Heresthetic Equilibrium: Collective Choice under Universal Manipulation

Heresthetics changes the choice problem before votes are counted: an actor may introduce a criterion, alter which alternatives remain organizationally executable, set an agenda, and vote strategically. The missing object has been an equilibrium in which every actor uses every instrument at once under finite budgets. This paper constructs one common finite game that retains full payoff-relevant history and asks what is durably installable in equilibrium, joining five earlier components of the same program: durable deletion and carrier hysteresis, the two-player security price, verified criteria versus voter response, append-only spatial reachability, and fixed-dimensional Condorcet-domain limits. An exact set-valued recursion defines ordinary heresthetic equilibrium and makes the global-consistency requirement for Markov compression explicit. The central result is a budgeted-core characterization: coalitionally secure durable Condorcet installation succeeds if and only if every hostile coalition's pooled budget is strictly below its route-specific blocking price, proved for necessity as well as sufficiency under an explicit product architecture. It nests the two-player opposed-interest contest and the reshape-free planner cost and locates its fixed-menu boundary inside the heresthetically stable and Banks sets. The paper is explicit about what the price theorem does not do. It does not characterize the possible outcomes of unrestricted general-sum subgame-perfect equilibrium: a public-good counterexample shows that affordable opposition may fail to coordinate, so no budget-only blocking test can succeed without a coalition protocol or preference restrictions. The dynamics of capacity are then resolved by resource regime rather than left to a single conjecture. In the primitive integer-stock model, restorative access neutralizes a one-time deletion while a two-basin law yields an exact lock-in inequality. With hard nonrenewable budgets, a positive intervention-price floor, and the stated horizon scaling, a sharp pathwise bound drives the occupation share of incomplete menus to zero, yet a separate construction shows that occupation share and terminal installation are distinct: one saved intervention placed just before handoff can still install a nonfundamental alternative for the whole durability window. With per-date replenishing resources, an exact two-player, three-alternative game has a unique role-equivariant equilibrium that deters a globally contractive passive recovery process at every date of every finite truncation, so the deep Banks winner is kept out of office with probability one, whereas the same holder under a fixed budget fails at a rate governed by a binomial tail. What remains open is stated narrowly, as a semialgebraic controlled-kernel frontier separating exhaustion, deterrence, and hysteretic lock-in, with an explicit obstruction rather than a broad claim. Local criterion-gap richness alone does not force dynamic chaos; universal installability follows only under stated history-compatible richness and equilibrium-lifting conditions. Possible pure heresthetic-equilibrium membership is PSPACE-complete in a succinct polynomial-horizon encoding and polynomial on an explicit tree, so fixed alternatives alone do not make the problem easy. A dynamic Banks correspondence and a limited identification result connect endogenous menus to sophisticated agendas and destroyed counterfactuals.