Pre-registration
Countries with large tertiary-attainment gains from 2000 to 2023 should usually register sizable growth in output per worker.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
SUPPORTED if n>=15, at least 70% of large-tertiary-gain countries increase output per worker by >=20%, and the median gain is >=25%. REFUTED if fewer than 50% pass or the median gain is <5%. Otherwise PARTIAL.
formal test & threshold
test: wdi_tertiary_attainment_labor_productivity_2000_2023 threshold: n >= 15 AND pass_rate >= 0.70 AND median_labor_productivity_growth_pct >= 25
Method
- Template
descriptive- Clustering
none- Sample
- 21 countries · 2000 – 2023
- Evidence type
- descriptive
Custom endpoint/mean panel replication using local WDI/OWID vintages and fixed country-selection thresholds.
Data
| Variable | Source | Transform |
|---|---|---|
output_per_worker outcome | world_bank_wdi:SL.GDP.PCAP.EM.KDtier 2 | endpoint percent growth |
tertiary_attainment_gain treatment | world_bank_wdi:SE.TER.CUAT.BA.ZStier 2 | country selected if endpoint gain >= 10pp |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Result card - wdi_tertiary_attainment_labor_productivity_2000_2023
Verdict: supported - 17 of 21 countries passed (81.0%); median labor_productivity_growth_pct = 43.78
Predeclared Threshold
SUPPORTED if n>=15, at least 70% of large-tertiary-gain countries increase output per worker by >=20%, and the median gain is >=25%. REFUTED if fewer than 50% pass or the median gain is <5%. Otherwise PARTIAL.
Threshold expression: n >= 15 AND pass_rate >= 0.70 AND median_labor_productivity_growth_pct >= 25
Metrics
- n_countries: 21
- countries_passing: 17
- pass_rate: 0.8095238095238095
- median_labor_productivity_growth_pct: 43.77821269416413
Country Panel
| country_iso3 | country_name | tertiary_attainment_gain_pp | labor_productivity_growth_pct | pass | |---|---|---|---|---| | ALB | Albania | 10.08 | 129.13 | yes | | AUS | Australia | 15.86 | 19.49 | no | | BOL | Bolivia | 12.10 | 23.83 | yes | | CAN | Canada | 15.14 | 11.55 | no | | CRI | Costa Rica | 11.16 | 69.71 | yes | | CZE | Czechia | 21.11 | 49.89 | yes | | IRL | Ireland | 11.09 | 106.08 | yes | | IRN | Iran, Islamic Rep. | 11.83 | 28.80 | yes | | KOR | Korea, Rep. | 13.65 | 69.24 | yes | | LTU | Lithuania | 14.92 | 137.73 | yes | | MKD | North Macedonia | 11.40 | 43.78 | yes | | MLT | Malta | 13.22 | 34.74 | yes | | MNG | Mongolia | 18.93 | 160.21 | yes | | MUS | Mauritius | 10.99 | 66.27 | yes | | PRT | Portugal | 16.22 | 22.12 | yes | | PSE | West Bank and Gaza | 14.53 | 15.41 | no | | SAU | Saudi Arabia | 15.96 | -30.16 | no | | SGP | Singapore | 18.67 | 64.02 | yes | | SRB | Serbia | 10.80 | 106.78 | yes | | SWE | Sweden | 12.84 | 24.97 | yes | | USA | United States | 12.73 | 36.84 | yes |
Interpretation
This is a descriptive structural-screen verdict using local WDI/OWID vintages. It grades the predeclared pattern, not a causal effect of a single policy lever.
Steelman
See hypotheses/steelman/wdi_tertiary_attainment_labor_productivity_2000_2023.md.
Strongest opposing argument
Every hypothesis ships with its charitable opposing argument. The framework earns credibility by handling objections at their strongest, not weakest.