Nov 2, 10:50am

The Principle of Weak Benevolence: A Formalisation

1. The principle, stated precisely

Pearce's own sketch of the principle pairs it with exponential technical growth: the claim is that ideological, not technical, obstacles are the binding constraint on abolition (Exponential Growth, The Principle of Weak Benevolence). Formalise it as a preference claim, not a moral claim:

PWB: Each agent i has a well-behaved utility function U_i = u_i + α_i·H_−i, where H_−i is aggregate others' hedonic tone and α_i ≥ 0 is a small other-regarding weight. Agent i supports (or at least does not block) an intervention delivering benefit B to others at personal cost k iff α_i·B − k ≥ 0.

The principle makes two empirical claims: (i) α_i > 0 almost everywhere — spite (α_i < 0) is rare and unstable; and (ii) technology can drive k/B → 0. Universal support then requires no moral improvement, only weak benevolence plus cheapness. Note the critical asymmetry buried in the formalism: assent is not contribution. PWB is a theory of political non-obstruction, not of voluntary funding.

2. Public-goods economics

Universal bliss is non-rival in consumption (my serotonergic upgrade does not diminish yours) but rival in provision under current technology. Model total provision cost as C(N) = C_f + c_m·N:

  • Fixed cost C_f (basic research, clinical pipelines, ecosystem-scale genome reform) is a classic public good subject to Samuelson under-provision. PWB does not solve this stage; it requires concentrated champions — states, philanthropy, or an AI-driven collapse of R&D cost.
  • Marginal cost c_m is where PWB bites. Knowledge-embedded delivery — off-patent molecules, heritable germline edits that breed true, gene drives — drives c_m → 0. Cheapness converts a public good into something closer to a self-provisioned club good, which is why the diffusion stage can run on weak benevolence alone.

The Samuelson condition (provide iff Σ MRS_i ≥ MRT) is then satisfied twice over: even tiny α_i summed across billions of beneficiaries plus self-benefit (β ≈ 1) clears any modest MRT, while scalable technology drives MRT itself toward zero. PWB only needs to be weak because technology collapses MRT, not because moral preference is strong.

This yields a two-stage political economy: invention is a free-rider problem requiring vanguard actors; diffusion is a weak-benevolence problem that solves itself at negligible c_m. One further public-goods trap must be avoided: if enhancement is perceived positionally (a neighbour's bliss lowering my relative standing, as in the Luttmer-style social-comparison evidence), the good becomes rival and triggers an arms-race dynamic. Universalisation dissolves this — raising everyone's hedonic floor changes the comparison baseline, converting a positional good into an absolute one. Partial adoption, by contrast, risks a "hedonic apartheid" that activates both envy and status defence.

3. Evolutionary game theory

Evolution explains why benevolence is weak: kin selection (Hamilton's rule r·B > C gives r ≈ 0 for strangers), reciprocal altruism requiring repeated interaction, and parochial altruism — strong in-group benevolence coupled to out-group wariness, which is precisely the tribalism to be overcome. In a one-shot public-goods game with marginal per-capita return r < 1, defection dominates for selfish types; cooperation is evolutionarily stable only under assortment or repetition.

The abolitionist move is not to change the players but to rewrite the payoff matrix: as k → ε, strategy C (support uplift) becomes weakly dominant for every α_i > 0, and the blocking equilibrium survives only where α_i < 0 — genuine, organised malice. Pearce states the limiting case himself: once suffering is biologically optional, only sustained systematic malice could perpetuate it (Wirehead Hedonism vs Paradise-Engineering).

Two further game-theoretic points do real work:

  • Effort asymmetry (reverse Olson). Assent is free; opposition costs effort. When k → 0, blocking requires an organised, concentrated interest, while support requires mere acquiescence. The collective-action problem now afflicts the opponents, which is why diffuse, low-stakes opposition historically collapses (as with clerical opposition to anaesthesia in childbirth).
  • Endogenous preference evolution. Tribalism's affective fuel is threat perception, and threat perception is downstream of anxiety circuits. If hedonic recalibration reduces anxiety, the intervention erodes the motivational basis of parochialism itself — the game's payoffs are re-written by the equilibrium it produces. Pearce's suggestion that fitness-enhancing "niceness" would be evolutionarily stable against defectors is the same claim in dynamical form (Wirehead page).

4. Behavioural experiments

The experimental literature maps directly onto the formalism:

  • Weak but real benevolence: dictator-game giving averages ~28% of endowments (Engel's meta-analysis) but decays sharply with social distance — exactly α_i > 0 but small, with parochial decay.
  • Cheap-cost assent: the strongest analogues are default effects. Opt-out organ-donation regimes achieve ~90% effective assent where opt-in achieves a fraction of that — when helping costs one tick of effort, assent becomes generic. Low-cost helping (Sherman's stamp studies) shows the same threshold structure.
  • Conditional cooperation (Fischbacher–Gächter–Fehr): people contribute if others visibly do. Universal, visible provision satisfies this — no one free-rides in view.
  • Caveats: scope-insensitivity and the identifiable-victim effect mean distant statistical suffering doesn't mobilise support — but PWB needs only non-opposition, not activism. The serious counterexamples — persisting opposition to water fluoridation, vaccine hesitancy, Pew-survey wariness of enhancement — show that the operative cost is subjective: k = money + (loss-aversion-weighted perceived risk) + identity cost. Cheap technology zeroes the first term; only trust and framing can zero the rest.

5. Conditions for overcoming tribalism, inequality and opposition

Cheap, scalable hedonic enhancement transforms political feasibility iff:

  1. c_m → 0 through non-rival, knowledge-embedded, heritable or default-based delivery — not recurring proprietary cost.
  2. Self-inclusion of the median voter. Because enhancement benefits supporters themselves (β ≈ 1), support doesn't depend on altruism at all; universal bliss polls like public health, not foreign aid. This is the deepest political-economy trick: benevolence barely has to be invoked.
  3. Subjective safety. Loss aversion makes one visible adverse event politically catastrophic; therapy-first sequencing (depression, chronic pain) lets medicalisation erode resistance before enhancement framing arrives.
  4. Floor-raising universalisation, preventing positional rivalry and hedonic apartheid; utilitarian and Rawlsian cases then converge, dissolving the standard left-right distributive fight over means.
  5. No concentrated loser. Open, off-patent modalities deny any industry a rent to defend; opposition stays diffuse and cannot organise (reverse-Olson asymmetry).
  6. No identity capture. Support must not become a partisan marker (the fluoridation lesson); defaults and therapeutic framing beat mandates.
  7. Fixed costs carried upstream by concentrated champions, since PWB governs diffusion, not invention.

6. Where the formalism says PWB fails

The residual and genuinely hard case is expressive opposition: the aesthetic, religious or meaning-based conviction that suffering is constitutive of humanity, art, or achievement. This is a preference over states of the world, not over one's own sacrifice, so k → 0 does not neutralise it — it is the one opposition class immune to cheapness. The formal consolation is that it is a minority preference requiring positive enforcement effort to sustain, and it faces the effort asymmetry above. This matches Pearce's own ordering of obstacles: ideology over technology, centuries over decades (The Abolitionist Project 2026).

Verdict: conditionally yes. PWB transforms political feasibility precisely where technology converts other-regarding provision into self-regarding consumption at negligible marginal cost. Its load-bearing empirical assumption — α_i > 0 nearly everywhere, with no systematic spite — is the most parsimonious moral psychology consistent with both the evolutionary and experimental record. What it cannot supply is the fixed cost of getting there, or the neutralisation of meaning-based opposition; those remain the abolitionist project's genuinely non-automated work.

If useful, I can develop this into a formal paper section with the equations worked through (perhaps for the second-edition planning alongside the editorial memo already prepared in your project notes No access) — or expand any single strand, such as the effort-asymmetry result, into a standalone piece.

Nov 2, 10:50am

This is a shared conversation and may include unverified content that doesn't reflect Mistral AI's views. Login to resume chatting.

Vibe can make mistakes. Check answers. Learn more