) are entering democratic contexts as instruments of governance, where the challenges at hand are ill-structured, marked by ambiguity and contestation. Ill-structured democratic problems demand more than factual precision; they call for intersubjective reasoning: context-sensitive judgments that others can understand and publicly accept. Using the Deliberative Reason Index (DRI), this study evaluates 60 LLMs against human deliberation across nine policy scenarios. Only four models consistently exceed the permutation-based null benchmark for alignment with human patterns of reason-giving. Most models fall short: their reason-preference structures rarely clear this threshold, even though their outputs can still appear coherent and persuasive. Yet outputs can appear reasonable even when this alignment is absent. The observed gap between surface plausibility and deliberative coherence urges caution: deploying LLMs in governance contexts requires prior assessment of their deliberative reasoning capacity, not just their surface outputs.
JPART — 2026-09-09
Original Institutional Conversion and Digital Transformation in Local Governments: A Longitudinal Analysis of Trajectories
Adalmir de Oliveira Gomes
Abstract
Abstract This article asks a central question in digital government research: why do local governments operating under similar institutional conditions follow sharply different digital transformation trajectories over time? Drawing on theories of gradual institutional change, the study conceptualizes digital transformation not as a discrete adoption outcome, but as a process of institutional conversion through which technological capacity is translated into sustained digital service provision. Information governance is theorized as the key conditioning mechanism linking technological resources to durable organizational outcomes. The analysis combines longitudinal panel models, fixed-effects estimators, and trajectory-based clustering using a panel of all Brazilian municipalities (MUNIC/IBGE; N = 5,569) across seven waves between 2004 and 2024. Results show that increases in technological capacity are associated with subsequent gains in digital services, but these effects are gradual and conditional. Information governance significantly shapes how municipalities mobilize and stabilize technological resources over time, while interaction models reveal non-linear and diminishing returns consistent with institutional coordination limits. Trajectory analyses identify four distinct patterns of digital transformation, ranging from ‘institutionalized pioneers’ to ‘persistent stagnation’, demonstrating that digital development unfolds through path-dependent and institutionally anchored trajectories rather than through simple convergence. The article contributes to public administration research in three ways: provides rare longitudinal evidence on local digital transformation over two decades; integrates institutional change theory into digital government research by conceptualizing governance as a mechanism of institutional conversion; and advances a trajectory-based explanation of digital transformation that shifts attention from episodic technological adoption to cumulative institutional processes shaping long-run administrative performance.
PA — 2026-09-09
Original How Were Governance Pathways Formed for Public Service Co‐Production? A Contingency Perspective
PSJ 54/4 2026-09-09
Original Balancing Particularism and Universalism in Welfare Policy: Insights From China's Hukou Reform
Liu Chen, Tom Christensen, Xufeng Zhu
Abstract
ABSTRACT Welfare policies often combine particularistic and universalistic elements, yet we know less about how shifts in this balance shape public attitudes toward vulnerable low‐status groups. This article develops a three‐dimensional framework for analyzing welfare policy, distinguishing among welfare scope, welfare benefits, and inclusion criteria. We argue that public attitudes toward lower‐status groups are shaped by the combined signal about the welfare deservingness hierarchy generated by the configuration of particularistic and universalistic elements across these dimensions. Because policy design packages convey a broader normative stance about who belongs and who deserves, they can influence perceptions of groups that are not directly targeted by a policy change, lowering or raising their perceived deservingness. Using difference‐in‐differences and regression discontinuity methods, we analyze Weibo posts during China's two‐round hukou reform. We find that shifts toward more particularistic policy configurations lead to more negative attitudes toward low‐skilled migrant workers, whereas shifts toward more universalistic configurations yield more positive attitudes. Subgroup analyses yield suggestive patterns across local economic contexts and individual characteristics. These findings have broader implications for research on welfare policy systems and for reform efforts regarding social integration.
PNAS 123/37 pp. e2600311123-e2600311123 2026-09-09 Original A role for self-regulatory control in the absence of ingroup preference
Samuel A. W. Klein, Tessa Elizabeth Sadie Charlesworth, Daniel W. Heck, Mahzarin R. Banaji, Jeffrey W. Sherman
JCC — pp. 1-18 2026-09-08 Original Chinese Intellectual Influencers and the Construction of Algorithmized Nationalism: The Case of Jinri Toutiao
Haiyan Wang, Zhengqing Yan, Lin Wu, Carwyn Morris
JOP — 2026-09-08
Original Beliefs and Racial Identity Norms among Black Americans: An Examination of How Black Americans Understand Racism
Antoine Banks, Eric Groenendyk, Ja'nae Jackson, Erik Kimbrough, Mark Pickup
JPART — 2026-09-08
Original A Model of Agency Behavior in Participatory Budgeting: A Resource Dependence Perspective
Seunghyun Kim
Abstract
Abstract Based on resource dependence theory, this article theorizes why government agencies adopt citizen proposals in participatory budgeting. Agencies depend on the executive budget office and the legislature for budget approval, which makes them face uncertainty in securing budgets from these political principals. In this context, agencies can adopt citizen proposals to present their requests as reflecting citizen demands, thereby increasing the probability of approval. Building on this logic, this study argues that the greater the uncertainty agencies face in their dependencies on principals, the more likely they are to depend on proposals. This argument is tested using multilevel mixed-effects logistic regression based on citizen proposals (N = 3,609) submitted through Korea’s national participatory budgeting initiative from 2020 through 2023. The results show that agencies with less financial support from the legislature or those with smaller budgets are more likely to adopt proposals, as they face greater uncertainty due to unstable budgetary flows or fewer response options to manage resource instability. Agencies are also more likely to adopt proposals when they previously secured greater funding through proposal adoption, but this effect weakens for larger-budget agencies or those receiving greater legislative support. These findings emphasize the strategic nature of proposal adoption and the role of resource dependence in agency responses to proposals.
PMR — pp. 1-27 2026-09-08 Original A fish rots from the head down: unpacking the trickle-down effect of leaders’ unethical pro-organizational behaviour (UPB) in the Chinese public sector
PMR — pp. 1-37 2026-09-07 Original When tasks feel unjustified: effects of illegitimate tasks on public servants’ citizen-oriented boundary-spanning behaviours