IESET.
Hypotheses·regulatory·chicago_school_voucher_choice_test_score_gain_meta

Across the empirical universe of school-voucher and means-tested school-choice programmes that have been evaluated with random-assignment or quasi-random-assignment designs (USA Milwaukee, Cleveland, DC OSP, New York School Choice Scholarships, Louisiana, Indiana; Chile post-1980 voucher; Sweden post-1992 friskolor; Colombia PACES), the population- weighted mean effect on standardised reading and mathematics test scores is positive at +0.05 to +0.15 standard deviations after 2-4 years of programme exposure, with effects largest for low-income participants drawn from the worst-performing public schools.

The hypothesis is Friedman's (1962 Capitalism and Freedom) prediction that vouchers permit competitive sorting that raises measured human capital outcomes for the marginal participant.

INCONCLUSIVEengine/runs/chicago_school_voucher_choice_test_score_gain_meta

INCONCLUSIVE_DATA_PENDING — no outcome variable loaded; missing: ['academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct']

confidence cueResult card produced; verdict unclassified.

policy briefCoverage too thin

In ordinary language

Over a long period, do more market-oriented institutions translate into higher income or productivity, once the comparison looks beyond a single success story?

plain answer

This test cannot make a firm call yet. no outcome variable loaded; missing: ['academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct']

why it matters

This matters because regulatory claims should change belief only when they survive a pre-declared empirical test.

how the test works

It compares 5 country or place units from 1990 to 2024, using a panel fe decomposition design, with fixed effects for study id.

what was measured
What changed
  • Voucher or choice lottery assignment
What we checked
  • Standardised reading score effect size
  • Standardised maths score effect size
  • High school graduation rate
what this does not prove

A single test is not the whole truth. It narrows the claim under a specific sample, time period, and method. Strong policy conclusions need the pattern to survive nearby tests, alternative data, and serious objections.

verification

No evidence packet has been generated yet.

Results

engine/runs/chicago_school_voucher_choice_test_score_gain_meta
1007550250199020072024USACHLSWECOLIND
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show standardised_reading_score_effect_size across 5 sampled countries over 1990–2024.
The shapes above are stylised — none of the lines are real data.
Placeholder for chicago_school_voucher_choice_test_score_gain_meta. Published chart will be generated from engine/runs/chicago_school_voucher_choice_test_score_gain_meta/chart_data.json.

Pre-registration

pre-registered
first-spec commit 098ce96 · 2026-04-30T12:57:33Z
run generated · 2026-06-29T17:54:28Z

Across the empirical universe of school-voucher and means-tested school-choice programmes that have been evaluated with random-assignment or quasi-random-assignment designs (USA Milwaukee, Cleveland, DC OSP, New York School Choice Scholarships, Louisiana, Indiana; Chile post-1980 voucher; Sweden post-1992 friskolor; Colombia PACES), the population- weighted mean effect on standardised reading and mathematics test scores is positive at +0.05 to +0.15 standard deviations after 2-4 years of programme exposure, with effects largest for low-income participants drawn from the worst-performing public schools. The hypothesis is Friedman's (1962 Capitalism and Freedom) prediction that vouchers permit competitive sorting that raises measured human capital outcomes for the marginal participant.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

Not supported if the random-effects pooled effect on EITHER reading OR maths is below +0.05 SD with confidence interval including zero, OR if the moderator-analysis does not show larger effects in bottom-quartile-public-school counterfactual contexts, OR if the Chilean post-1980 panel (long-run, large-N case) shows pooled null or negative results. A heterodox / Hirschman exit-vs-voice reading wins cleanly if pooled reading effect < 0 OR pooled maths effect < 0 at p<0.05. Publication-bias correction (trim-and-fill, PET-PEESE) must leave the pooled estimate above +0.03 SD; if it does not, the result is treated as publication-driven and insufficient.

formal test & threshold
test:      random_effects_meta_voucher_test_score_with_publication_correction
threshold: pooled_reading_effect >= 0.05 SD with CI excluding zero AND pooled_maths_effect >= 0.05 SD with CI excluding zero AND PET-PEESE bias-corrected pooled effect >= 0.03 SD AND moderator effect: bottom-quartile public school context > top-quartile by >= 0.05 SD

Method

Template
panel_fe_decomposition
Fixed effects
study_id
Clustering
study_id
Sample
5 countries · 1990 – 2024
Evidence type
causal

Random-effects meta-analysis pooling RCT and RD-design effect estimates with study-level random intercepts and inverse-variance weights. Forest plot reported per outcome (reading, maths, graduation). Moderator analysis: counterfactual public-school quality, programme age, country-level institutional quality. PRISMA-style search protocol pre-registered: search the academic-publisher corpus (Angrist, Hoxby, Chetty, Hanushek, Wolf) for all RCT / lottery / RD designs published 1995-2024.

Data

VariableSourceTransform
standardised_reading_score_effect_size
outcome
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
hedges_g_random_effects
standardised_maths_score_effect_size
outcome
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
hedges_g_random_effects
high_school_graduation_rate
outcome
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
marginal_effect_pp
voucher_or_choice_lottery_assignment
treatment
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
indicator_treatment_assignment
programme_funding_per_pupil
control
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
log_usd_2020
counterfactual_public_school_quality
control
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
standardised_test_score_baseline
programme_age_years
control
academic:hoxby_chetty_angrist_meta_voucher_rcttier 4
integer_years
country_gdp_per_capita
control
world_bank_wdi:NY.GDP.PCAP.PP.KDtier 2
log

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — chicago_school_voucher_choice_test_score_gain_meta

Verdict: INCONCLUSIVE_DATA_PENDING — no outcome variable loaded; missing: ['academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct']

Pre-registration

  • Claim: Across the empirical universe of school-voucher and means-tested school-choice programmes that have been evaluated with random-assignment or quasi-random-assignment designs (USA Milwaukee, Cleveland, DC OSP, New York School Choice Scholarships, Louisiana, Indiana; Chile post-1980 voucher; Sweden post-1992 friskolor; Colombia PACES), the population- weighted mean effect on standardised reading and mathematics test scores is positive at +0.05 to +0.15 standard deviations after 2-4 years of programme exposure, with effects largest for low-income participants drawn from the worst-performing public schools. The hypothesis is Friedman's (1962 Capitalism and Freedom) prediction that vouchers permit competitive sorting that raises measured human capital outcomes for the marginal participant.
  • Falsification rule: Not supported if the random-effects pooled effect on EITHER reading OR maths is below +0.05 SD with confidence interval including zero, OR if the moderator-analysis does not show larger effects in bottom-quartile-public-school counterfactual contexts, OR if the Chilean post-1980 panel (long-run, large-N case) shows pooled null or negative results. A heterodox / Hirschman exit-vs-voice reading wins cleanly if pooled reading effect < 0 OR pooled maths effect < 0 at p<0.05. Publication-bias correction (trim-and-fill, PET-PEESE) must leave the pooled estimate above +0.03 SD; if it does not, the result is treated as publication-driven and insufficient.
  • Falsification test: random_effects_meta_voucher_test_score_with_publication_correction

Estimate

  • Error: no outcome variable loaded; missing: ['academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct', 'academic:hoxby_chetty_angrist_meta_voucher_rct']

Variables resolved

  • world_bank_wdi:NY.GDP.PCAP.PP.KD → country_gdp_per_capita (controls, publisher=world_bank_wdi, n=8325)

Variables missing data

  • academic:hoxby_chetty_angrist_meta_voucher_rct (outcome, name=standardised_reading_score_effect_size) — vintage not on disk
  • academic:hoxby_chetty_angrist_meta_voucher_rct (outcome, name=standardised_maths_score_effect_size) — vintage not on disk
  • academic:hoxby_chetty_angrist_meta_voucher_rct (outcome, name=high_school_graduation_rate) — vintage not on disk
  • academic:hoxby_chetty_angrist_meta_voucher_rct (treatment, name=voucher_or_choice_lottery_assignment) — vintage not on disk
  • academic:hoxby_chetty_angrist_meta_voucher_rct (controls, name=programme_funding_per_pupil) — vintage not on disk
  • academic:hoxby_chetty_angrist_meta_voucher_rct (controls, name=counterfactual_public_school_quality) — vintage not on disk
  • academic:hoxby_chetty_angrist_meta_voucher_rct (controls, name=programme_age_years) — vintage not on disk

Generated by scripts/run_panel_fe.py at 2026-06-29T17:54:28+00:00

Strongest opposing argument

Every hypothesis ships with its charitable opposing argument. The framework earns credibility by handling objections at their strongest, not weakest.

Notes

Hoxby (2003), Angrist-Bettinger-Kremer (2006 PACES), Wolf et al. (2010 DC OSP), Abdulkadiroglu-Pathak-Walters (2018 Boston charters), Chingos-Peterson (2015 NY SCSF) form the core US RCT corpus. Hsieh-Urquiola (2006) is the canonical sceptical Chilean voucher study and is included as a falsifying-direction reference. Specifying publication-bias correction in advance is the structural defence against author's pro-voucher prior.

Authored framework. Read the transparency note.