IESET.
Hypotheses·institutional quality·owid_internet_schooling_followthrough_2000_2022

Countries with very large internet-use diffusion after 2000 should usually also show gains in average years of schooling, consistent with broader public-goods and capability expansion.

SUPPORTEDengine/runs/owid_internet_schooling_followthrough_2000_2022

supported

confidence cueThis is a clear pass for the claim as written. It still applies only to this sample, period, and method.

policy briefNeeds review

In ordinary language

In plain terms, this asks whether internet use gain is actually linked to better or worse mean years schooling from 2000 to 2022.

plain answer

The data clearly moved in the predicted direction. supported

why it matters

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

how the test works

It compares 89 country or place units from 2000 to 2022, using a descriptive design.

what was measured
What changed
  • Internet use gain
What we checked
  • Mean years schooling
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/owid_internet_schooling_followthrough_2000_2022
1007550250200020112022ALBARGAUSAUTBELBGDBGR
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show mean_years_schooling across 89 sampled countries over 2000–2022.
The shapes above are stylised — none of the lines are real data.
Placeholder for owid_internet_schooling_followthrough_2000_2022. Published chart will be generated from engine/runs/owid_internet_schooling_followthrough_2000_2022/chart_data.json.

Who has skin in the game — schools predicting on this

1 school list this hypothesis as a test of their position. The chips below are school-level scoreboard outcomes, not a second hypothesis verdict.

hypothesis verdict vs scoreboard outcome

The banner verdict judges this hypothesis as written. The scoreboard asks whether each school's polarity-corrected prediction was right. Raw status is not a school win: SUPPORTED supports schools that needed SUPPORTED, but refutes schools that needed REFUTED.

New Institutional Economicswins if SUPPORTEDschool supported

Pre-registration

pre-registered
first-spec commit 3fc4c7e · 2026-05-02T17:38:10Z

Countries with very large internet-use diffusion after 2000 should usually also show gains in average years of schooling, consistent with broader public-goods and capability expansion.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

SUPPORTED if n>=50, at least 65% of large-internet-gain countries add >=1 average year of schooling, and the median gain is >=1 year. REFUTED if fewer than 45% pass or the median gain is <0.25 years. Otherwise PARTIAL.

formal test & threshold
test:      owid_internet_schooling_followthrough_2000_2022
threshold: n >= 50 AND pass_rate >= 0.65 AND median_schooling_gain_years >= 1

Method

Template
descriptive
Clustering
none
Sample
89 countries · 2000 – 2022
Evidence type
descriptive

Custom endpoint/mean panel replication using local WDI/OWID vintages and fixed country-selection thresholds.

Data

VariableSourceTransform
mean_years_schooling
outcome
owid:mean-years-of-schooling-long-runtier 2
endpoint years gained
internet_use_gain
treatment
owid:share-of-individuals-using-the-internettier 2
country selected if endpoint gain >= 40pp

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card - owid_internet_schooling_followthrough_2000_2022

Verdict: supported - 65 of 89 countries passed (73.0%); median schooling_gain_years = 1.66

Predeclared Threshold

SUPPORTED if n>=50, at least 65% of large-internet-gain countries add >=1 average year of schooling, and the median gain is >=1 year. REFUTED if fewer than 45% pass or the median gain is <0.25 years. Otherwise PARTIAL.

Threshold expression: n >= 50 AND pass_rate >= 0.65 AND median_schooling_gain_years >= 1

Metrics

  • n_countries: 89
  • countries_passing: 65
  • pass_rate: 0.7303370786516854
  • median_schooling_gain_years: 1.6600000000000001

Country Panel

| country_iso3 | country_name | internet_use_gain_pp | schooling_gain_years | pass | |---|---|---|---|---| | ALB | Albania | 82.50 | 0.39 | no | | ARG | Argentina | 81.34 | 0.78 | no | | AUS | Australia | 50.29 | 1.56 | yes | | AUT | Austria | 59.88 | 0.96 | no | | BEL | Belgium | 64.58 | 0.62 | no | | BGD | Bangladesh | 41.55 | 2.46 | yes | | BGR | Bulgaria | 73.76 | 1.10 | yes | | BLZ | Belize | 65.85 | 0.52 | no | | BOL | Bolivia | 65.92 | 0.33 | no | | BRA | Brazil | 77.66 | 1.24 | yes | | BRB | Barbados | 75.47 | 0.44 | no | | CAN | Canada | 42.70 | 1.54 | yes | | CHE | Switzerland | 49.35 | 2.83 | yes | | CHL | Chile | 76.94 | 1.33 | yes | | CHN | China | 73.84 | 1.20 | yes | | CMR | Cameroon | 41.52 | 1.61 | yes | | COL | Colombia | 70.59 | 3.08 | yes | | CRI | Costa Rica | 76.80 | 1.16 | yes | | CUB | Cuba | 67.71 | 1.73 | yes | | CYP | Cyprus | 74.35 | 1.60 | yes | | CZE | Czechia | 74.76 | 0.15 | no | | DEU | Germany | 61.41 | 2.51 | yes | | DNK | Denmark | 58.69 | 0.19 | no | | DOM | Dominican Republic | 77.81 | 2.01 | yes | | DZA | Algeria | 74.34 | 2.12 | yes | | ECU | Ecuador | 68.26 | 1.66 | yes | | EGY | Egypt | 71.56 | 1.71 | yes | | ESP | Spain | 80.86 | 1.86 | yes | | FIN | Finland | 55.74 | 1.15 | yes | | FJI | Fiji | 76.75 | 1.55 | yes | | FRA | France | 71.03 | 1.67 | yes | | GBR | United Kingdom | 69.48 | 2.93 | yes | | GHA | Ghana | 69.21 | 1.72 | yes | | GMB | Gambia | 43.12 | 2.15 | yes | | GRC | Greece | 74.03 | 1.96 | yes | | GTM | Guatemala | 54.70 | 1.91 | yes | | GUY | Guyana | 77.79 | 0.09 | no | | HKG | Hong Kong | 67.79 | 2.46 | yes | | HND | Honduras | 55.19 | 1.64 | yes | | HUN | Hungary | 82.14 | 0.39 | no | | IDN | Indonesia | 65.56 | 3.65 | yes | | IND | India | 55.37 | 2.50 | yes | | IRL | Ireland | 78.77 | 2.71 | yes | | IRN | Iran | 78.13 | 2.14 | yes | | IRQ | Iraq | 78.62 | 2.04 | yes | | ISL | Iceland | 55.34 | 3.36 | yes | | ITA | Italy | 61.95 | 1.78 | yes | | JAM | Jamaica | 79.33 | 0.51 | no | | JOR | Jordan | 87.88 | 2.09 | yes | | JPN | Japan | 54.93 | 1.06 | yes | | KHM | Cambodia | 59.61 | 2.17 | yes | | KOR | South Korea | 52.47 | 1.88 | yes | | KWT | Kuwait | 92.99 | 0.50 | no | | LBY | Libya | 88.45 | 2.53 | yes | | LKA | Sri Lanka | 47.65 | -0.44 | no | | LSO | Lesotho | 46.29 | 0.51 | no | | LUX | Luxembourg | 75.35 | 1.61 | yes | | MAR | Morocco | 89.21 | 2.74 | yes | | MEX | Mexico | 73.55 | 2.22 | yes | | MLT | Malta | 78.43 | 1.66 | yes | | MMR | Myanmar | 58.01 | 2.59 | yes | | MUS | Mauritius | 68.22 | 3.33 | yes | | MYS | Malaysia | 76.01 | 2.47 | yes | | NIC | Nicaragua | 48.15 | 2.06 | yes | | NLD | Netherlands | 48.54 | 0.68 | no | | NOR | Norway | 47.00 | 0.52 | no | | NPL | Nepal | 53.68 | 3.06 | yes | | NZL | New Zealand | 48.81 | -1.17 | no | | PAN | Panama | 71.15 | 1.20 | yes | | PER | Peru | 71.60 | 0.91 | no | | PHL | Philippines | 73.23 | 1.48 | yes | | POL | Poland | 79.66 | 1.27 | yes | | PRT | Portugal | 68.07 | 1.15 | yes | | PRY | Paraguay | 75.51 | 2.28 | yes | | ROU | Romania | 81.89 | 0.58 | no | | RUS | Russia | 88.44 | 0.06 | no | | SEN | Senegal | 58.82 | 2.42 | yes | | SLV | El Salvador | 64.35 | 1.06 | yes | | SRB | Serbia | 60.04 | 2.20 | yes | | SWE | Sweden | 49.32 | 0.21 | no | | SWZ | Eswatini | 56.54 | 1.84 | yes | | THA | Thailand | 84.29 | 3.23 | yes | | TTO | Trinidad and Tobago | 76.77 | 1.56 | yes | | TUN | Tunisia | 67.80 | 2.97 | yes | | TUR | Turkey | 79.68 | 2.12 | yes | | URY | Uruguay | 79.33 | 0.40 | no | | USA | United States | 49.12 | 0.22 | no | | VEN | Venezuela | 58.24 | 2.32 | yes | | ZAF | South Africa | 70.11 | 2.55 | 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/owid_internet_schooling_followthrough_2000_2022.md.

Strongest opposing argument

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

Authored framework. Read the transparency note.