IESET.
Hypotheses·labour·working_time_reduction_employment_neutral_oecd

Across OECD economies 1980-2022, country-year reductions in average annual hours worked per employed person are NOT proportionally associated with declines in the employment-to-population ratio.

The implied employment elasticity with respect to hours is small (|epsilon| < 0.5), refuting the lump-of-labour-fallacy critique that shorter-hours policy must shed jobs proportionally.

REFUTEDengine/runs/working_time_reduction_employment_neutral_oecd

REFUTED — coef=+0.3272 (sign opposite claim -), p=0.0667

confidence cueThis test cuts against the claim as written or misses its pre-declared threshold.

policy briefNeeds review

In ordinary language

In plain terms, this asks whether log avg annual hours worked is actually linked to better or worse log employment to population ratio from 1980 to 2022.

plain answer

The data did not support the prediction. coef=+0.3272 (sign opposite claim -), p=0.0667

why it matters

Labor-market rules often help some workers while risking job loss or slower hiring for others. This test looks for that tradeoff in observable employment or unemployment data.

how the test works

It compares 23 country or place units from 1980 to 2022, using a panel fe design, with fixed effects for country and year.

what was measured
What changed
  • Log avg annual hours worked
  • French 35 hour indicator
What we checked
  • Log employment to population ratio
  • Log unemployment rate
  • Log real income per employed
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/working_time_reduction_employment_neutral_oecd
1007550250198020012022FRADEUNLDBELAUTSWENOR
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show log_employment_to_population_ratio across 23 sampled countries over 1980–2022.
The shapes above are stylised — none of the lines are real data.
Placeholder for working_time_reduction_employment_neutral_oecd. Published chart will be generated from engine/runs/working_time_reduction_employment_neutral_oecd/chart_data.json.

Pre-registration

registration ordering unverified
first-spec commit 4c8ce8e · 2026-07-18T22:11:21Z
run generated · 2026-06-29T17:54:05Z
Run timestamp predates this path's first git-add commit (rebase, rename, or pre-git local run). Spec hash is still the path's first-add commit — not repository HEAD — but ordering is not a clean pre-registration proof.

Across OECD economies 1980-2022, country-year reductions in average annual hours worked per employed person are NOT proportionally associated with declines in the employment-to-population ratio. The implied employment elasticity with respect to hours is small (|epsilon| < 0.5), refuting the lump-of-labour-fallacy critique that shorter-hours policy must shed jobs proportionally.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

SUPPORTED iff the estimated within-country elasticity of log(emp/pop) on log(hours) is in (-0.5, +0.5) at p<0.10 in the primary specification AND the FRA event-study around 2000 shows no negative deviation in employment trajectory worse than -2% at 5-year horizon. REFUTED if the elasticity is < -0.7 (close to lump-of-labour prediction of -1) at p<0.10 OR if the FRA event-study shows a negative employment deviation > 5% at 5-year horizon. PARTIAL between these. METHOD_VALID requires PWT avh + ILOSTAT employment-to-population on disk for >= 18 of 23 panel countries 1990-2022.

formal test & threshold
test:      oecd_panel_employment_elasticity_to_hours_1980_2022
threshold: PRIMARY: |beta_log_hours_on_log_emp_pop_ratio| < 0.5 at p<0.10 AND FRA 35-hour event-study employment deviation worse than -2% at 5y NOT realised.

Method

Template
panel_fe
Fixed effects
country, year
Clustering
country
Sample
23 countries · 1980 – 2022
Evidence type
associational

Two-way FE panel regression of log(employment_to_population_ratio) on log(avg_annual_hours) with country and year fixed effects, clustered on country. The estimated coefficient is the within-country elasticity of employment to hours; the lump-of-labour prediction is coefficient = -1 (each hour cut translates to a proportional employment decline only if the wage bill is fully redistributed across new hires). The hypothesis predicts |coefficient| < 0.5. Sensitivity specifications: (i) drop GFC + COVID windows, (ii) add OECD lagged GDP-growth as control, (iii) FRA-only event-study around 2000 35-hour week, (iv) Iceland 2015-2019 4-day-week trial.

Data

VariableSourceTransform
log_employment_to_population_ratio
outcome
ilostat:employment_to_population_ratiotier 2
log
log_unemployment_rate
outcome
world_bank_wdi:SL.UEM.TOTL.ZStier 2
log
log_real_gdp_per_employed
outcome
pwt:rgdpotier 3
pwt:emptier 3
log_ratio
log_avg_annual_hours_worked
treatment
pwt:avhtier 3
log
french_35_hour_indicator
treatment
constructed:1 for FRA 2000+ (Aubry laws). Used in the named-policy event-study spec.tier 5
indicator
log_real_gdp_per_capita
control
world_bank_wdi:NY.GDP.PCAP.KDtier 2
log
log_population
control
world_bank_wdi:SP.POP.TOTLtier 2
log
trade_openness
control
world_bank_wdi:NE.TRD.GNFS.ZStier 2
level
services_share_gdp
control
world_bank_wdi:NV.SRV.TOTL.ZStier 2
level

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — working_time_reduction_employment_neutral_oecd

Verdict: REFUTED — coef=+0.3272 (sign opposite claim -), p=0.0667

Pre-registration

  • Claim: Across OECD economies 1980-2022, country-year reductions in average annual hours worked per employed person are NOT proportionally associated with declines in the employment-to-population ratio. The implied employment elasticity with respect to hours is small (|epsilon| < 0.5), refuting the lump-of-labour-fallacy critique that shorter-hours policy must shed jobs proportionally.
  • Falsification rule: SUPPORTED iff the estimated within-country elasticity of log(emp/pop) on log(hours) is in (-0.5, +0.5) at p<0.10 in the primary specification AND the FRA event-study around 2000 shows no negative deviation in employment trajectory worse than -2% at 5-year horizon. REFUTED if the elasticity is < -0.7 (close to lump-of-labour prediction of -1) at p<0.10 OR if the FRA event-study shows a negative employment deviation > 5% at 5-year horizon. PARTIAL between these. METHOD_VALID requires PWT avh + ILOSTAT employment-to-population on disk for >= 18 of 23 panel countries 1990-2022.
  • Falsification test: oecd_panel_employment_elasticity_to_hours_1980_2022

Estimate

  • Method: linearmodels.PanelOLS
  • Coefficient (treatment): +0.3272
  • Std error: 0.178
  • p-value: 0.0667
  • Observations: 493, countries: 18
  • Within R²: 0.386
  • Fixed effects: entity=True, time=True
  • Clustering: country

Variables resolved

  • ilostat:employment_to_population_ratio → log_employment_to_population_ratio (outcome, publisher=world_bank_wdi, n=8071)
  • world_bank_wdi:SL.UEM.TOTL.ZS → log_unemployment_rate (outcome, publisher=world_bank_wdi, n=6874)
  • pwt:rgdpo; pwt:emp → log_real_gdp_per_employed (outcome, publisher=pwt, n=10399)
  • pwt:avh → log_avg_annual_hours_worked (treatment, publisher=pwt, n=3492)
  • constructed: 1 for FRA 2000+ (Aubry laws). Used in the named-policy event-study spec. → french_35_hour_indicator (treatment, publisher=constructed, n=989)
  • world_bank_wdi:NY.GDP.PCAP.KD → log_real_gdp_per_capita (controls, publisher=world_bank_wdi, n=12104)
  • world_bank_wdi:SP.POP.TOTL → log_population (controls, publisher=world_bank_wdi, n=14447)
  • world_bank_wdi:NE.TRD.GNFS.ZS → trade_openness (controls, publisher=world_bank_wdi, n=10714)
  • world_bank_wdi:NV.SRV.TOTL.ZS → services_share_gdp (controls, publisher=world_bank_wdi, n=10330)

Generated by scripts/run_panel_fe.py at 2026-06-29T17:54:05+00:00

Notes

Distinct from reduced_working_time_output_employment which tests only FRA 2000 + ISL 2015 event-studies on a 13-country panel. This spec generalises to a 23-country OECD panel-FE elasticity estimate (different identification: within-country variation rather than two named treatments) and adds employment-to- population as primary outcome. PWT avh + ILOSTAT emp/pop are registered publishers; PWT avh coverage is good for OECD 1980-2019, with 2020-2022 patchy.

Authored framework. Read the transparency note.