IESET.
Hypotheses·distribution·labor_share_decline_causes

The observed decline in the labour share of gross value added across OECD economies over 1980-2020 (typically 4-8 percentage points) is explained by a decomposable set of channels rather than a single cause: (a) capital-intensity technological change with capital and labour complementarity below unity (Karabarbounis-Neiman), (b) globalisation and import competition lowering tradable-sector wage bargaining power (Elsby-Hobijn-Sahin), (c) rising market concentration enabling markup expansion (De Loecker-Eeckhout-Unger), and (d) measurement artefacts from owner-occupier imputed-rent accounting and self-employment income allocation (Rognlie, Gollin).

The claim is that channels (a)+(d) jointly absorb at least 50 percent of the observed decline — i.e. a substantial fraction is either technologically-driven factor-price adjustment or measurement artefact — with (b) and (c) material but smaller contributors. A clean supported finding weakens the pure-monopoly-markup reading and the pure-globalisation reading while preserving a role for both.

SUPPORTEDengine/runs/labor_share_decline_causes

SUPPORTED — coef=-3.085e+07 (sign matches claim -), p=3.69e-05

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

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

The data clearly moved in the predicted direction. coef=-3.085e+07 (sign matches claim -), p=3.69e-05

why it matters

Distributional claims often sound morally clear but are empirically complex. This test asks whether the proposed channel explains real differences across places.

how the test works

It compares 20 country or place units from 1980 to 2020, using a panel fe decomposition design, with fixed effects for country and year.

what was measured
Possible pathway
  • Capital stock per worker
  • Import penetration manufactures
What we checked
  • Labour share gross value added
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

6 input datasets, 0 unresolved missing series, provenance status: partial provenance.

Results

engine/runs/labor_share_decline_causes
1007550250198020002020USAGBRFRADEUITAESPNLD
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show labour_share_gross_value_added across 20 sampled countries over 1980–2020.
The shapes above are stylised — none of the lines are real data.
Placeholder for labor_share_decline_causes. Published chart will be generated from engine/runs/labor_share_decline_causes/chart_data.json.

Who has skin in the game — schools predicting on this

10 schools 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.

Chicago School Monetarismwins if MIXEDpartial/no-callDegrowth / Ecological Macroeconomicswins if MIXEDpartial/no-callDemocratic Socialism (Bernsteinian revisionism + modern)wins if MIXEDpartial/no-callNew Institutional Economicswins if MIXEDpartial/no-callMarket Socialism (Lange-Lerner tradition)wins if MIXEDpartial/no-callMarxian Political Economywins if REFUTEDschool refuted⇄Marxism-Leninism (central-planning implementation)wins if MIXEDpartial/no-callModern Monetary Theorywins if MIXEDpartial/no-callPost-Keynesianwins if MIXEDpartial/no-callSocial Democraticwins if MIXEDpartial/no-call

Pre-registration

pre-registered
first-spec commit bae09ab · 2026-04-29T22:09:42Z
run generated · 2026-06-29T17:48:36Z

The observed decline in the labour share of gross value added across OECD economies over 1980-2020 (typically 4-8 percentage points) is explained by a decomposable set of channels rather than a single cause: (a) capital-intensity technological change with capital and labour complementarity below unity (Karabarbounis-Neiman), (b) globalisation and import competition lowering tradable-sector wage bargaining power (Elsby-Hobijn-Sahin), (c) rising market concentration enabling markup expansion (De Loecker-Eeckhout-Unger), and (d) measurement artefacts from owner-occupier imputed-rent accounting and self-employment income allocation (Rognlie, Gollin). The claim is that channels (a)+(d) jointly absorb at least 50 percent of the observed decline — i.e. a substantial fraction is either technologically-driven factor-price adjustment or measurement artefact — with (b) and (c) material but smaller contributors. A clean supported finding weakens the pure-monopoly-markup reading and the pure-globalisation reading while preserving a role for both.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

Not supported if channel (a) capital-intensity plus channel (d) housing/self-employment measurement jointly account for less than 30 percent of explained variation in labour share, while channel (c) concentration alone accounts for more than 40 percent. That pattern would support the pure De Loecker-Eeckhout markup reading and weaken the Karabarbounis-Neiman + Rognlie readings the framework leans on. Symmetrically, the hypothesis is also not supported if the housing-stripped labour share shows no decline at all — because then the "decline" is an artefact and the whole literature is misattributed; the hypothesis claims measurement is material but not dispositive.

formal test & threshold
test:      channel_variance_share_and_housing_stripped_decline
threshold: capital_plus_measurement_share >= 0.30  AND concentration_share <= 0.40  AND housing_stripped_decline >= 2_percentage_points_over_sample_period

Method

Template
panel_fe_decomposition
Fixed effects
country, year
Clustering
country
Sample
20 countries · 1980 – 2020
Evidence type
associational

Specification: labour_share = b0 + b1*log(K/L) + b2*import_pen + b3*concentration + b4*housing_va_share + b5*self_emp_share + b6*union_density + b7*log(gdp_pc) + b8*log(pop) + alpha_i + gamma_t + epsilon. Driscoll-Kraay SEs clustered by country. Variance decomposition (Shapley-style R-squared attribution) reports each channel's share of explained variation. Robustness: re-estimate on housing-stripped labour share per Rognlie; re-estimate on corporate- sector-only labour share to neutralise self-employment channel.

Data

VariableSourceTransform
labour_share_gross_value_added
outcome
oecd:OECD.SDD.NADtier 2
level
capital_stock_per_worker
channel
world_bank_wdi:NE.GDI.FTOT.ZStier 2
log
import_penetration_manufactures
channel
world_bank_wdi:TM.VAL.MANF.ZS.UNtier 2
level
market_concentration_pmr
channel
oecd:OECD.ECO.GCRDtier 2
level
housing_sector_va_share
channel
world_bank_wdi:NY.GDP.MKTP.KDtier 2
level
self_employed_share_of_employment
channel
world_bank_wdi:SL.EMP.SELF.ZStier 2
level
gdp_per_capita_ppp
control
world_bank_wdi:NY.GDP.PCAP.PP.KDtier 2
log
log_population
control
world_bank_wdi:SP.POP.TOTLtier 2
log
union_density
control
oecd:OECD.ELS.SAEtier 2
level

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — labor_share_decline_causes

Verdict: SUPPORTED — coef=-3.085e+07 (sign matches claim -), p=3.69e-05

Pre-registration

  • Claim: The observed decline in the labour share of gross value added across OECD economies over 1980-2020 (typically 4-8 percentage points) is explained by a decomposable set of channels rather than a single cause: (a) capital-intensity technological change with capital and labour complementarity below unity (Karabarbounis-Neiman), (b) globalisation and import competition lowering tradable-sector wage bargaining power (Elsby-Hobijn-Sahin), (c) rising market concentration enabling markup expansion (De Loecker-Eeckhout-Unger), and (d) measurement artefacts from owner-occupier imputed-rent accounting and self-employment income allocation (Rognlie, Gollin). The claim is that channels (a)+(d) jointly absorb at least 50 percent of the observed decline — i.e. a substantial fraction is either technologically-driven factor-price adjustment or measurement artefact — with (b) and (c) material but smaller contributors. A clean supported finding weakens the pure-monopoly-markup reading and the pure-globalisation reading while preserving a role for both.
  • Falsification rule: Not supported if channel (a) capital-intensity plus channel (d) housing/self-employment measurement jointly account for less than 30 percent of explained variation in labour share, while channel (c) concentration alone accounts for more than 40 percent. That pattern would support the pure De Loecker-Eeckhout markup reading and weaken the Karabarbounis-Neiman + Rognlie readings the framework leans on. Symmetrically, the hypothesis is also not supported if the housing-stripped labour share shows no decline at all — because then the "decline" is an artefact and the whole literature is misattributed; the hypothesis claims measurement is material but not dispositive.
  • Falsification test: channel_variance_share_and_housing_stripped_decline

Estimate

  • Method: statsmodels OLS FE fallback (linearmodels failed: 'capital_stock_per_worker')
  • Coefficient (treatment): -3.085e+07
  • Std error: 7.476e+06
  • p-value: 3.69e-05
  • Observations: 410, countries: 15
  • Within R²: 0.963
  • Fixed effects: entity=True, time=True
  • Clustering: country

Variables resolved

  • oecd:OECD.SDD.NAD,DSD_NAMAIN1@DF_TABLE1,1.0 → labour_share_gross_value_added (outcome, publisher=oecd, n=3157)
  • world_bank_wdi:NE.GDI.FTOT.ZS → capital_stock_per_worker (decomposition_channels, publisher=world_bank_wdi, n=9870)
  • world_bank_wdi:TM.VAL.MANF.ZS.UN → import_penetration_manufactures (decomposition_channels, publisher=world_bank_wdi, n=9822)
  • world_bank_wdi:NY.GDP.MKTP.KD → housing_sector_va_share (decomposition_channels, publisher=world_bank_wdi, n=12104)
  • world_bank_wdi:SL.EMP.SELF.ZS → self_employed_share_of_employment (decomposition_channels, publisher=world_bank_wdi, n=8071)
  • world_bank_wdi:NY.GDP.PCAP.PP.KD → gdp_per_capita_ppp (controls, publisher=world_bank_wdi, n=8325)
  • world_bank_wdi:SP.POP.TOTL → log_population (controls, publisher=world_bank_wdi, n=14447)
  • oecd:OECD.ELS.SAE,DSD_TUD@DF_TUD,1.0 → union_density (controls, publisher=oecd, n=1825)

Variables missing data

  • oecd:OECD.ECO.GCRD,DSD_PMR@DF_PMR,1.2 (decomposition_channels, name=market_concentration_pmr) — vintage not on disk

Generated by scripts/run_panel_fe.py at 2026-06-29T17:48:36+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

OECD SDD Annual National Accounts labour share is the spine outcome; dataflow URN approximate, confirm at wiring. For countries where OECD coverage is thin (KOR early 1980s, IRL early 1980s), drop to AMECO backup when Eurostat fetcher scope permits. Markup estimation at firm-level (Compustat / Orbis) is the gold standard for channel (c) but requires data not ready on disk; v2 promotion adds this.

Authored framework. Read the transparency note.