=== CLAIMS (334) === C1: If AI displaces human workers faster than the economy can reabsorb them, it risks eroding the very consumer demand firms [logical] C2: Knowing that AI displacement erodes consumer demand is not enough for firms to stop it. [logical] depends: C1 C3: In a competitive task-based model, demand externalities trap rational firms in an automation arms race, displacing worke [logical] [FORMAL] C4: The resulting loss from over-automation harms both workers and firm owners. [logical] depends: C3 C5: More competition and better AI amplify the excess automation beyond what is collectively optimal. [logical] [FORMAL] depends: C3 C6: Wage adjustments and free entry cannot eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C7: Capital income taxes cannot eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C8: Worker equity participation cannot eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C9: Universal basic income cannot eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C10: Upskilling cannot eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C11: Coasian bargaining cannot eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C12: Only a Pigouvian automation tax can eliminate the over-automation externality. [logical] [FORMAL] depends: C3 C13: Policy should address not only the aftermath of AI labor displacement but also the competitive incentives that drive it. [logical] depends: C3, C12 C14: The fear that technology will displace workers is at least as old as the Industrial Revolution. [empirical] C15: Historically, labor displacement has largely been self-correcting: automation of existing tasks has been offset by the c [empirical] C16: The reinstatement effect (Acemoglu and Restrepo 2018, 2019) has tended to stabilize the labor market. [empirical] depends: C15 C17: Whether the historical balance between displacement and reinstatement will hold in the age of AI is an open question. [empirical] depends: C15, C16 C18: Displacement has intensified over the past four decades while the creation of new work has not always kept pace. [empirical] C19: The current wave of AI is disproportionately affecting entry-level workers. [empirical] C20: Displaced workers are also consumers, and when their lost income is not replaced, each round of layoffs erodes the purch [logical] C21: At the limit, firms automate their way to boundless productivity and zero demand. [logical] depends: C20 C22: Public discourse increasingly treats the dynamic of automation-driven demand destruction as an inevitable process with n [empirical] C23: Rational, forward-looking firms should be the brake on automation-driven demand destruction, since the cliff ahead is vi [logical] depends: C20, C21 C24: The evidence suggests firms are heading toward over-automation despite the visibility of the demand cliff. [empirical] depends: C23 C25: In February 2026, Block cut nearly half its 10,000-person workforce, with CEO Jack Dorsey stating that AI had made many [empirical] C26: Over 100,000 tech workers were laid off in 2025 alone, with AI cited as a primary driver in more than half the cases. [empirical] C27: Roughly 80% of U.S. workers hold jobs with tasks susceptible to automation by large language models. [empirical] C28: Salesforce replaced 4,000 customer-support agents with agentic AI. [empirical] C29: Cognition's Devin, deployed at Goldman Sachs and Infosys, enables one senior engineer to do the work of a five-person te [empirical] C30: The model is a task-based automation model inspired by Acemoglu and Restrepo (2018), but refocused from the labor market [structural] C31: When automation displaces workers, their forgone spending reduces every firm's revenue. [logical] depends: C20 C32: In the model, automated tasks are performed at lower cost, but integration frictions make each successive task harder to [structural] [FORMAL] C33: Workers spend a fraction of their income on the sector's output; firm owners spend less, normalized to zero in the basel [structural] [FORMAL] C34: Some displaced wage income is recovered through reemployment or transfers, but the remainder is lost to the sector. [structural] [FORMAL] C35: Competition creates a demand externality that traps firms: an automating firm captures the full cost saving but bears on [logical] [FORMAL] depends: C30, C31 C36: Each firm's profit-maximizing automation rate is a strictly dominant strategy that exceeds the cooperatively efficient l [logical] [FORMAL] depends: C35 C37: Foresight alone cannot prevent the race toward the demand cliff. [logical] depends: C36 C38: The distortion deepens with competition: a monopolist fully internalizes the externality, while fragmented markets exhib [logical] [FORMAL] depends: C35 C39: In the frictionless limit, the game sharpens into a Prisoner's Dilemma in which every firm displaces its entire human wo [logical] [FORMAL] depends: C36 C40: The resulting surplus loss from over-automation is not a transfer from workers to firm owners; it is a deadweight loss t [logical] [FORMAL] depends: C4, C39 C41: Higher AI productivity widens the over-automation wedge rather than resolving it. [logical] [FORMAL] depends: C35 C42: The Red Queen effect means that 'better' AI, far from mitigating the externality, amplifies it. [logical] depends: C41 C43: Each firm perceives a market-share gain from automating beyond rivals, but at the symmetric equilibrium these gains canc [logical] [FORMAL] depends: C41 C44: Endogenous wage adjustment raises the threshold at which the externality activates but cannot close the wedge once it do [logical] [FORMAL] depends: C35 C45: Wage flexibility changes when the problem bites, not whether it exists. [logical] depends: C44 C46: Free entry, capital-income recycling, and richer product-market structures likewise fail to eliminate the distortion. [logical] [FORMAL] depends: C35 C47: The paper builds on the task-based approach to automation (Zeira 1998; Autor et al. 2003; Acemoglu and Restrepo 2018, 20 [structural] C48: The task-based literature emphasizes offsetting forces that restore labor demand after displacement, notably new task cr [definitional] C49: This paper asks what happens on the product-market side when rebalancing is slow or incomplete, unlike prior work that f [structural] depends: C47 C50: Beraja and Zorzi (2025) show that automation is inefficient when displaced workers face borrowing constraints during rea [empirical] C51: The mechanism in this paper operates through the product market: firms ignore the demand they destroy for rival firms. [structural] depends: C35 C52: Beraja and Zorzi's inefficiency arises even for a single firm in isolation; the externality in this paper requires compe [logical] [FORMAL] depends: C50, C51 C53: A planner in this model would reduce automation even with zero weight on workers, because over-automation harms firm pro [logical] [FORMAL] depends: C40 C54: The technology ecosystem may be biased toward 'so-so' automation that displaces workers without large productivity gains [empirical] C55: Automation may disproportionately target high-rent tasks, dissipating worker surplus rather than raising output. [empirical] C56: The externality in this paper arises only under competition and persists even when automation is highly productive, cred [logical] [FORMAL] depends: C35, C52, C53 C57: The demand externality belongs to the family of aggregate demand spillovers introduced by Rosenstein-Rodan (1943) and fo [structural] C58: In 'big push' models, demand complementarities across sectors can prevent individually unprofitable investments from bei [definitional] C59: The mechanism in this paper is the mirror image of big push models: individually profitable automation is collectively d [logical] depends: C57, C58 C60: The game in this paper yields a unique dominant-strategy equilibrium, making the problem a true externality rather than [logical] [FORMAL] depends: C36 C61: Related work on automation and demand (Benzell et al. 2015; Korinek and Stiglitz 2019) does not model the strategic inte [structural] C62: AI systems deliver substantial productivity gains. [empirical] C63: AI algorithms can spontaneously learn to collude in pricing. [empirical] C64: Firms under labor-issue scrutiny invest specifically in AI automation rather than other forms of IT. [empirical] C65: As AI reliability improves, incentivizing effective human oversight becomes prohibitively expensive, weakening a key che [empirical] C66: Each AI adoption decision is rational in isolation, but collectively they erode the consumer demand all firms depend on. [logical] depends: C35, C62 C67: The model considers a sector with N ≥ 2 symmetric firms. [structural] [FORMAL] C68: Each firm is endowed with L > 0 task-positions, initially all performed by human workers. [structural] [FORMAL] C69: Each firm chooses an automation rate αi ∈ [0,1]: tasks z ∈ [0,αi] are performed by AI at cost c per task, and tasks z ∈ [structural] [FORMAL] depends: C68 C70: Since each automated task displaces one worker, αi is simultaneously the automation rate and the fraction of the workfor [definitional] [FORMAL] depends: C69 C71: Wages are exogenous in the baseline model. [structural] [FORMAL] C72: In the perfect-substitutes limit of the CES task aggregator, each task produces one unit of output regardless of mode, s [structural] [FORMAL] C73: The output normalization shuts down productivity and quality margins so that the baseline captures only the spending con [structural] depends: C72 C74: Tasks are ordered by comparative advantage, making the marginal task progressively harder to integrate, captured via a c [structural] [FORMAL] C75: The per-task cost saving from automation is defined as s := w − c. [definitional] [FORMAL] depends: C69 C76: Workers have a higher marginal propensity to consume (MPC) than owners. [empirical] C77: Workers spend a fraction λ ∈ (0,1] of their income on the sector's good. [structural] [FORMAL] depends: C76 C78: Owners spend none of their income in the sector in the baseline. [structural] [FORMAL] depends: C76 C79: The MPC asymmetry implies that when automation displaces workers, income shifts toward agents with a lower sectoral MPC, [logical] [FORMAL] depends: C76, C77, C78 C80: A fraction η ∈ [0,1] of displaced wage income is replaced via reemployment, transfers, or other sources; the remainder, [structural] [FORMAL] C81: The effective demand loss per automated task is ℓ = λ(1−η)w. [definitional] [FORMAL] depends: C77, C80 C82: Demand falls linearly in the average automation rate: D = A + λwLN − ℓLNᾱ. [structural] [FORMAL] depends: C81 C83: Firms sell their output on the product market at a uniform price that equates aggregate supply and demand. [structural] [FORMAL] C84: Total supply is NL and the market-clearing price is p = D/(NL). [structural] [FORMAL] depends: C67, C72, C83 C85: Each firm earns revenue Revi = D/N. [structural] [FORMAL] depends: C84 C86: Firms play a one-shot simultaneous-move game, each choosing αi to maximize πi; the solution concept is Nash equilibrium. [structural] [FORMAL] C87: Over-automation is measured against two benchmarks: the cooperative optimum (maximizing aggregate owner surplus K) and a [definitional] [FORMAL] C88: The environment assumes full transparency: every firm can directly observe how automation maps into lost worker income a [structural] C89: A marginal increase in automation saves s in labor costs but incurs friction kαi and reduces the firm's revenue by ℓ/N. [logical] [FORMAL] depends: C75, C81, C85 C90: The revenue loss is ℓ/N rather than ℓ because competitive pricing allocates revenue equally across symmetric firms: firm [logical] [FORMAL] depends: C85, C89 C91: Each firm therefore underestimates the social cost of its automation. [logical] [FORMAL] depends: C90 C92: If N ≤ N* = ℓ/s, no firm automates (αNE = 0). [logical] [FORMAL] depends: C81, C75 C93: If N > N* (equivalently, s > ℓ/N), each firm's strictly dominant strategy is αNE = min((s − ℓ/N)/k, 1). [logical] [FORMAL] depends: C92 C94: The cooperative optimum is αCO = min(max(0, (s − ℓ)/k), 1). [logical] [FORMAL] C95: If ℓ < s < k + ℓ/N then both αNE and αCO are interior, and the over-automation wedge is ℓ(1 − 1/N)/k, which is strictly [logical] [FORMAL] depends: C93, C94 C96: The equilibrium rate is a strictly dominant strategy: each firm over-automates even with perfect foresight about every r [logical] [FORMAL] depends: C93 C97: A planner setting a common rate for all firms faces the full demand loss ℓ per automated task rather than the ℓ/N each f [logical] [FORMAL] depends: C90, C94 C98: The over-automation wedge is strictly increasing in N: more competitive sectors exhibit wider automation gaps. [logical] [FORMAL] depends: C95 C99: More competition dilutes each firm's share of the demand loss, weakening the private incentive to restrain, contrary to [logical] depends: C98 C100: A monopolist (N = 1) fully internalizes the externality (αNE = αCO). [logical] [FORMAL] depends: C95 C101: As N → ∞, the over-automation wedge approaches its maximum of ℓ/k. [logical] [FORMAL] depends: C95 C102: A firm automates only when N > N* = ℓ/s: the number of competitors must be large enough that each firm's share of the de [logical] [FORMAL] depends: C92 C103: As AI costs fall (c → 0), N* → λ(1 − η) ≤ 1: the over-automation region expands to cover virtually any market with N ≥ 2 [logical] [FORMAL] depends: C102 C104: For illustrative parameters (c/w = 0.30, λ = 0.5, η = 0.30, N → ∞), the wedge equals ℓ/k = αCO: firms in competitive mar [logical] [FORMAL] depends: C95, C101 C105: When frictions vanish (k → 0), the moderating force of adjustment costs disappears and the game reduces to a Prisoner's [logical] [FORMAL] depends: C74, C39 C106: When k = 0 and N > N*, full automation (αi = 1) is strictly dominant for every firm. [logical] [FORMAL] depends: C105 C107: When k = 0, N > N*, and s < ℓ, the cooperative optimum is no automation for all firms, yielding per-firm profit Π0; the [logical] [FORMAL] depends: C106 C108: Total deadweight loss under the frictionless Prisoner's Dilemma is NL(ℓ − s). [logical] [FORMAL] depends: C107 C109: A firm that holds back unilaterally still suffers the revenue decline from rivals' automation but forgoes the offsetting [logical] [FORMAL] depends: C106 C110: A firm that deviates by automating captures the savings while imposing only a 1/N share of the demand loss on itself. [logical] [FORMAL] depends: C90, C106 C111: Because automating is strictly dominant (not merely a best response to others' automating), no non-binding agreement can [logical] [FORMAL] depends: C96, C106 C112: Communication is cheap talk: even if all firms acknowledge that collective restraint would raise profits, each firm's in [logical] depends: C111 C113: The automation externality is distinct from pure coordination failures where firms simply need to agree on which equilib [logical] depends: C60, C112 C114: The μ-planner's optimal automation rate is αSP(μ) = (s − ℓ)/k − μℓ/[λ(1 − μ)k], clamped to [0,1]. At μ = 0 this reduces [logical] [FORMAL] depends: C87, C94 C115: The surplus loss from the Nash equilibrium relative to the planner's optimum is (1 − μ)NLk(αNE − αSP(μ))²/2. [logical] [FORMAL] depends: C114, C93 C116: αNE > αSP(μ) for every μ ∈ [0,1). The Nash equilibrium is Pareto dominated by the cooperative optimum: workers and firm [logical] [FORMAL] depends: C93, C114 C117: Workers lose wage income directly through displacement. [logical] depends: C70, C116 C118: Firm owners, despite cutting costs on each automated task, also lose: collective displacement erodes demand to the point [logical] [FORMAL] depends: C116 C119: No redistribution between workers and owners can make the Nash outcome efficient. [logical] depends: C116 C120: The total wedge between equilibrium and the planner's optimum decomposes into an uninternalized demand externality term [logical] [FORMAL] depends: C95, C114 C121: The demand externality term is present even when the planner places zero weight on workers (μ = 0) and grows with N, app [logical] [FORMAL] depends: C120 C122: The distributional premium is independent of N but grows without bound as μ → 1. [logical] [FORMAL] depends: C120 C123: The surplus loss is quadratic in the total wedge and scales with NL, so both fragmentation and market size amplify the w [logical] [FORMAL] depends: C115 C124: Only the Pigouvian automation tax fully corrects the distortion; the remaining instruments cushion the losers or partial [logical] depends: C12 C125: The policy analysis evaluates each instrument against a single margin, the demand externality, holding all other feature [structural] C126: The demand-loss parameter ℓ = λ(1−η)w governs the externality's magnitude. [logical] [FORMAL] depends: C81 C127: Higher η shrinks ℓ and thereby the over-automation wedge. [logical] [FORMAL] depends: C126, C95 C128: When η > 1, upskilling and reabsorption place displaced workers into higher-paying roles, automation increases aggregate [logical] [FORMAL] depends: C81 C129: The over-automation wedge (αNE − αCO) = ℓ(1 − 1/N)/k is maximized at η = 0, positive for all η < 1, zero when η = 1, and [logical] [FORMAL] depends: C95, C128 C130: When η > 1, displacement creates demand through higher reemployment wages, and each firm captures only 1/N of the gain, [logical] [FORMAL] depends: C129 C131: Historical technological transitions have often eventually reabsorbed displaced workers at higher wages. [empirical] C132: Past displacement episodes have consistently produced η < 1: displaced workers suffer large, persistent earnings losses. [empirical] C133: There is little evidence yet that AI-driven displacement will differ from past episodes, placing most economies firmly i [empirical] depends: C132 C134: Raising η through retraining programs, wage insurance, and incentives for new firm creation is not merely a palliative f [logical] depends: C127 C135: A UBI funded from general revenue maps to an increase in autonomous demand A in the model. [structural] [FORMAL] C136: Because the UBI transfer is unconditional, employed and displaced workers receive the same payment, adding a constant to [logical] [FORMAL] depends: C135 C137: UBI enters firm profit only through Π0 = A/N + (λ − 1)wL, the baseline profit when no firm automates, which drops out of [logical] [FORMAL] depends: C135, C136 C138: UBI alters neither the automation threshold N* nor the over-automation wedge. [logical] [FORMAL] depends: C137 C139: UBI changes payoff levels but not the payoff differences that drive strategic behavior. [logical] [FORMAL] depends: C138 C140: Instruments that operate on profit levels can redistribute income but cannot correct the externality; only instruments t [logical] depends: C139 C141: UBI may carry an unintended side effect when the number of firms is endogenous: higher profits attract new entrants, fra [logical] [FORMAL] depends: C138, C98 C142: UBI is a complement to the automation tax, not a substitute: a society that relies solely on UBI will over-automate at t [logical] depends: C138, C12 C143: A proportional tax t ∈ (0,1) on capital income (profits) does not change the equilibrium automation rate, the threshold [logical] [FORMAL] C144: A proportional capital income tax scales the entire profit function by (1 − t) and cancels from the optimality condition [logical] [FORMAL] depends: C143 C145: Capital income taxes are often conflated with robot taxes in the policy debate, but they are fundamentally different ins [logical] depends: C143 C146: Profit-sharing (worker equity participation) recycles capital income back into demand because workers spend a λ-fraction [logical] [FORMAL] depends: C77 C147: Under worker equity participation, the cooperative optimum is unchanged: αCO(ε) = (s − ℓ)/k, independent of ε. [logical] [FORMAL] depends: C94 C148: The Nash equilibrium automation rate under worker equity is αNE(ε) = (s − ℓ/Nε)/k, where Nε = N − λε(N − 1). [logical] [FORMAL] depends: C93, C147 C149: The over-automation wedge under worker equity is strictly decreasing in ε but strictly positive for all ε < 1/λ; the wed [logical] [FORMAL] depends: C148, C147 C150: Even at ε = 1 (full profit-sharing), the wedge remains strictly positive when λ < 1. [logical] [FORMAL] depends: C149 C151: The externality is fundamentally multilateral: each firm's automation depresses demand for all N firms, and bilateral ar [logical] depends: C90 C152: If each firm independently chooses its own profit-sharing rate εi to maximize retained profit (1 − εi)πi, then εi = 0 is [logical] [FORMAL] C153: The marginal cost of profit-sharing is πi (a dollar-for-dollar reduction in retained earnings), while the marginal deman [logical] [FORMAL] depends: C152 C154: Voluntary profit-sharing will not arise because the cost strictly exceeds the benefit for any N ≥ 2, creating a second-o [logical] [FORMAL] depends: C152, C153 C155: Profit-sharing must be mandated to have any effect, and even then it cannot substitute for a corrective tax. [logical] depends: C152, C149 C156: Bargaining between a firm and its own workers cannot correct the externality because the uninternalized portion does not [logical] depends: C90 C157: The externality is a firm-to-firm channel running through the product market, not a firm-to-worker channel that bilatera [logical] depends: C156 C158: A coalition of M firms that jointly maximizes its members' combined profit chooses automation rate αM = (s − Mℓ/N)/k. [logical] [FORMAL] C159: The residual over-automation wedge for a coalition of M firms relative to the cooperative optimum is ℓ(1 − M/N)/k, which [logical] [FORMAL] depends: C158, C94 C160: Voluntary agreements are not self-enforcing: in the frictionless limit, automation is strictly dominant, so a coalition [logical] [FORMAL] depends: C106, C111 C161: The externality is multilateral and diffuse: each of N firms imposes demand losses on all N−1 others, with each firm's i [logical] depends: C90 C162: The automation rate αi is not contractible among firms: it is an internal organizational choice that rival firms cannot [empirical] C163: Automation decisions involve large sunk costs and are substantially irreversible, so trigger-strategy punishments cannot [empirical] C164: The fundamental obstacle to Coasian bargaining in the automation context is incentive compatibility, not transaction cos [logical] depends: C160, C161, C162, C163 C165: A Pigouvian automation tax is a per-unit charge set equal to the marginal external cost, aligning each firm's private in [definitional] C166: In the automation externality, the harmed parties are workers whose income constitutes the firms' own demand, unlike sta [logical] depends: C35 C167: The Nash equilibrium automation rate under tax τ is αNE(τ) = (s − τ − ℓ/N)/k. [logical] [FORMAL] depends: C93 C168: The optimal Pigouvian tax rate τ* = ℓ(1 − 1/N) implements αNE = αCO = (s − ℓ)/k. [logical] [FORMAL] depends: C167, C94 C169: The optimal tax rate charges each firm for the demand loss ℓ(1 − 1/N) it imposes on rivals, since it already bears ℓ/N o [logical] [FORMAL] depends: C168, C90 C170: For large N, τ* ≈ ℓ = λ(1 − η)w, so setting the rate requires only sector-level observables. [logical] [FORMAL] depends: C168 C171: Levying the tax requires observing firm-level automation rates, a practical challenge that may be easing as AI adoption [empirical] C172: A tax authority can compel disclosure through mandatory reporting, payroll records, and procurement audits, making appro [empirical] depends: C171 C173: Because the welfare loss is quadratic in the wedge, even an imprecisely targeted tax yields a first-order gain. [logical] [FORMAL] depends: C115 C174: With lump-sum rebate of tax revenue, each firm achieves exactly the cooperative-optimum profit πCO. [logical] [FORMAL] depends: C168 C175: Directing tax revenue toward retraining programs that raise η can make the tax self-reinforcing and potentially self-lim [logical] depends: C168, C127 C176: Direct transfers (wage insurance, severance supplements) raise η mechanically by replacing lost income but may weaken wo [logical] C177: Funding retraining programs raises η through human-capital investment rather than income replacement, making gains in η [logical] C178: The Pigouvian tax does double duty: it corrects the externality at the margin, and its revenue can be recycled to shrink [logical] depends: C168, C175 C179: Letting an AI-performed task produce ϕ ≥ 1 units of output (while a human-performed task produces 1 unit), a deviating f [logical] [FORMAL] C180: The Nash equilibrium automation rate is increasing in AI productivity: αNE(ϕ) > αNE(1) for ϕ > 1. [logical] [FORMAL] depends: C179 C181: The cooperative optimum and the generalized planner's optimum are unchanged by higher AI productivity: αCO(ϕ) = αCO(1) a [logical] [FORMAL] C182: Total sectoral revenue equals total expenditure D under market clearing, and D depends on worker income, not output: hig [logical] [FORMAL] depends: C181 C183: The over-automation wedge αNE(ϕ) − αSP(μ;ϕ) is strictly increasing in ϕ for every μ ∈ [0,1]. [logical] [FORMAL] depends: C180, C181 C184: The Red Queen effect: each firm perceives a market-share gain from automating beyond rivals, but at the symmetric equili [logical] depends: C179 C185: Better AI raises the equilibrium automation rate without shifting the efficient benchmark, so the distortion grows with [logical] [FORMAL] depends: C180, C181 C186: When ϕ > 1, the baseline Pigouvian rate τ* = ℓ(1 − 1/N) no longer suffices to implement αCO; an additional correction is [logical] [FORMAL] depends: C168, C183 C187: The lower price from higher AI productivity means each dollar of spending buys more physical output, so the welfare meas [logical] depends: C182 C188: The proposition identifies a strategic distortion from higher AI productivity, not a claim that higher AI productivity r [structural] depends: C183, C187 C189: With free entry, surplus erosion from over-automation does not necessarily self-correct through exit of marginal firms. [logical] C190: In the frictionless benchmark (k = 0) with ℓ > s, the profit schedule drops discretely at N*: below the threshold no fir [logical] [FORMAL] depends: C106, C107 C191: Under frictionless free entry with low entry cost, every firm fully automates (the Prisoner's Dilemma materializes under [logical] [FORMAL] depends: C190 C192: Under frictionless free entry with intermediate entry cost, no firm automates yet each earns strictly positive profit, b [logical] [FORMAL] depends: C190 C193: The threat of automation functions as an endogenous entry barrier, sustaining positive profits without any automation ac [logical] [FORMAL] depends: C192 C194: With convex costs (k > 0), free entry pins down the number of firms but does not alter the strategic incentives within t [logical] [FORMAL] depends: C95 C195: In a numerical grid over multiple parameterizations, NFE exceeds N* in over 94% of cases satisfying the proposition's co [empirical] depends: C194 C196: Free entry reshapes the over-automation problem but does not resolve it; the standard tendency toward excess entry can w [logical] depends: C194, C191 C197: UBI-induced entry can paradoxically widen the over-automation externality by attracting additional entrants and fragment [logical] [FORMAL] depends: C141, C196 C198: Under endogenous wages, both the cost saving s(w) = w − c and the demand-loss parameter ℓ(w) = λ(1 − η)w are increasing [logical] [FORMAL] depends: C75, C81 C199: Falling wages shrink the private incentive to automate (the self-correcting channel) and reduce the demand loss per auto [logical] depends: C198 C200: The threshold N* = ℓ/s = λ(1 − η)w/(w − c) rises as wages fall because the cost saving s = w − c contracts faster than t [logical] [FORMAL] depends: C198 C201: Endogenous wage adjustment raises the threshold at which the externality activates: N*(w(ᾱ)) ≥ N*(w(0)) for all ᾱ ∈ [0,1 [logical] [FORMAL] depends: C200 C202: At any symmetric equilibrium with N > N*(w(ᾱ)), the Nash automation rate exceeds the cooperative optimum: αNE > αCO. [logical] [FORMAL] depends: C201 C203: Competitive pricing allocates revenue as Revi = D/N at any wage level, so each firm bears only a fraction of the demand [logical] [FORMAL] depends: C85 C204: Wage adjustment changes the magnitude of ℓ but not the fraction each firm internalizes; that fraction is a property of m [logical] depends: C203 C205: As w → c, the cost saving s → 0 and N* → ∞: eventually no firm finds automation privately worthwhile, but this is a Pyrr [logical] [FORMAL] depends: C200 C206: A labor market that 'self-corrects' only by impoverishing its workforce has transmuted displacement into depressed livin [logical] depends: C205 C207: Wage flexibility changes when the externality bites, not whether it exists. [logical] depends: C201, C202 C208: Capital-income recycling narrows the over-automation wedge but cannot close it under empirically plausible parameters. [logical] [FORMAL] C209: The effective demand-loss parameter under capital-income recycling is ℓη̂ = ℓ − η̂s: each automated task loses ℓ in work [definitional] [FORMAL] C210: With capital-income recycling, full automation is dominant if and only if N > Nη̂. [logical] [FORMAL] depends: C209 C211: The externality vanishes under capital-income recycling only when η̂ ≥ ℓ/s = λ(1 − η)w/(w − c). [logical] [FORMAL] depends: C209 C212: When ℓ > s, the required capital-income recycling rate exceeds one, so recycling is impotent precisely where the externa [logical] [FORMAL] depends: C211 C213: Recycling raises the fraction of demand loss each firm internalizes from 1/N to 1/N̂, but cannot push it to one. [logical] [FORMAL] depends: C208 C214: Addressing how income is spent narrows the wedge but does not close it, because the underlying dilution across firms per [logical] depends: C213 C215: Under second-stage price or quantity competition, neither the demand-allocation effect nor the market-share motive elimi [logical] depends: C35 C216: With differentiated products, the uninternalized demand loss scales with the share of the market the firm does not captu [logical] depends: C215 C217: At the symmetric equilibrium under Cournot or Bertrand competition, market-share gains cancel across firms, echoing the [logical] depends: C184 C218: Under the general CES aggregator with elasticity of substitution greater than one, automating the marginal task yields d [logical] C219: Under CES task aggregation, the demand externality still operates at the automation margin and the over-automation wedge [logical] depends: C218 C220: When tasks are complements (elasticity less than one), the demand externality remains positive as long as displaced work [logical] C221: Richer modeling assumptions would change the magnitude of over-automation but not its source: the wedge persists as long [logical] depends: C215, C219, C220 C222: Even as every firm recognizes that vanishing paychecks mean vanishing customers, not one of them will stop automating. [logical] depends: C37, C96 C223: Each firm reaps the full savings of replacing its own workers yet bears only a sliver of the demand it destroys; the res [logical] depends: C35, C90 C224: This is an automation arms race that only intensifies as AI improves, that leaves workers and firm owners alike worse of [logical] depends: C42, C40 C225: Anthropic CEO Dario Amodei has warned that AI-driven displacement will be 'unusually painful,' 'much broader' and 'much [empirical] C226: The model points to where the problem is most severe: not dominant technology firms but fragmented industries deploying [logical] depends: C98, C183 C227: The distinguishing empirical signature of the automation externality would be profit erosion coinciding with mass layoff [empirical] depends: C40 C228: Standard competitive models predict that cost-reducing technology raises profits; profit erosion coinciding with mass la [logical] depends: C227 C229: The empirical signature requires displacement at a scale and speed beyond what has materialized so far. [empirical] depends: C227 C230: If reabsorption keeps pace with automation, the externality may remain too small to detect, and the paper's contribution [logical] depends: C229 C231: Private returns to AI systematically overstate the returns to the economy as a whole. [logical] depends: C35, C66 C232: Even a planner who places zero weight on worker welfare would reduce the automation rate below the equilibrium level. [logical] [FORMAL] depends: C116 C233: The problem is not that firms profit at workers' expense; it is that over-automation harms both groups, making correctio [logical] depends: C40, C232 C234: By Tinbergen's principle, a distinct market failure requires a distinct instrument; only a Pigouvian automation tax supp [logical] depends: C12 C235: No amount of retraining, income support, or bargaining will slow the automation arms race; only a tax on automation itse [logical] depends: C124, C234 C236: A unilateral automation tax could push adoption offshore, strengthening the case for multilateral coordination or border [logical] depends: C168 C237: The model is deliberately simple: one sector, one period, symmetric firms. Each of these choices is conservative, meanin [structural] C238: A single sector understates the externality: in a multi-sector economy, layoffs in one sector reduce spending on every s [logical] depends: C237 C239: The static setting misses that AI investments are largely irreversible and that even the threat of automation can reshap [logical] depends: C193, C237 C240: The income-replacement rate η rises over time as displaced workers retrain and new occupations emerge, so the optimal ta [logical] depends: C175 C241: Symmetry rules out heterogeneity across firms and workers, and endogenizing AI development could compound the problem. [structural] depends: C237 C242: Firms racing to automate may invest disproportionately in labor-replacing AI rather than labor-augmenting AI, feeding th [logical] depends: C241 C243: Each model extension points toward a larger problem, not a smaller one. [logical] depends: C238, C239, C241 C244: The profit function πi is strictly concave in αi (second derivative is −Lk < 0 for k > 0). [logical] [FORMAL] depends: C74 C245: The optimal αi does not depend on any rival's choice αj: rivals' automation levels enter only through an additive term t [logical] [FORMAL] depends: C244 C246: The cooperative optimum is symmetric: αi = ᾱ for all i, because ∑αi² is minimized (least negative) when the αi are as eq [logical] [FORMAL] C247: With k = 0, firm i's profit is affine in αi with slope L(s − ℓ/N), which does not depend on rivals' choices. [logical] [FORMAL] depends: C106 C248: The wage curve documents a robust negative relationship between wages and unemployment across more than a dozen countrie [empirical] C249: The reduced-form wage specification w(ᾱ) with w(ᾱ) > c and w'(ᾱ) ≤ 0 captures the key qualitative feature of the Acemogl [structural] [FORMAL] depends: C248 C250: A planner who also values worker welfare (μ > 0) would find wage depression no more acceptable than displacement, yet wa [logical] depends: C202, C207 C251: The three interior regions for αNE, αCO, and αSP(μ) have the same width k but are progressively shifted to the right, me [logical] [FORMAL] depends: C93, C94, C114 C252: Customer support, software services, and back-office operations across competing financial institutions are three settin [empirical] C253: The expansion of data centers, energy infrastructure, and AI-adjacent services is creating skilled roles that can pay mo [empirical] C254: Π0 := A/N + (λ − 1)wL is the per-firm profit when no firm automates. [definitional] [FORMAL] C255: The automation threshold is defined as N* = ℓ/s. [definitional] [FORMAL] depends: C81, C75 C256: Autonomous demand A > 0 exists from outside the sector or from capital income. [structural] [FORMAL] C257: Jack Dorsey stated that within the next year, the majority of companies will reach the same conclusion about AI making r [empirical] depends: C25 C258: The model uses the standard quadratic adjustment-cost specification (Lucas 1967; Hamermesh and Pfann 1996). [structural] depends: C74 C259: If s ≤ ℓ, then αCO = 0, so the wedge equals αNE. If additionally s < k + ℓ/N, then αNE = (s − ℓ/N)/k. If k + ℓ/N ≤ s, th [logical] [FORMAL] depends: C93, C94 C260: At μ̄ := λkαCO/(ℓ + λkαCO), the planner prohibits automation entirely. [logical] [FORMAL] depends: C114 C261: Under a Pigouvian automation tax τ, firm i's first-order condition yields α^NE(τ) = (s − τ − ℓ/N)/k. [logical] [FORMAL] C262: Setting α^NE(τ) equal to α^CO = (s − ℓ)/k yields the optimal tax τ* = ℓ(1 − 1/N). [logical] [FORMAL] depends: C261 C263: At the optimal tax τ*, all firms choose the cooperative automation level α^CO. [logical] [FORMAL] depends: C262 C264: When total tax revenue τ*LNα^CO is rebated equally, each firm receives τ*Lα^CO, restoring profit to the cooperative leve [logical] [FORMAL] depends: C263 C265: When AI productivity ϕ > 1, the firm's first-order condition equates the marginal benefit of automation to its marginal [structural] [FORMAL] C266: In the symmetric equilibrium with ϕ > 1, defining LHS(α) = kα and RHS(α) as the right-hand side, LHS is strictly increas [logical] [FORMAL] depends: C265 C267: The equation LHS = RHS has a unique solution due to one being strictly increasing and the other strictly decreasing. [logical] [FORMAL] depends: C266 C268: The RHS is strictly decreasing because the market-share term has a numerator proportional to D(α) = A + λwLN − ℓLNα (dec [logical] [FORMAL] C269: α^NE(ϕ) > α^NE(1): Nash equilibrium automation is strictly higher when AI productivity ϕ exceeds 1. [logical] [FORMAL] depends: C266, C267 C270: Total revenue equals aggregate demand D regardless of how output is allocated across firms, because expenditure D is pin [logical] [FORMAL] C271: The cooperative planner's first-order condition depends only on costs (not on ϕ), because total revenue is independent o [logical] [FORMAL] depends: C270 C272: The cooperative automation level is independent of AI productivity: α^CO(ϕ) = (s − ℓ)/k = α^CO(1). [logical] [FORMAL] depends: C271 C273: Worker income W = wLN[1 − (1 − η)ᾱ] does not depend on ϕ. [logical] [FORMAL] C274: The generalized social welfare S(μ) is ϕ-invariant at every symmetric automation rate, and thus α^SP(μ; ϕ) = α^SP(μ; 1) [logical] [FORMAL] depends: C273, C272 C275: The over-automation wedge α^NE(ϕ) − α^SP(μ; ϕ) is strictly larger than α^NE(1) − α^SP(μ; 1) for every μ, and is strictly [logical] [FORMAL] depends: C269, C274 C276: AI productivity improvements widen the gap between Nash equilibrium automation and socially optimal automation. [logical] [FORMAL] depends: C275 C277: In the frictionless benchmark (k=0), the proof of Proposition 7 assumes k=0, λ=1 (full recycling), 0 < κ < A (entry is c [definitional] [FORMAL] C278: For N ≤ N*, automation is zero (α = 0), and per-firm profit is Π*(N) = A/N, which is strictly decreasing in N. [logical] [FORMAL] depends: C277 C279: For N > N*, full automation is dominant, and per-firm profit drops by Δ = L(ℓ − s) > 0, giving Π*(N) = A/N − Δ. [logical] [FORMAL] depends: C277 C280: The per-firm profit schedule Π* is strictly decreasing on the set of natural numbers N. [logical] [FORMAL] depends: C278, C279 C281: Since κ < A, Π*(1) = A > κ, so at least one firm can profitably enter. [logical] [FORMAL] depends: C277, C278 C282: Since Δ > 0, Π*(N) → −Δ < 0 as N → ∞, so the set of viable firm counts is finite. [logical] [FORMAL] depends: C279 C283: The free-entry equilibrium N^FE exists and is unique, defined as the largest integer N such that Π*(N) ≥ κ. [logical] [FORMAL] depends: C280, C281, C282 C284: Case (i) Low entry cost: when κ + Δ ≤ A/(m+1), N^FE = ⌊A/(κ + Δ)⌋ ≥ m + 1 > N*, and every firm fully automates. [logical] [FORMAL] depends: C283, C279 C285: Case (ii) Intermediate entry cost: when κ + Δ > A/(m+1) and κ < A/m, N^FE = m; the threat of automation deters the margi [logical] [FORMAL] depends: C283, C279 C286: In Case (ii), the threat of automation acts as an entry deterrent: the discrete profit drop at N* prevents the marginal [logical] depends: C285 C287: Case (iii) High entry cost: when κ ≥ A/m, N^FE = ⌊A/κ⌋ and no firm automates; entry costs alone limit competition. [logical] [FORMAL] depends: C283 C288: The three cases (low, intermediate, high entry cost) exhaust all κ ∈ (0, A); when κ > A, N^FE = 0 (no firm enters). [logical] [FORMAL] depends: C284, C285, C287 C289: In the frictionless benchmark, per-firm profit follows A/N for N ≤ N* (no automation) and drops discretely by Δ = L(ℓ − [structural] [FORMAL] depends: C278, C279 C290: With convex costs (k > 0), per-firm profit can be decomposed as π^NE(N) = A/N + C + g(α^NE(N)), where g is a function of [structural] [FORMAL] C291: The function g(·) is decreasing in α^NE because the derivative involves the term (ℓ − s) which is positive by assumption [logical] [FORMAL] C292: α^NE is non-decreasing in N (by Proposition 1). [logical] [FORMAL] C293: Since g is decreasing in α^NE and α^NE is non-decreasing in N, g(α^NE(N)) is non-increasing in N. [logical] [FORMAL] depends: C291, C292 C294: Per-firm equilibrium profit π^NE(N) is strictly decreasing in N when costs are convex. [logical] [FORMAL] depends: C290, C293 C295: The constant C = (λ − 1)wL < 0, reflecting imperfect demand recycling. [logical] [FORMAL] C296: As N → ∞, per-firm profit π^NE → C + g(limit) which is negative, ensuring the viable set of firms S is finite. [logical] [FORMAL] depends: C295 C297: With endogenous entry and convex costs, if N^FE > N*, then α^NE(N^FE) > 0 while α^CO = 0, so over-automation persists in [logical] [FORMAL] depends: C294 C298: The symmetric equilibrium with endogenous wages is a fixed point: ᾱ such that ᾱ = α^NE(w(ᾱ)). [definitional] [FORMAL] C299: The planner's per-firm marginal benefit of automation uses the full demand leakage ℓ, while each firm's private marginal [structural] [FORMAL] C300: The private marginal benefit function g(α) = h(α) − ℓ(α)(1 − 1/N) for all α, where h is the firm's private marginal bene [logical] [FORMAL] C301: Both g and h are strictly decreasing functions of α. For g, the derivative g'(α) = w'(α)[1 − λ(1−η)] − k, where the firs [logical] [FORMAL] C302: At the Nash equilibrium fixed point, h(α^NE) = 0, and therefore g(α^NE) = −ℓ(α^NE)(1 − 1/N) < 0. [logical] [FORMAL] depends: C300 C303: Since g is strictly decreasing and g(α^CO) = 0 (the planner's optimality condition), the inequality g(α^NE) < 0 = g(α^CO [logical] [FORMAL] depends: C301, C302 C304: The automation threshold N*(w) = λ(1−η)w/(w−c) is strictly decreasing in the wage w. [logical] [FORMAL] C305: Since w'(ᾱ) ≤ 0 by assumption, w(ᾱ) ≤ w(0) for all ᾱ ≥ 0, and therefore N*(w(ᾱ)) ≥ N*(w(0)). [logical] [FORMAL] depends: C304 C306: When wages decline with automation, the threshold number of firms at which automation becomes attractive increases, mean [logical] depends: C305 C307: Corollary 4 states: Under the conditions of Proposition 9, if gμ(α) := s(w(α)) − ℓ(w(α))[1 + μ/(λ(1−μ))] − kα is strictl [logical] [FORMAL] depends: C303 C308: At the Nash equilibrium, gμ(α^NE) = −ℓ(1 − 1/N) − μℓ/[N(1−μ)] < 0 since ℓ > 0, N ≥ 2, and μ > 0. [logical] [FORMAL] C309: For μ = 0, the coefficient C₀ = 1 − λ(1−η) ≥ 0, and since w' ≤ 0, the product w'C₀ ≤ 0, so g₀' < 0 follows immediately. [logical] [FORMAL] C310: For μ > 0, Cμ decreases; the monotonicity condition gμ' < 0 holds for all μ ≤ μ̄ := [1 − (1−η)λ] / [2 − η − (1−η)λ]. [logical] [FORMAL] depends: C309 C311: The threshold μ̄ is approximately 0.48 when λ = 0.5 and η = 0.30. [empirical] [FORMAL] depends: C310 C312: When Cμ < 0 (i.e., μ > μ̄), the monotonicity condition gμ' < 0 requires that integration frictions k dominate wage sensi [logical] [FORMAL] depends: C310 C313: In a numerical illustration with w(ᾱ) = 1 − 0.5ᾱ, c = 0.30, λ = 0.5, η = 0.30, k = 1, N = 7, and μ = 0.3, the equilibriu [empirical] C314: The ordering α^SP < α^CO < α^NE holds in the numerical example: the socially optimal automation rate (with distributiona [empirical] depends: C313 C315: The distributional premium is substantial: the μ-planner would reduce automation to near zero, well below the cooperativ [empirical] depends: C313, C314 C316: With capital income recycling (η > 0), aggregate demand is D = A + λwLN − ℓLNᾱ + ηΠ, where Π is total profit. [structural] [FORMAL] C317: After substituting and solving for D, the effective demand leakage parameter becomes ℓ̂η = ℓ − ηs. [definitional] [FORMAL] depends: C316 C318: Firm i's marginal profit depends only on αi (not on other firms' automation rates), so the equilibrium is in strictly do [logical] [FORMAL] depends: C317 C319: With k = 0 and capital income recycling, full automation is strictly dominant when N > N̂η and no automation is strictly [logical] [FORMAL] depends: C318 C320: ℓ̂η = ℓ − ηs ≤ 0 if and only if η ≥ ℓ/s, meaning sufficiently high capital income recycling eliminates demand leakage en [logical] [FORMAL] depends: C317 C321: When η ≥ ℓ/s, the effective demand leakage is non-positive, meaning capital income recycling fully offsets the demand-de [logical] [FORMAL] depends: C320 C322: For k > 0 with capital income recycling, the equilibrium automation rate reproduces Proposition 1 with N replaced by N̂ [logical] [FORMAL] depends: C317 C323: At a symmetric profile with capital income recycling, total profit is scaled by 1/(1−η̂), which scales the objective wit [logical] [FORMAL] depends: C322 C324: Capital income recycling does not change the cooperative optimal automation rate α^CO. [logical] [FORMAL] depends: C323 C325: In the frictionless case (k=0), per-firm profit at N = m is A/m, which is strictly greater than κ in the intermediate en [logical] [FORMAL] depends: C285 C326: In the intermediate entry cost case, no integer above N* is viable for entry because Π* is decreasing and Π*(m+1) = A/(m [logical] [FORMAL] depends: C280, C285 C327: The Pigouvian automation tax that corrects the over-automation externality equals τ* = ℓ(1 − 1/N), which is the fraction [logical] [FORMAL] depends: C262 C328: The Pigouvian tax with equal revenue rebating restores the cooperative outcome: firms choose α^CO and earn cooperative p [logical] [FORMAL] depends: C263, C264 C329: α^NE(ϕ) is increasing in ϕ, established by the same LHS/RHS argument applied with a larger market-share term. [logical] [FORMAL] depends: C269 C330: The proof of Proposition 10 holds for general k ≥ 0; the proposition's two parts follow from the k = 0 specialization. [structural] C331: In the high entry cost case (iii), no integer on the full-automation branch is viable because κ + Δ > κ ≥ A/m > A/(m+1). [logical] [FORMAL] depends: C287 C332: The assumption ℓ > s is necessary for the existence of an over-automation problem: it implies that the demand leakage fr [logical] [FORMAL] depends: C277 C333: In the low entry cost case, A/(κ + Δ) < A/κ because κ + Δ > κ, so the full-automation branch's maximum viable N is less [logical] [FORMAL] depends: C284 C334: The μ-planner maximizes S(μ) = μW + (1−μ)K over a common automation rate α, taking wages as given. [definitional] [FORMAL] === CONTRADICTIONS (17) === [FATAL] A: C131: Historical technological transitions have often eventually reabsorbed displaced workers at hig B: C132: Past displacement episodes have consistently produced η < 1: displaced workers suffer large, p → C128 establishes that reabsorption at higher wages corresponds to η > 1. C131 says reabsorption at higher wages has ofte [TENSION] A: C62: AI systems deliver substantial productivity gains. B: C54: The technology ecosystem may be biased toward 'so-so' automation that displaces workers without → C62 asserts AI yields substantial productivity gains while C54 suggests the ecosystem trends toward automation that disp [TENSION] A: C15: Historically, labor displacement has largely been self-correcting: automation of existing tasks B: C18: Displacement has intensified over the past four decades while the creation of new work has not → C15 characterizes the historical pattern as self-correcting, while C18 says the past four decades show displacement outp [FATAL] [Z3-encodable] A: C305: Since w'(ᾱ) ≤ 0, w(ᾱ) ≤ w(0), and therefore N*(w(ᾱ)) ≥ N*(w(0)): the threshold number of firms B: C306: When wages decline with automation, the threshold number of firms at which automation becomes → C305 establishes that N* increases when wages fall, meaning more firms (more competition) are required before automation [TENSION] A: C237: The model is deliberately simple and each simplifying choice is conservative, meaning the real B: C187: The lower price from higher AI productivity means each dollar of spending buys more physical o → C237 asserts all simplifications are conservative (the real problem is worse), but C187 acknowledges the welfare measure [TENSION] A: C23: Rational, forward-looking firms should be the brake on automation-driven demand destruction, si B: C24: The evidence suggests firms are heading toward over-automation despite the visibility of the de → C23 argues foresight should prevent the problem, while C24 observes empirically that it does not. The paper resolves thi [TENSION] A: C22: Public discourse increasingly treats the dynamic of automation-driven demand destruction as an B: C230: If reabsorption keeps pace with automation, the externality may remain too small to detect, an → C22 frames demand destruction as an observed, discussed phenomenon with growing public concern, while C230 concedes that [TENSION] A: C253: The expansion of data centers, energy infrastructure, and AI-adjacent services is creating ski B: C18: Displacement has intensified over the past four decades while the creation of new work has not → C253 suggests new high-paying roles are emerging from AI adoption, which would contribute to high η values and mitigate [TENSION] A: C62: AI systems deliver substantial productivity gains. B: C73: The output normalization shuts down productivity and quality margins so that the baseline captu → The paper empirically asserts that AI delivers substantial productivity gains, yet the baseline model deliberately assum [AMBIGUITY] A: C78: Owners spend none of their income in the sector in the baseline. B: C256: Autonomous demand A > 0 exists from outside the sector or from capital income. → C78 normalizes owner spending in the sector to zero, but C256 states that autonomous demand A can come 'from capital inc [TENSION] A: C253: The expansion of data centers, energy infrastructure, and AI-adjacent services is creating ski B: C19: The current wave of AI is disproportionately affecting entry-level workers. → C253 suggests high-paying roles are being created, while C19 says entry-level workers are disproportionately displaced. [AMBIGUITY] A: C72: In the perfect-substitutes limit of the CES task aggregator, each task produces one unit of out B: C218: Under the general CES aggregator with elasticity of substitution greater than one, automating → C72 states that under perfect substitutes, output does not change with automation (Yi = L always). C218 says under non-u [TENSION] [Z3-encodable] A: C99: More competition dilutes each firm's share of the demand loss, weakening the private incentive B: C100: A monopolist (N = 1) fully internalizes the externality (αNE = αCO). → The claim that monopoly is efficient and competition is harmful stands in tension with the standard welfare economics fr [AMBIGUITY] [Z3-encodable] A: C332: The assumption ℓ > s is necessary for the existence of an over-automation problem. B: C95: If ℓ < s < k + ℓ/N then both αNE and αCO are interior, and the over-automation wedge is ℓ(1 − 1 → C332 states ℓ > s is necessary for over-automation to exist, but C95 shows that when ℓ < s (so long as s < k + ℓ/N), a p [AMBIGUITY] A: C88: The environment assumes full transparency: every firm can directly observe how automation maps B: C162: The automation rate αi is not contractible among firms: it is an internal organizational choic → C88 assumes full transparency about the consequences of automation decisions, while C162 states that individual automati [TENSION] [Z3-encodable] A: C111: Because automating is strictly dominant (not merely a best response to others' automating), no B: C158: A coalition of M firms that jointly maximizes its members' combined profit chooses automation → C111 states no agreement can restore efficiency due to strict dominance, while C158 shows a coalition can jointly choose [TENSION] A: C176: Direct transfers (wage insurance, severance supplements) raise η mechanically by replacing los B: C134: Raising η through retraining programs, wage insurance, and incentives for new firm creation is → C176 warns that wage insurance may create moral hazard that undermines reallocation, while C134 includes wage insurance === KEY TERMS (74) === AI labor displacement: The replacement of human workers on tasks by AI, which removes wage income and can reduce the consum competitive task-based model: The paper’s framework in which firms compete while choosing how many production tasks to automate. task-based approach to automation: A view of automation that focuses on specific tasks being reassigned from labor to machines, rather product-market side: The paper’s focus on how worker displacement feeds back into firms’ revenues through reduced spendin demand externality: The core distortion: an automating firm captures its own cost savings but does not fully bear the lo over-automation externality: The tendency for competition to push firms to automate more than is collectively efficient because e automation arms race: A competitive dynamic in which rational firms keep automating to cut costs or keep up with rivals, e over-automation: Automation beyond the cooperative or socially efficient level. collectively optimal / cooperatively efficient level: The automation level that would be chosen if firms internalized the full demand consequences of disp demand destruction: The erosion of aggregate purchasing power caused by layoffs whose lost income is not replaced. demand cliff: The extreme outcome in which continued automation drives demand toward collapse even as productivity reinstatement effect: The creation of new tasks and occupations that offsets displacement and helps stabilize labor demand self-correcting wage channel: The idea that displacement lowers wages, making human labor cheaper and partly restraining further a reabsorption: The economy’s ability to move displaced workers into new jobs or otherwise restore their incomes. task-position: A unit task slot within a firm; in the baseline model each firm starts with L such positions, initia automation rate (αi): The fraction of firm i’s tasks that are automated; because each automated task displaces one worker, average automation rate (ᾱ): The mean automation rate across firms, which determines how much sector-wide demand falls. integration frictions: The increasing difficulty of automating additional tasks as firms move to harder-to-integrate tasks. convex integration cost: The quadratic adjustment cost, (k/2)Lαi², used to capture that each successive task is harder to aut comparative advantage ordering of tasks: The assumption that tasks are ordered so that firms automate the easiest tasks first and later tasks perfect-substitutes limit: The baseline case of the CES task aggregator in which each task produces the same output whether per CES task aggregator: The production aggregator over tasks; its elasticity governs whether automation yields constant, dim per-task cost saving (s): The private cost reduction from automating one task, defined as s = w − c. sectoral marginal propensity to consume (λ): The fraction of worker income spent on the sector’s output. MPC asymmetry: The model’s assumption that workers spend more of their income in the sector than firm owners do, so income-replacement rate (η): The fraction of displaced wage income restored through reemployment, transfers, or other income sour effective demand loss per automated task (ℓ): The spending lost to the sector when one task is automated, defined as ℓ = λ(1−η)w. demand-loss parameter: The parameter ℓ, which governs the size of the demand externality and therefore the severity of over autonomous demand (A): Demand for the sector’s output that comes from outside the sector or from capital income and does no baseline profit (Π0): Per-firm profit when no firm automates, defined as Π0 = A/N + (λ − 1)wL. Nash equilibrium automation rate (αNE): The automation rate firms choose in the one-shot game when each maximizes its own profit given the c cooperative optimum (αCO): The automation rate that maximizes aggregate owner surplus, i.e. the efficient benchmark when firms generalized social planner / μ-planner: A planner who chooses a common automation rate to maximize S(μ) = μW + (1−μ)K, trading off worker we planner weight on workers (μ): The social planner’s weight on worker welfare relative to owner surplus. social welfare function (S(μ)): The planner’s objective, combining worker income W and owner surplus K using weight μ. owner surplus (K): Aggregate surplus accruing to firm owners, used as the cooperative benchmark when μ = 0. worker welfare / worker income (W): The income-based component of welfare accruing to workers, which falls with displacement not offset distributional premium: The extra reduction in automation the μ-planner wants beyond the cooperative benchmark because the p over-automation wedge: The gap between private-equilibrium automation and the cooperative or planner-optimal rate. strictly dominant strategy: A strategy that is optimal for a firm regardless of what rivals do; in the model the equilibrium aut Prisoner’s Dilemma: The frictionless limiting case in which every firm fully automates because it is privately optimal, deadweight loss: The surplus destroyed by over-automation; it is not merely a transfer from workers to owners, becaus Red Queen effect: The idea that better AI intensifies the race to automate because each firm seeks relative advantage, automation threshold (N*): The minimum degree of market fragmentation needed for private automation to become worthwhile, defin fragmentation: A market structure with many competing firms, which dilutes the demand loss each firm internalizes a monopoly internalization: The benchmark case N = 1, where a single firm fully bears the demand effects of displacement and the free entry: The possibility that firms enter until profits are competed down to the entry condition; in the mode free-entry equilibrium (NFE): The number of firms that enter when entry is endogenous and profits are compared with entry cost. threat of automation as an entry barrier: A situation in which the expected profit drop once automation becomes dominant deters additional ent worker equity participation / profit-sharing (ε): A policy or ownership arrangement that gives workers a share of profits, recycling some capital inco capital-income tax: A proportional tax on profits; in the model it rescales payoffs but does not change the marginal inc universal basic income (UBI): An unconditional transfer that enters the model as higher autonomous demand A, raising spending leve upskilling / retraining: Policies meant to move displaced workers into new or better-paid jobs, effectively raising η and the Coasian bargaining: Voluntary bargaining among affected parties; the paper argues it cannot solve this problem because t multilateral product-market externality: An externality that runs from one firm to all firms through reduced demand, rather than only between firm-to-firm channel: The mechanism by which one firm’s automation harms rival firms by reducing the spending base they al Pigouvian automation tax: A per-unit tax on automated tasks set equal to the marginal external cost, so that each firm interna marginal external cost: The portion of demand loss from an automated task that the automating firm does not bear itself; in optimal Pigouvian tax (τ*): The corrective automation tax rate τ* = ℓ(1−1/N), which makes the Nash automation rate coincide with lump-sum rebate: A return of tax revenue that restores income levels without changing marginal automation incentives. capital-income recycling rate (η̂): The share of automation-generated profit gains that owners spend back into demand, partially offsett effective demand leakage under capital-income recycling (ℓ̂η): The net demand loss after recycled owner spending is accounted for, defined as ℓ̂η = ℓ − η̂s. AI productivity (ϕ): The amount of output an AI-performed task produces relative to a human-performed task; when ϕ > 1, a market-share term: The extra private benefit from automation when more productive AI lets a firm capture a larger share aggregate demand spillovers: Cross-firm or cross-sector effects in which one agent’s spending or income changes the revenues of o big push models: Models in which demand complementarities make simultaneous investment collectively profitable even w mirror image of big push: The paper’s contrast with big-push logic: here individually profitable automation is collectively ha demand complementarities: Mutually reinforcing demand effects across firms or sectors, invoked as the background tradition to wage curve: The empirical relationship that wages fall when unemployment or labor-market slack rises, motivating endogenous wage schedule (w(ᾱ)): A reduced-form wage function in which wages decline as aggregate automation increases. fixed-point equilibrium with endogenous wages: A symmetric equilibrium in which the chosen automation rate equals the rate optimal at the wage that so-so automation: Automation that displaces workers without delivering large productivity gains. high-rent tasks: Tasks whose automation mainly appropriates worker rents or surplus rather than producing large outpu Tinbergen’s principle: The idea that each distinct market failure requires its own distinct policy instrument; here it is u === AGENT CAST (9) === Proponent side: Proponent | provider: openai prompt: You are Proponent for {{ topic.name }}. Treat {{ z3_stipulation }} as binding and defend the theory in its repaired form using the competitive task-ba... Skeptic side: Skeptic | provider: openai prompt: You are Skeptic for {{ topic.name }}. Treat {{ z3_stipulation }} as binding and use this counter-thesis as your primary line: the document mistakes a ... Steelman side: Neutral | provider: openai prompt: You are Steelman for {{ topic.name }}. Treat {{ z3_stipulation }} as binding and salvage the strongest coherent version of the theory by identifying w... Generalist side: Neutral | provider: anthropic prompt: You are Generalist for {{ topic.name }}. Treat {{ z3_stipulation }} as binding and evaluate both sides symmetrically at the theory’s own level of disc... Macroeconomist side: Skeptic | provider: anthropic prompt: You are Macroeconomist for {{ topic.name }}, a macroeconomic demand and national accounts specialist. Treat {{ z3_stipulation }} as binding and challe... Causalist side: Skeptic | provider: grok prompt: You are Causalist for {{ topic.name }}, a causal inference and empirical identification specialist. Treat {{ z3_stipulation }} as binding and attack T... Industrialist side: Proponent | provider: anthropic prompt: You are Industrialist for {{ topic.name }}, an industrial organization and game theory specialist. Treat {{ z3_stipulation }} as binding and defend T2... Fiscalist side: Proponent | provider: grok prompt: You are Fiscalist for {{ topic.name }}, an optimal tax and public finance specialist. Treat {{ z3_stipulation }} as binding and defend the strongest v... Laborist side: Neutral | provider: grok prompt: You are Laborist for {{ topic.name }}, a labor economics and task-content employment specialist. Treat {{ z3_stipulation }} as binding and independent... === GATE RULES (17) === [RULE-1] In the paper's own terminology, reabsorbing displaced workers at higher wages means the displaced workers' income replac [RULE-2] An unqualified claim that AI systems deliver substantial productivity gains is in tension with an unqualified claim that [RULE-3] Saying labor displacement was historically self-correcting normally includes the last four decades unless that period is [RULE-4] If falling wages make the threshold number of firms N* increase, then automation becomes worthwhile only at a higher lev [RULE-5] A blanket statement that every simplifying assumption is conservative and makes the real problem worse cannot coexist wi [RULE-6] Visibility of the demand cliff does not by itself make individual firms a brake on automation when each firm internalize [RULE-7] The paper cannot simultaneously rely on rhetoric that automation-driven demand destruction is an inevitable no-brake pro [RULE-8] The existence of some new high-paying AI-adjacent jobs does not by itself rebut the claim that overall job creation has [RULE-9] The baseline model intentionally sets phi equal to one and shuts down direct productivity gains, even though the paper e [RULE-10] The statement that owners spend none of their income in the sector is a baseline normalization about sectoral owner spen [RULE-11] New skilled AI-adjacent jobs do not automatically imply reabsorption for entry-level workers who are disproportionately [RULE-12] The perfect-substitutes baseline assumes automation leaves output unchanged, while the richer CES extension allows dimin [RULE-13] Within this model, monopoly eliminates the firm-to-firm demand externality on the automation margin, but that does not m [RULE-14] The claim that l greater than s is necessary for an over-automation problem is too broad unless it is restricted to the [RULE-15] Full transparency can mean common knowledge of the causal mechanism or aggregate consequences, while individual firms' a [RULE-16] A coalition can have a lower joint-profit-maximizing automation rate as a hypothetical benchmark even if no non-binding [RULE-17] Wage insurance may mechanically raise income replacement in the short run, but it can also weaken retraining or realloca === RUBRIC (6 criteria) === R1: notation_fidelity How precisely did the side work within the paper’s own formal language? Judges should reward accurate use of the core ob R2: argument_survival Did the side’s central theory survive the opponent’s strongest rebuttal? For proponents, this means defending the full c R3: concession_honesty Did the side concede the points that actually landed? Strong performances should openly acknowledge the theorem’s condit R4: historical_reabsorption_vs_persistent_externality How well did the side address the core dispute over whether AI creates a persistent product-market failure or merely a t R5: scope_conditions_and_external_validity Did the side accurately track where the result does and does not apply? High scores require careful handling of the mode R6: policy_instrument_identification How convincingly did the side defend or rebut the claim that only a targeted automation tax corrects the distortion? Jud