ATT Internal Terrain Assessment v3.1 — United States, Q4 2026

This version 3.1 supersedes v2.1 after independent review on October 29. See the Review log, which lists every change. Executive assessment What is measured is clear: federal trust is at record lows and split by party, while local trust holds. What the read infers from it, that the federal foreign-policy node is decoupling, is a hypothesis it has not yet measured. The mode of the broader institutional floor is undetermined. [Extension] In the Codex's grammar, the central claim is a coherence claim: the terrain drivers for the civic membrane (the war, the energy market, the alliance system) are moving faster than the institutions inside it can recover. Neither coherence (C) nor its critical value (C*) has been estimated for any civic node, so this read treats the claim as a hypothesis to test, not a finding. The two hypotheses, and what they rest on: Decoupling at the foreign-policy node (hypothesized). Trust in the federal handling of international problems fell 12 points in one year, about 10 in every party group, under a war that began in late February. That fits decoupling. It also fits ordinary war-weariness, which standard opinion models explain without ATT, so it does not single out decoupling. [Established data, Extension reading] The institutional floor (undetermined). Average confidence in 14 institutions fell from as high as 43% in the early 2000s to below 30% every year since 2022, and has held at 26–28% for four years (Gallup, 2024). The Codex's flattening signature is slower recovery, not a low level. A basin that drifted down slowly and is now stable fits these data as well as an eroding one. Recovery-time estimates (dashboard) will decide. [Established data, Extension reading] Five observations carry the read: Federal trust is at a floor, and the floor is partisan. The executive is at 33% and the judiciary at 46%, both record lows. Republicans trust every federal entity except Congress at 70% or more; Democrats are in single digits for the executive. Independents' trust on all five federal measures is the lowest Gallup has recorded for them. That is a large third group outside both partisan basins, which complicates the two-basin picture. [Established] Abroad, views of the U.S. are coupled to its president. Across 23 countries, national favorability toward the U.S. and confidence in its president correlate at 0.90; within countries, the median individual-level correlation is 0.55 (Pew Decoded). That is strong coupling, not a one-for-one relationship. [Established] The war is the main terrain driver, and it is in a pause. October is the most intense month of maritime attacks around Hormuz since March. On October 8 the president pledged no U.S. strikes on Iran before the November 3 election, and talks over a U.S. counterproposal are under way, while a third carrier moves to the region (WDBJ/AP). [Established] Inflation is an energy shock carried by the same driver. Headline CPI is 3.7%, core 2.6%. [Established] Local membranes are holding, not gaining. Local government trust is 67%, inside its 1972–2026 range of 63–77%. [Established] Why the mode would matter. If the hypotheses hold, the remedies differ: decoupling is met by slowing the drivers, flattening by rebuilding institutions' capacity to recover. Those remedies are normative recommendations that follow from the framework, not findings. [Extension] Hysteresis. In bistable systems, reversing the driver that moved a system into a new basin does not by itself bring it back. [Established for bistable systems] Whether civic trust is bistable is untested here. [Extension] Near-term test. The November 3 midterms and battleground certification fall inside this read's window. Two live inputs: a lawsuit filed September 23 over possible armed federal agents at polling sites (WJLA/Sinclair), and the DC National Guard deployment, extended to January 2029 (PBS). [Established] Trajectory judgment. Extended Interstice (Scenario B) is the most likely path, at 45%, revised up after independent review. The scenario set now has a precedence rule and an "Other" category, so every outcome scores to exactly one scenario. [Extension]