Rhetoric of Data Center Resistance: Causal Stories, GSS Archetypes & Municipal Moratoriums
Combining Deborah Stone's Theory of Causal Stories (Accidental, Mechanical, Inadvertent, Intentional), the Narrative Policy Framework (NPF), empirical citizen archetypes from the General Social Survey (GSS 2022, N=3,784), and Charles Ragin's Qualitative Comparative Analysis (QCA) to discover minimal equifinal persuasion recipes across 428 US municipal data center moratoriums and utility disputes.
data_center_moratoriums.csv)K=4 Sociodemographic Citizen Archetypes1. Core Empirical Findings: Minimal Equifinal Persuasion Recipes
ΔY ≥ +1.50) across four GSS 2022 public archetypesWorking-Class Community Defenders
Exhibits textbook equifinality (100.0% consistency, avg ΔY = +1.91). Whether the causal story is Inadvertent, Intentional, or Mechanical, persuasion succeeds whenever anchored by Fairness & Anti-Cheating (refusing residential electric rate subsidies for Big Tech) + Lived Neighbor Testimonials.
Pro-Growth Property Conservatives
Climate and regulatory frames trigger negative backlash (ΔY < 0). Persuasion succeeds exclusively (85.7% consistency, avg ΔY = +1.63) when hyperscale transmission corridors and zoning overrides are framed through Liberty & Private Property Rights + Lived Farmer Testimonials.
Educated Progressive Conservationists
An Intentional Causal Story (challenging NDAs and corporate externalization) is a strict necessary condition. Succeeds when paired with Fairness + Lived Testimonial (100% consistency) or Harm Prevention + Legal Precedent (CEQA) (80% consistency).
Conservative Conservationist & Ratepayer Skeptic
Right-leaning rural/exurban voters (polviews = 5.05) with the lowest institutional trust (2.54) and highest demand for local watershed protection (1.00). Exposure to Accidental + Economic Pragmatism ("inevitable AI progress") triggers severe backlash (ΔY = -0.65).
2. Four-Stage Computational Social Science Architecture
Empirical Sociodemographic Personas
Ingests NORC GSS 2022 microdata (N=3,784 complete respondents across polviews, partyid, degree, coninc, xnorcsiz, conbus, confed, consci, natenvir) and extracts 4 empirical citizen archetypes via standardized K-Means clustering.
1,600 Causal Story Configurations
Prevents unconstrained LLM token drift by defining a structured discrete action space combining Deborah Stone's 4 Causal Stories with NPF Heroes, Villains, Victims, Moral Frames, Evidence Modes, and Tones.
500 Calibrated Deliberation Episodes
Evaluates pre-intervention moratorium support (Y_pre on a 1–7 scale) and post-hearing shift (ΔY = Y_post - Y_pre) across 500 Monte Carlo episodes (125 per archetype), flagging persuasion success at ΔY ≥ +1.50.
Quine-McCluskey Truth Tables & Ground Truth
Applies Charles Ragin's QCA truth-table synthesis (consistency ≥ 0.75) to extract minimal sufficient Boolean formulas and cross-validates them against 428 real-world municipal moratoriums.
3. Datasets, Research Report & Python Pipeline Files
projects/rhetoric_of_data_center_resistance/Comprehensive national tracker of 428 US municipal and county data center moratoriums, zoning bans, utility disputes, durations, causal rationales, and primary news/legal source URLs.
Full log of all 500 simulated public hearing deliberation episodes, including GSS respondent demographics, NPF action vectors, synthesized testimony, Y_pre, Y_post, ΔY, and cognitive reasoning.
Serialized Quine-McCluskey minimal Boolean formulas, consistency/frequency scores, GSS 2022 cluster centroids, and sampled respondent persona records (personas_sample.jsonl).
Complete academic writeup detailing the Deborah Stone causal story formalization, GSS 2022 clustering table, Boolean truth-table equations, and ground-truth alignment with local moratoriums.
Defines the 1,600 discrete NPF × Deborah Stone causal story configurations and synthesizes parameterized public-comment hearing testimonies.
End-to-end pipeline scripts for clustering GSS 2022 microdata, running the 500 deliberation episodes in persona_environment.py, and executing Quine-McCluskey minimization.