---
title: "The Selection Question"
subtitle: "What does this policy make easier to reproduce?"
author: Elias Kunnas
description: "Every durable policy changes what can reproduce itself. The Selection Question asks what a mechanism makes easier to continue, harder to continue, and more likely to recur—across four evidentiary channels."
canonical: https://kunnas.com/articles/selection-question
url: https://kunnas.com/articles/selection-question.md
date_published: 2026-01-09
date_modified: 2026-08-31
corpus_frame_url: https://kunnas.com/articles/how-to-read-this.md
---
## How to read this corpus

The corpus applies one lens to many domains: what mechanisms produce the outcome? It shares four methodological commitments and one explicit directional commitment. Each linked page argues for its part; the links are derivations and disputes, not evidence inherited by every page. The directional commitment does not by itself settle system boundary, distribution, sacrifice, or institutional authority.

1. **Mechanisms are what act.** Incentive gradients, selection pressures, feedback loops, and capital stocks produce the distribution of outcomes. Intentions, labels, official categories, and stated values are evidence about mechanisms, or are themselves coordination mechanisms. They are not causal substitutes. — [Mechanism Realism](https://kunnas.com/articles/mechanism-realism.md) · [Only Selection](https://kunnas.com/articles/only-selection.md)
2. **The reference telos is sustained flourishing.** The broadest achievable adaptive safety margin over deep time — not the continuity of any incumbent state, coalition, institution, or doctrine. A mechanism's own stated goal can still serve as a local proof obligation — showing that its incentives defeat even the purpose it claims is a bounded finding — but meeting that goal establishes nothing about the margin. — [Flourishing Is Maximum Safety Margin](https://kunnas.com/articles/flourishing-is-maximum-safety-margin.md)
3. **Law, rights, legitimacy, democracy, markets, and sovereignty are mechanisms under evaluation.** They are constraints, carriers, or proxies inside the analysis. None is a terminal value or a boundary of what is real. Treating one as terminal ends the mechanism search before it starts. Evaluation carries current function, replacement cost, path dependence, uncertainty, capture risk, reversibility, and who bears model error into the ledger. — [The Stack](https://kunnas.com/articles/the-stack.md) · [Mechanism Space](https://kunnas.com/articles/mechanism-space.md)
4. **Optimization is a system function.** A civilization has to build, exercise, and revise metamechanisms that search mechanism-space, discard dominated options, install, observe effects, and repair under uncertainty. Not running that loop leaves margin unrealized, and that is itself the failure. No single component — analyst, model, or institution — is presumed to contain a global optimum; the capacity is a property of the system. — [From Telos to Policy](https://kunnas.com/articles/from-telos-to-policy.md) · [The Three-Layer Architecture](https://kunnas.com/articles/three-layer-architecture.md)
5. **Uncertainty is preserved, not spent.** Partial orders, binding constraints, unknowns, and residuals stay explicit. An unmeasured effect is not a favorable default. — [The Compression Paradox](https://kunnas.com/articles/compression-paradox.md) · [Cargo Cult Epistemology](https://kunnas.com/articles/cargo-cult-epistemology.md)

*Each essay bears its own evidence. Links carry definitions, derivations, applications, and disputes; they do not transfer proof. Criticism is answered on its substance.*

Canonical: <https://kunnas.com/articles/how-to-read-this.md>

---

# The Selection Question

*What does this policy make easier to reproduce?*

Elias Kunnas

## Thesis {#thesis}

Every durable policy changes what can reproduce itself: behaviors, household forms, skills, institutions, and sometimes demographic lineages. The Selection Question is not “which people should reproduce?” It is: what does this mechanism make easier to continue, harder to continue, and more likely to recur?

## Standard objections addressed in this essay

- “Heritability does not predict a response to a specific policy.” — [§III](#iii-biological-differential-reproduction) (Population estimates do not identify intervention effects.)
- “Policy effects may be small beside environment and culture.” — [§III](#iii-biological-differential-reproduction), [§IV](#iv-developmental-effects), [§V](#v-cultural-transmission), [§VI](#vi-institutional-selection)(Each channel has its own magnitude question.)
- “This is eugenics through incentives rather than force.” — [§VII](#vii-authority-boundary) (Equal standing and anti-coercion constrain every proposal.)
- “Selective exit from welfare is not genetic selection.” — [§VI](#vi-institutional-selection) (Exit belongs in institutional selection; see [The Capability Trap](capability-trap.md).)
- “Capability is multidimensional and politically contestable.” — [§VIII](#viii-the-operational-test) (No policy target reduces people to one profile.)

---

## I. The Diagnostic {#i-the-diagnostic}

Standard policy analysis asks who benefits now, what it costs, and whether the mechanism works as intended. Long-run analysis adds a different question: **what becomes more common because this policy exists?**

Reproduction here is not limited to genes. A practice, skill, household form, institution, or behavior pattern “reproduces” when it persists, spreads, or recurs across people and time. Tax design, housing rules, childcare subsidies, benefit formulas, school curricula, and migration law all change the costs and payoffs of different ways of living. Those changes alter what continues.

The Selection Question is the diagnostic that makes those effects inspectable. It does not ask which people ought to have children. It asks what a mechanism makes **easier to continue**, **harder to continue**, and **more likely to recur**.

## II. Four Channels {#ii-four-channels}

Selection effects arrive through distinct pathways. Treating them as one undifferentiated “selection pressure” is how genetic claims smuggle in on cultural evidence, or institutional sorting gets mistaken for biological change.

Separate the channels. Keep the evidence in its channel.

| Channel | What varies | What “reproduces” | Evidence required |
|----|----|----|----|
| **1. Biological differential reproduction** | Allele or lineage frequency | Genotypes represented in the next generation | Trait-specific differential fertility or mortality; transmission model; attribution to the policy |
| **2. Developmental effects** | Skills, health, stress physiology, epigenetic marks | Capabilities and constraints in the same life and across generations | Exposure pathway; developmental window; measured outcome; persistence horizon |
| **3. Cultural transmission** | Norms, practices, narratives, identities | Behaviors and beliefs copied between people | Transmission mechanism; adoption and retention rates; competing carriers |
| **4. Institutional selection** | Rules, organizations, eligibility categories, career paths | Institutions and the behaviors they reward or filter | Incentive structure; sorting and exit; administrative burden; fiscal and political feedback |

A single policy can touch all four channels. That does not make them interchangeable. A change in who qualifies for a benefit is institutional selection until direct reproduction and transmission evidence place part of the effect in the biological channel.

## III. Biological Differential Reproduction {#iii-biological-differential-reproduction}

Biological selection is change in who leaves descendants and what they transmit. If a trait correlates with higher fertility, lower child mortality, or earlier reproduction, its frequency can shift. If it correlates with the opposite, it can decline. This is arithmetic, not ideology.

Policy can alter those correlations. Child benefits, housing costs, partner-market structure, health access, and time pressure all change the feasible fertility set. Whether a given reform changes allele frequencies is a separate, harder claim.

**Heritability is channel-specific evidence, not a policy verdict.** Twin and adoption studies estimate how much variation in a trait within a population is associated with genetic differences. They do not, by themselves, predict how a policy will move that trait, let alone its frequency in the gene pool.

To claim biological selection from a policy, name:

- the **trait** (for example time preference, executive function, or a measured behavioral phenotype);
- the **population** and time window;
- the **pathway** through which the policy changes differential reproduction or survival;
- the **transmission model** (direct inheritance, assortative mating, mortality, migration);
- and the **counterfactual** that isolates the policy from confounds.

Illustrative magnitudes, not policy conclusions: executive-function differences show very high heritability in some twin designs; cognitive ability heritability tends to rise with age as people select environments; savings behavior shows measurable genetic association in population samples. None of these establish that a particular welfare rule, tax credit, or school reform will shift the next generation’s capability distribution through the biological channel. That requires the full chain above.

Future gene-editing tools change the biological channel directly. They do not dissolve the other three. Development, culture, and institutions still shape what edited or unedited humans become.

## IV. Developmental Effects {#iv-developmental-effects}

Developmental effects change what a person can do without requiring a change in allele frequency. Nutrition, stress, infection, sleep, parenting quality, classroom practice, and neighborhood safety all shape brain development, health, and skill formation. Some effects fade. Some persist for a life. Some show up in the next generation through biology, parenting behavior, or both.

This channel owns the developmental evidence. Do not treat a measured skill gap as proof of genetic selection without a transmission argument. Do not treat a heritability estimate as proof that environment does not matter.

Policy enters through exposure: what conditions children and parents actually face once a rule is in force. Prenatal famine cohorts show persistent epigenetic marks decades later. Severe early adversity can alter stress physiology. “Genetic nurture” means parental genotypes shape child environments even when particular alleles are not inherited—an interaction that belongs here and in [§III](#iii-biological-differential-reproduction), with the pathway made explicit.

Heckman-style skill formation is the constructive version of the same channel: early investment can raise later capability when the investment hits a plastic window and is not undone by later institutions. Whether a given public program does that is empirical. The channel is real either way.

## V. Cultural Transmission {#v-cultural-transmission}

Cultures reproduce through copying. Parents, peers, schools, media, employers, and subcultures transmit norms about work, thrift, trust, family formation, risk, and status. A policy changes cultural transmission when it alters what is admired, shamed, rewarded, or imaginable.

Examples belong in this channel: a benefit formula that makes single-earner households non-viable; a credential rule that elevates one kind of excellence; a housing policy that makes multi-generational living rare; a rhetoric of “deserving” and “undeserving” poor that children absorb. The mechanism is social learning and status, not allele frequency.

Cultural selection can move faster than biological selection and reverse faster too. It can also lock in: norms that made sense under one economic regime persist after the regime changes, because institutions and identities lag.

When an analyst says a policy “rewards dependency” or “rewards independence,” they are usually making a cultural-transmission claim, sometimes joined to an institutional claim. Keep it here until reproduction evidence says otherwise.

## VI. Institutional Selection {#vi-institutional-selection}

Institutions select by sorting people into categories and attaching consequences to category membership. Eligibility rules, marginal tax rates, asset tests, licensing, zoning, accreditation, hiring pipelines, and enforcement intensity decide which behaviors are viable inside the law.

This channel owns incentive traps, administrative burden, and selective exit. A system that requires asset liquidation before help, imposes high effective marginal tax rates on earnings, or makes incapacity legible to qualify creates institutional selection toward particular household strategies. That is a design question about incentives, dignity, and exit routes—not, by itself, a genetic claim.

The Finnish income-support mechanisms—asset treatment, earnings withdrawal, surveillance, and trajectory dependence—are analyzed in [The Capability Trap](capability-trap.md). That essay owns the institutional channel for that case: how the formula consumes runway and shapes behavior. This essay states the separation rule: selective exit from a benefit system belongs here unless trait-specific differential reproduction and transmission have been demonstrated elsewhere.

Institutional selection also applies to organizations. Agencies, firms, parties, and professions reproduce when budgets, careers, and status flows favor their continued existence. A policy that feeds one organizational form and starves another is selecting at the institutional layer even when individual genetics are untouched.

## VII. Authority Boundary {#vii-authority-boundary}

The Selection Question is politically dangerous because historical eugenics answered a different question with state force. The boundary is compact:

- **Legitimate:** inspect what a mechanism makes easier to continue; measure effects by channel; revise design under equal standing, anti-coercion, uncertainty, and reversibility.
- **Illegitimate:** rank human beings by reproductive worth; use state power to control who may have children; treat a uncertain model as a mandate to override reproductive autonomy.

Blank-slate denial is the mirror error. Refusing to examine recurrence does not make recurrence stop. It makes mechanisms unmeasured.

Proposals that “build capability” through education, health, and opportunity belong on the table. Proposals that specify an optimal genotype do not. The first works through development, culture, and institutions. The second repeats the central-planning mistake in biology.

## VIII. The Operational Test {#viii-the-operational-test}

For any durable policy, run four parallel checks:

1.  **Biological:** Does it change differential reproduction or survival for a named trait and population, with a transmission model? If not, stop inferring genetic direction.
2.  **Developmental:** What exposures change, in which windows, with what measured persistence?
3.  **Cultural:** What norms, identities, and practices become easier to copy?
4.  **Institutional:** What behaviors and organizations does the rule reward, filter, or extinguish?

Then ask the cross-channel questions: magnitude, horizon, distribution, reversibility, and who bears model error. Present wellbeing, rights, fiscal sustainability, and democratic legitimacy still matter. No channel settles the others.

Capability is multidimensional. No serious analysis reduces people to a single score. The operational test is not a ranking of human worth. It is a discipline for seeing what a mechanism compounds.

For designed systems—training stacks, evaluation harnesses, markets with explicit fitness—see [Engineered Selection](engineered-selection.md): given an intended effect, make causal contribution to it determine what recurs. The present essay is the policy-side diagnostic; that essay is the construction-side response.

## IX. Conclusion {#ix-conclusion}

Every durable policy changes what can reproduce itself. The Selection Question makes that fact inspectable without smuggling biology into culture or culture into genetics.

Separate the four channels. State the evidence each channel requires. Hold equal standing, anti-coercion, uncertainty, and reversibility as constraints—not as excuses to avoid the question.

What a mechanism makes easier to continue is not a side effect of governance. It is part of what governance is.

---

## Synthesis {#synthesis}

**The argument in four sentences:** Durable policy changes what behaviors, forms, skills, institutions, and sometimes lineages recur. Biological, developmental, cultural, and institutional channels require separate evidence. Heritability and benefit design do not, by themselves, establish genetic selection from a specific reform. The legitimate task is to inspect recurrence under equal standing, anti-coercion, and reversibility—not to rank who should reproduce.

*Capability & Selection series: [Diagnostic](capability-trap.md) → [Prescriptive](capability-architecture.md) → **Selection** → [Institutional](fourth-branch.md)*

**Related:**

- [Engineered Selection](engineered-selection.md) — Given the intended effect, make causal contribution to it determine what recurs
- [The Capability Trap](capability-trap.md) — Institutional selection in Finnish income-support design
- [Only Selection](only-selection.md) — Variants, filters, and differential persistence across domains
- [Fertility Requires Margin](fertility-requires-margin.md) — Why household margin shapes demographic recurrence
- [The Axiological Malthusian Trap](axiological-malthusian-trap.md) — When abundance suppresses the feedback that would force repair

## Sources and Notes

**Heritability (biological channel only):** Executive function (Common EF factor): high heritability in Friedman et al. ([2008](https://pmc.ncbi.nlm.nih.gov/articles/PMC2762790/)). Cognitive ability: the “Wilson effect”—heritability tends to rise with age (Bouchard, 2004).

Savings behavior: genetic association in population samples ([Cronqvist & Siegel, 2015](https://www.journals.uchicago.edu/doi/10.1086/679284)). These are trait-in-population estimates, not predictions about a named policy’s genetic effect.

**Developmental and intergenerational transmission:** Genetic nurture (Kong et al., *Science*, 2018). Dutch Hunger Winter epigenetic persistence (Heijmans et al., *PNAS*, 2008). Holocaust-survivor offspring FKBP5 methylation (Yehuda et al., 2016). Scarcity effects on parental bandwidth (Mullainathan & Shafir, *Scarcity*, 2013).

**Institutional channel (Finland):** Basic Income Experiment reports (Kangas et al., 2019) measure wellbeing and employment within the experiment design; they do not identify genetic or multigenerational selection. Income-support formulas and effective marginal rates are documented in Kela guidelines and OECD reviews. Trajectory analysis: THL / Blomgren et al. on social-assistance dynamics.