Pre-registration
Credible fiscal consolidation episodes — defined as sustained primary balance improvement of at least 2% of GDP over 3 years, not reversed within 5 years, and accompanied by declining debt-to-GDP trajectories — predict stronger subsequent 10-year real GDP per capita growth and private investment than repeated discretionary fiscal stimulus in mature economies, in an OECD panel 1980-2020. The directional claim is that credible consolidators experience cumulative growth 5-15% above matched non-consolidating peers over the post-consolidation decade, while stimulus-reliant economies show no significant gain or a relative loss.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
SUPPORTED if the average synthetic-control gap for real GDP per capita growth is positive and significant (permutation p<0.10) over the 10-year post-consolidation window, AND the private-investment gap is also positive and significant. PARTIAL if growth gap is positive but investment gap is not (consumption-driven expansion). REFUTED if the growth gap is negative and significant at p<0.10. INFORMATIVE: excluding the 2010-2015 European austerity episodes should not eliminate the result; if it does, the finding is a post-GFC Eurozone story rather than a general consolidation effect.
formal test & threshold
test: synthetic_control_credible_consolidation_vs_stimulus_growth_10yr threshold: Synthetic-control ATT (GDP pc growth, 10yr) > 0 at permutation p<=0.10 AND Synthetic-control ATT (private investment, 10yr) > 0 at permutation p<=0.10 AND Ex-Eurozone-austerity robustness retains sign of both ATTs.
Method
- Template
synthetic_control- Clustering
country- Sample
- 35 countries · 1980 – 2020
- Evidence type
- causal
Primary: synthetic control for each credible consolidation episode, matching on pre-consolidation GDP per capita, debt, investment, institutional quality, and unemployment over a 10-year donor window. Donor pool: mature OECD economies that did not consolidate in the same period. Secondary: Callaway-Sant'Anna staggered DiD with never-treated and not-yet-treated controls. Tertiary: local projections tracing dynamic response to consolidation initiation.
Data
| Variable | Source | Transform |
|---|---|---|
real_gdp_per_capita_growth outcome | world_bank_wdi:NY.GDP.PCAP.KDtier 2 | annual_log_change |
private_investment_share_gdp outcome | world_bank_wdi:NE.GDI.FTOT.ZStier 2 | level |
long_term_interest_rate outcome | imf:long_term_interest_ratetier 2 | level |
credible_consolidation_indicator treatment | constructed:indicator = 1 during and 10 years after credible consolidation episodetier 5 | indicator |
repeated_stimulus_indicator treatment | constructed:indicator = 1 for countries with >3 discretionary fiscal expansions (>1.5pp cyclically-adjusted primary balance deteriortier 5 | indicator |
sovereign_bond_yield_spread channel | constructed:10yr government bond yield minus German bund (EU) or US Treasury (non-EU)tier 5 | level |
business_confidence_index channel | oecd:business_confidencetier 2 | level |
initial_log_gdp_per_capita control | world_bank_wdi:NY.GDP.PCAP.KDtier 2 | log |
institutional_quality control | wgi:RL.ESTtier 4 | level |
trade_openness control | world_bank_wdi:NE.TRD.GNFS.ZStier 2 | level |
initial_debt_to_gdp control | imf:GGXWDG_NGDPtier 2 | pre_episode_level |
unemployment_rate control | ilo:unemployment_ratetier 2 | pre_episode_mean |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Result card — fiscal_consolidation_credibility_growth
Verdict: supported — High-credibility consolidators +2.76%/yr vs comparator median +1.78%/yr (diff +0.98pp, n=287).
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 readiness: - IMF FAD / OECD Economic Outlook primary balance (ready) - Alesina-Ardagna consolidation dataset (ready) - WDI GDP pc, private investment (ready) - IMF WEO debt, long-term interest rate (ready) - WGI RL.EST (ready) - ILO unemployment (ready) - Sovereign bond yields (ready) - OECD BCI business confidence (pending)