Wednesday, March 4, 2026

Activist movement model and strategy

A thread by eigenrobot had outlined a model of activist movements and their capture by their least sympathetic elements. This post cleans up that model and tries to use the result to answer strategic questions that were raised by Maya. It is stated in abstract terms, and applies variably to immigration, trans issues, etc. As with some of my other posts, the only custom instruction is “User language: English”.

I think Rothbard’s memorandum on activist movement strategy (and other Leninist writings on strategy) is also relevant to the problem in some way, but I did not try to incorporate anything from it.

Eigenrobot’s model, improved

Try to clean up and develop the model described below.

The following is a tentative rational agent explanation for why the public perception of identitarian activist groups tends to be dominated by the least sympathetic members of whatever community they purport to represent. The key intuition here is that it’s an act of self-defense for the high-friction group.

Imagine some excluded group A with measure α < 0.5 and an excluding majority B with measure (1-α). To become “included,” at least 0.5 of the population must agree to inclusion. inclusion imposes a cost C(Ai) borne by B, but produces a benefit Y accruing to included members of A. But A is not homogeneous: some fraction γ∈[0, 1] of A imposes a cost Ch on B when included; fraction (1-γ) imposes a cost Cl on B when included. Here, Ch > Cl.

To interpret these ideas, A is some minority group. If they were accepted by wider society, which currently shuns them, they would benefit; but it would moderately irritate the majority. The amount of irritation increases with the number of A-members accepted and varies by type. Some A-members are low-friction A-members. They’re already well integrated with the majority in some ways and basically chill. However, the high-friction A-members? They’re—say—crazy criminals who sacrifice dogs in their backyards.

Crucially, let’s stipulate that Ch > Y > Cl. That is: the benefit accruing to an included low-friction A-member is greater than the cost imposed on the majority by his inclusion. This means he can in theory bargain with the majority to gain his own inclusion in a Pareto improvement. This isn’t possible for a high-friction A-member because whatever he gets from inclusion is insufficient to compensate the majority for the inconveniences his dog sacrifice ceremonies are causing them.

In this base model, individual low-friction A-members each compensate the majority from some fraction of their inclusion returns, and high-friction A-members are left out in the cold. Obviously this doesn’t explain our phenomenon; we need an story for why the high-friction A-members are in charge. A richer model driving the observed equilibrium might include cost sensitivity heterogeneity in the majority group leading to some kind of coalition politics in the minority, or a structure in the game that requires a representative role for a subset of the minority group.

Regardless of what form a solution actually looks like, the intuition this is groping toward is that the high-cost subpopulation is motivated to dominate activism by the possibility that they’ll be shut out of the integration process and left in the cold by a majority-compatible subgroup.

Narrative variants: the high-cost subgroup has lower opportunity costs to activism (eg, no career or whatever) and pay that cost to control group bargaining position, using it to trade surplus accruing to the low cost subgroup in the base model in for their own inclusion; some kind of strategic lock into a coalition with superior bargaining power—maybe their benefit from inclusion is lower?; possession of some lever for extorting the majority. Regardless, the usual result is transfer of surplus from the low cost group to the high cost group.

Here’s a cleaner “workhorse” model that captures your intuition with a minimal set of added ingredients. I’ll keep your primitives, but I’ll (i) make the policy non-separable, and (ii) explicitly model who becomes visible / representative as an equilibrium selection outcome.

1) Baseline primitives (yours, slightly normalized)

Population mass 1. Minority group $A$ has mass $\alpha<1/2$; majority $B$ has mass $1-\alpha$.

Types in $A$:

  • Low-friction type $l$ has mass $\alpha(1-\gamma)$ and imposes per-capita cost $C_l$ on $B$ if included.
  • High-friction type $h$ has mass $\alpha\gamma$ and imposes per-capita cost $C_h$ on $B$ if included, with $C_h>C_l$.

Each included $A$-member gets benefit $Y$. Your key inequality: $$C_h > Y > C_l.$$ So (in a separable world) including $l$ is potentially Pareto-improvable; including $h$ is not.

2) The missing structural assumption: inclusion is not separable

To get “high-friction dominates the bargaining/branding,” you need a reason the world can’t just include $l$ and exclude $h$.

A crisp way to do that:

Non-separability / imperfect screening: the political/organizational decision is a single binary policy $I \in {0,1}$ (“include the group”), and if $I=1$ then all who can claim membership get included. Formally, the majority cannot condition inclusion on type: $$ I=1 \Rightarrow \text{all } A\text{-types included.} $$

This can represent: legal non-discrimination constraints, unverifiable “type,” administratively costly screening, or a norm that “inclusion must be group-wide.”

Then the majority’s cost of inclusion is $$ \text{Cost to }B = \alpha\left[(1-\gamma)C_l+\gamma C_h\right]. $$ Even if $l$ is “efficient,” the presence of $h$ can make the group’s inclusion unattractive in aggregate, which sets up a real conflict inside $A$: $l$ would like a carve-out; $h$ needs bundling.

3) Who becomes “the face” of the group: a simple collective-action/attention stage

Add a standard Olson-style activism stage with heterogeneous opportunity costs.

Stage 0: outside options

Let $u_t^0$ be the outside option for an $A$-member of type $t\in{l,h}$ if the policy fails ($I=0$). A natural asymmetry: $$ u_l^0 > u_h^0, $$ meaning low-friction members can partially “self-include” via assimilation, private markets, passing, norm-conforming behavior, etc., while high-friction members can’t.

So the incremental value of policy success is $$ \Delta_t := (Y - u_t^0), $$ with $\Delta_h > \Delta_l$.

Stage 1: activism as a public good with type-dependent cost

Each $A$-member chooses activism effort $e_i \ge 0$. Total effort $E=\sum e_i$ increases the probability of inclusion $p(E)$ where $p'>0, p''<0$.

Effort cost depends on type: $$ \text{cost}(e\mid t)=\frac{k_t}{2}e^2, $$ and crucially $k_h < k_l$ (lower opportunity cost / fewer career penalties / fewer foregone alternatives).

Each member solves $$ \max_{e_i\ge 0} ; p(E)\Delta_t ;-;\frac{k_t}{2}e_i^2. $$

Stage 2: visibility / representation

Public perception is shaped disproportionately by the most active/visible members. Model that as:

  • media/attention weight $w_i = \frac{e_i}{E}$ (share of attention proportional to effort), or
  • “the spokesperson” is the highest-effort individual (a tournament).

Either way: higher $e$ ⇒ more likely to be seen as representative.

4) Equilibrium logic: why the high-friction subgroup dominates visibility

In the standard public-good-with-heterogeneous-costs setup:

  • Those with lower marginal costs ($k$ small) and higher marginal benefits ($\Delta$ large) supply more effort.
  • Those with high $k$ (high opportunity cost) and low $\Delta$ free-ride.

Concretely, in any interior equilibrium, a type’s first-order condition is approximately $$ p'(E)\Delta_t \approx k_t e_t. $$ So effort scales like $$ e_t \propto \frac{\Delta_t}{k_t}. $$ Thus $e_h > e_l$ when $$ \frac{\Delta_h}{k_h} > \frac{\Delta_l}{k_l}. $$

Given $\Delta_h>\Delta_l$ and $k_h<k_l$, this inequality is very easy to satisfy. Then:

  1. High-friction members supply most activism effort (they’re the “marginal activists”).
  2. Attention/representation concentrates on them via $w_i$ or the tournament.
  3. Public perception becomes dominated by the least sympathetic members even if they’re a minority within $A$.

This is already enough to generate your phenomenon without assuming “high-friction is in charge” as a primitive. They become in charge because equilibrium activism supply selects them.

5) Bringing your “surplus transfer” intuition into the model

Now add an internal bargaining constraint that’s especially natural under non-separability.

If inclusion is group-wide, then low-friction members may want “moderation” to maximize acceptance probability by $B$. High-friction members may prefer aggressive tactics that (i) increase mobilization and internal solidarity, or (ii) impose costs on $B$ to force bargaining.

A clean way to formalize “transfer from $l$ to $h$” is:

  • There is a group organization that chooses a “line” (demands/tactics) $a$ and funds activism.
  • Low-friction members can pay dues/transfers $T$ to influence $a$ toward moderation.
  • High-friction members have a credible “spoiler” action (they can produce damaging visibility or disruption) unless $a$ and/or $T$ meet their participation constraint.

Participation/spoiler constraint for $h$: $$ \text{If } (a,T) \text{ not acceptable, }h\text{ exerts independent effort }e_h^{\text{spoiler}} \text{ that captures attention.} $$

Then equilibrium often has the structure you described:

  • $l$ would like a “separable bargain” (include $l$, exclude $h$) but can’t get it.
  • $h$ uses (low-cost, high-visibility) activism as leverage to keep the coalition bundled.
  • $l$ pays (resources, rhetorical compliance, reputational cover) to reduce the damage of $h$’s dominance, which is a transfer of surplus from $l$ to $h$.

6) What this model explains (and what it predicts)

Explains

  • Why visibility skews toward “worst” or least sympathetic members: they have the biggest stakes and lowest opportunity costs, so they out-supply activism and capture attention.
  • Why moderates often appear passive: it’s rational free-riding + higher reputational/career costs.
  • Why internal rhetoric can look “hostage-like”: under non-separability, $h$ has a credible veto/spoiler, so coalition policy shifts toward $h$’s constraints, financed by $l$.

Predictions (useful if you want to push toward an empirically testable story)

  1. When outside options for moderates improve ($u_l^0\uparrow$), activism and representation skew more extreme (because $\Delta_l\downarrow$).
  2. When opportunity costs of public activism fall for everyone (e.g., anonymized online activism), you should see more “moderate supply,” potentially reducing the skew—unless anonymity benefits $h$ disproportionately.
  3. When screening becomes feasible (separable inclusion), the extreme subgroup’s leverage collapses and “moderate representation” becomes easier to sustain.

7) One-line version of the refined intuition

If (i) inclusion is effectively group-wide and (ii) activism/visibility is supplied by those with the lowest opportunity costs and highest marginal benefits, then the subgroup that is hardest to integrate rationally becomes the most active and most visible—so public perception of the whole group gets anchored to them, and coalition bargaining shifts surplus away from the easy-to-integrate members.

If you want to take this one notch more “game-theory formal,” the next step is to explicitly model $B$’s decision as a voting game with heterogeneous tolerance (or moral payoffs) and make “tactics” affect both $p(E)$ and perceived expected cost. But the core mechanism above is already doing the main explanatory work with clean comparative statics.

Strategic implications

For a movement that ultimately wishes for inclusion of all A-members, it is better to have low-friction A-members as its public-facing identity, but incentives are against it, as we just modeled. Low-friction A-members also have the incentive to try to distinguish themselves and try to appease the B-members, but this is bad for the high-friction A-members. Think of how political movement strategy should account for this.

In your model, the movement is playing a two-level game:

  • External bargain (with B): you want B to believe “including A won’t be too costly / we can manage the costs.”
  • Internal bargain (within A): you need high-friction members not to veto/sabotage moderation, and low-friction members not to peel off into “separate-but-included” deals.

So “movement strategy” is mostly about engineering incentives and institutions so that (a) low-friction members reliably occupy the front stage (where persuasion matters), while (b) high-friction members still get credible inclusion and real voice (so they don’t rationally seize the mic).

Here are the main levers.

1) Keep “no carve-outs” credible without letting “worst examples” define the brand

Low-friction members distinguishing themselves (“I’m not like those people”) is locally rational but coalition-poisoning: it increases the perceived $\gamma$ and salience of $C_h$, and it tells high-friction members they’re about to be sold out.

A stable line is: “No carve-outs on personhood; lots of carve-outs on behavior.”

  • Don’t publicly split the identity into “good A” vs “bad A.”
  • Do publicly split conduct into “okay in civil society” vs “not okay,” and make that symmetrical (rules apply to everyone, not just A).
  • This lets the movement reassure B that costs are bounded without creating a permanent underclass inside A.

In your terms: you’re trying to reduce perceived $C_h$ (and perceived $\gamma$) without redefining $h$ out of A.

2) Convert “high-friction types” into “high-friction states”

Your model treats $h$ as a type. Movements do better if they treat much of it as a state that can be improved.

Strategically: invest movement resources in lowering $C_h$ rather than only arguing about it.

Examples (abstractly):

  • internal norms and accountability that prevent the most visibly costly behaviors
  • services/mutual aid that reduce dysfunction, instability, or antisocial spillovers
  • training/mentorship that increases members’ capacity to navigate mainstream institutions

If you can make $C_h$ move toward $Y$, you weaken the core “can’t be Pareto-improved” barrier that makes $h$ desperate to control bargaining.

3) Subsidize moderate participation (raise $e_l$ without asking them to be saints)

Your earlier equilibrium is driven by $\Delta_h/k_h > \Delta_l/k_l$: high-friction members have higher stakes and lower opportunity costs, so they out-supply activism and become “the movement.”

If you want low-friction members to front the movement, you must change those ratios:

  • Lower $k_l$: protect moderates from career/reputational costs (anonymity options, legal defense funds, employer shielding, professionalized comms that reduces personal exposure).
  • Raise $\Delta_l$: make movement success matter to them (status, leadership pipelines, selective benefits like networks, professional opportunities, real voice).

This sounds cynical, but it’s just mechanism design: if the movement needs “ambassadors,” it has to pay the opportunity costs of ambassadorship.

4) Separate “mobilization engine” from “public face” via roles and institutions

A common failure mode is pretending the same people can optimally do:

  • base mobilization (high arousal, high conflict)
  • elite persuasion (low arousal, high trust)
  • mass-image management (low variance, high discipline)

You can explicitly adopt a frontstage/backstage structure:

  • Frontstage: disciplined spokespeople; message coherence; high empathy; low variance; optimized for B’s median.
  • Backstage: broader internal deliberation; higher tolerance for anger/heterodoxy; optimized for keeping $h$ included and heard.

This only works if backstage has real power (agenda-setting, vetoes, budget shares), otherwise $h$ correctly treats it as a trick and fights for frontstage control.

5) Make internal representation robust to “activist capture”

If attention is proportional to effort, “whoever can shout all day becomes the movement.” Counter that with governance that weights membership more than activism hours:

  • leadership selection by election with high-turnout mechanisms, not whoever shows up to meetings
  • rotation/term limits for spokesperson roles
  • transparent decision rights: who sets demands, who chooses tactics, who speaks externally
  • formal “communication discipline” rules: not censoring beliefs, but controlling who represents the org publicly

The key is to decouple “time available” from “formal authority.”

6) Bundle demands in a sequence that avoids internal betrayal

Movements that want inclusion of all A-members usually need staging—but staging can look like betrayal to $h$.

A workable framing is:

  • Sequence policies, not people.
  • Start with policies that expand inclusion broadly while imposing minimal perceived costs on B.
  • Pair every early win with movement-visible investments that benefit $h$ (so $h$ sees progress, not abandonment).

So the coalition story becomes: “We win step 1 for everyone, then step 2 for everyone,” not “we win for the good ones first.”

7) Treat “appeasement” as a scarce resource; use it where it buys real votes

Low-friction members trying to appease B can backfire if it:

  • legitimizes B’s worst stereotypes
  • teaches B that outrage yields concessions
  • signals to $h$ that the coalition is conditional

So appeasement should be replaced with targeted reassurance:

  • acknowledge concrete concerns B has
  • offer universal safeguards (behavioral standards, enforcement symmetry, institutions that limit externalities)
  • keep moral posture consistent: “you don’t get to exclude people, but we do take costs seriously.”

A compact strategic slogan (model-faithful)

Keep identity inclusive, keep behavior governable, pay for moderation, and institutionalize who speaks.

If you want, I can sketch a simple formal extension where the movement chooses (i) a spokesperson-selection rule, (ii) an internal transfer $T$ to subsidize $e_l$, and (iii) a “cost-reduction investment” that lowers $C_h$, and then show the comparative statics for when “low-friction face, high-friction included” is an equilibrium.

at No comments:
Labels:
Subscribe to: Comments (Atom)