IESET.
Hypotheses·welfare architecture·welfare_architecture_comparative_effectiveness

Forced-saving welfare architectures (Singapore CPF since 1955, Chile AFP since 1981, Australia Superannuation Guarantee since 1992) and universal-transfer architectures (Nordic model, UK NHS) produce different profiles on two dimensions: (a) retirement-income adequacy for the median retiree, and (b) long-run fiscal sustainability as measured by projected net liabilities of the pension / welfare state over 30+ year horizons.

The hypothesis is that forced-saving systems deliver comparable or better adequacy for the median saver with materially lower sovereign fiscal liability, while universal-transfer systems deliver higher redistributive effect (lower elderly poverty rates) at the cost of larger sovereign liability exposure. Both architectures face political-economy pressure to under-save or over- promise; the empirical test compares realised outcomes rather than ideal types.

INCONCLUSIVEengine/runs/welfare_architecture_comparative_effectiveness

INCONCLUSIVE_DATA_PENDING — treatment 'welfare_architecture_category' has no within-country variation under country fixed effects

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. treatment 'welfare_architecture_category' has no within-country variation under country fixed effects

why it matters

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

how the test works

It compares 16 country or place units from 1990 to 2023, using a panel fe design, with fixed effects for country and year.

what was measured
What changed
  • Welfare architecture category
What we checked
  • Retirement income replacement rate median
  • Elderly poverty rate
  • Sovereign pension net liability projection
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/welfare_architecture_comparative_effectiveness
1007550250199020072023SGPCHLAUSNORSWEDNKFIN
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show retirement_income_replacement_rate_median across 16 sampled countries over 1990–2023.
The shapes above are stylised — none of the lines are real data.
Placeholder for welfare_architecture_comparative_effectiveness. Published chart will be generated from engine/runs/welfare_architecture_comparative_effectiveness/chart_data.json.

Who has skin in the game — schools predicting on this

3 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.

Structuralist / Developmentalistwins if MIXEDpartial/no-callEmpirical Pragmatist (the framework's house position)wins if SUPPORTEDpartial/no-callOrdoliberalism (Freiburg School)wins if SUPPORTEDpartial/no-call

Pre-registration

pre-registered
first-spec commit bae09ab · 2026-04-29T22:09:42Z
run generated · 2026-06-29T17:55:02Z

Forced-saving welfare architectures (Singapore CPF since 1955, Chile AFP since 1981, Australia Superannuation Guarantee since 1992) and universal-transfer architectures (Nordic model, UK NHS) produce different profiles on two dimensions: (a) retirement-income adequacy for the median retiree, and (b) long-run fiscal sustainability as measured by projected net liabilities of the pension / welfare state over 30+ year horizons. The hypothesis is that forced-saving systems deliver comparable or better adequacy for the median saver with materially lower sovereign fiscal liability, while universal-transfer systems deliver higher redistributive effect (lower elderly poverty rates) at the cost of larger sovereign liability exposure. Both architectures face political-economy pressure to under-save or over- promise; the empirical test compares realised outcomes rather than ideal types.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

Not supported if on pooled data (a) forced-saving systems do not show comparable-or-better median replacement rates than universal- transfer systems (gap < -10pp against forced-saving), OR (b) forced- saving systems do not show materially lower sovereign pension liability projections (gap < 15pp of GDP against universal-transfer), OR (c) universal-transfer systems do not show lower elderly poverty (gap > +3pp against universal-transfer). Support requires ALL three outcomes to fall in the predicted direction, consistent with the hypothesis that the architectures differ along these dimensions rather than being equivalent.

formal test & threshold
test:      three_outcome_joint_direction_test
threshold: median_replacement_rate(forced) - median_replacement_rate(universal) > -10pp AND sovereign_liability(universal) - sovereign_liability(forced) > 15pp of GDP AND elderly_poverty(forced) - elderly_poverty(universal) > 3pp

Method

Template
panel_fe
Fixed effects
country, year
Clustering
country
Sample
16 countries · 1990 – 2023
Evidence type
associational

Panel FE on the three outcomes with architecture category as a set of dummy variables (forced_saving, universal_transfer; mixed is reference). Controls for per-capita income, dependency ratio, female LFP. Primary specification. Secondary: between-country cross-sectional regression of outcomes on architecture category using country-mean values, to capture the level differences that within-country FE absorb. Both specifications reported.

Data

VariableSourceTransform
retirement_income_replacement_rate_median
outcome
oecd:OECD.ELS.SAEtier 2
level_pct_of_pre_retirement_earnings
elderly_poverty_rate
outcome
oecd:DSD_IDDtier 2
level_pct
sovereign_pension_net_liability_projection
outcome
imf:GGXWDG_NGDPtier 2
oecd:pension_spending_projectionstier 2
level_pct_of_gdp_2050_projection
welfare_architecture_category
treatment
constructed:categorical per country per epochtier 5
categorical
gdp_per_capita_ppp
control
world_bank_wdi:NY.GDP.PCAP.PP.KDtier 2
log
dependency_ratio
control
world_bank_wdi:SP.POP.DPND.OLtier 2
level
female_labour_force_participation
control
world_bank_wdi:SL.TLF.CACT.FE.ZStier 2
level

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Result card — welfare_architecture_comparative_effectiveness

Verdict: INCONCLUSIVE_DATA_PENDING — treatment 'welfare_architecture_category' has no within-country variation under country fixed effects

Pre-registration

  • Claim: Forced-saving welfare architectures (Singapore CPF since 1955, Chile AFP since 1981, Australia Superannuation Guarantee since 1992) and universal-transfer architectures (Nordic model, UK NHS) produce different profiles on two dimensions: (a) retirement-income adequacy for the median retiree, and (b) long-run fiscal sustainability as measured by projected net liabilities of the pension / welfare state over 30+ year horizons. The hypothesis is that forced-saving systems deliver comparable or better adequacy for the median saver with materially lower sovereign fiscal liability, while universal-transfer systems deliver higher redistributive effect (lower elderly poverty rates) at the cost of larger sovereign liability exposure. Both architectures face political-economy pressure to under-save or over- promise; the empirical test compares realised outcomes rather than ideal types.
  • Falsification rule: Not supported if on pooled data (a) forced-saving systems do not show comparable-or-better median replacement rates than universal- transfer systems (gap < -10pp against forced-saving), OR (b) forced- saving systems do not show materially lower sovereign pension liability projections (gap < 15pp of GDP against universal-transfer), OR (c) universal-transfer systems do not show lower elderly poverty (gap > +3pp against universal-transfer). Support requires ALL three outcomes to fall in the predicted direction, consistent with the hypothesis that the architectures differ along these dimensions rather than being equivalent.
  • Falsification test: three_outcome_joint_direction_test

Estimate

  • Error: treatment 'welfare_architecture_category' has no within-country variation under country fixed effects

Variables resolved

  • oecd:DSD_IDD@DF_IDD (poverty_rate, 66+ age group) → elderly_poverty_rate (outcome, publisher=oecd, n=902)
  • imf:GGXWDG_NGDP; oecd:pension_spending_projections → sovereign_pension_net_liability_projection (outcome, publisher=imf, n=8113)
  • constructed: categorical per country per epoch → welfare_architecture_category (treatment, publisher=constructed, n=544)
  • world_bank_wdi:NY.GDP.PCAP.PP.KD → gdp_per_capita_ppp (controls, publisher=world_bank_wdi, n=8325)
  • world_bank_wdi:SP.POP.DPND.OL → dependency_ratio (controls, publisher=world_bank_wdi, n=16935)
  • world_bank_wdi:SL.TLF.CACT.FE.ZS → female_labour_force_participation (controls, publisher=world_bank_wdi, n=8302)

Variables missing data

  • oecd:OECD.ELS.SAE,DSD_PENSIONS@DF_PENSIONS_REPL_RATE,1.0 (outcome, name=retirement_income_replacement_rate_median) — vintage not on disk

Generated by scripts/run_panel_fe.py at 2026-06-29T17:55:02+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

Data-gated. Primary publishers: oecd, world_bank_wdi, imf. Secondary (forced-saving country-specific): singapore_cpf (pending), chile_ spensiones (pending), apra (pending). Until those land, forced-saving measures rely on OECD Pensions at a Glance harmonised tabulations. The hypothesis is pre-registered now to lock the comparison logic; first-run execution waits on publisher readiness.

Authored framework. Read the transparency note.