Pre-registration
Israel's 2000-2024 transformation into a high-tech-export-led economy ("Start-Up Nation") produced a sustained labour-productivity acceleration, services-export composition shift, and high-skilled-employment expansion exceeding OECD peers without comparable defence-tech / venture-capital ecosystems. The pre-registered claim is that, in a synthetic-control design with small-open-economy OECD peers (Ireland, Denmark, Finland, New Zealand, Singapore is-non-OECD-but-comparable), Israel's cumulative labour-productivity (real GDP per hour worked) growth 2000-2024 exceeds the synthetic counterfactual by at least 25 log-points AND ICT services export share rises by at least 10 percentage points more than counterfactual. The null counter-claim is that Israel's headline outperformance is driven primarily by Russian/Soviet-Jewish-immigrant human-capital absorption (1989-1995 cohort) rather than continuing 2000-2024 policy content.
Falsification criterion — what would disprove this
This hypothesis is considered falsified if:
Not supported if EITHER (a) labour-productivity (real GDP per hour) cumulative gap 2000-2024 vs synthetic counterfactual is less than +25 log-points at p_perm < 0.10, OR (b) ICT services export share gap is less than +10 ppts, OR (c) the cumulative gap controlling for 1995-2000 Russian-Jewish-immigrant-cohort human- capital effect (proxied by post-1989 net-migration share) is reduced by more than 50% — i.e. the immigration channel dominates and the policy-content channel is not separable.
formal test & threshold
test: synth_did_with_immigration_attribution_check threshold: cumulative_log_productivity_gap_2000_2024 >= 0.25 at p_perm < 0.10 AND ict_services_export_share_gap >= 10 ppts AND immigration_adjusted_residual_gap >= 0.50 * raw_gap
Method
- Template
synth_did- Clustering
country- Sample
- 6 countries · 1995 – 2024
- Evidence type
- causal
Primary: synth_did with ISR treated from 2000 and small-open-economy OECD donor pool. Robustness drops Ireland (FDI-distortion confound on services exports) and drops Singapore (different geographical/institutional regime). Cumulative-growth contour reported at 2005, 2010, 2015, 2020, 2024 to show trajectory shape.
Data
| Variable | Source | Transform |
|---|---|---|
real_gdp_per_hour outcome | oecd:gdp_per_hourtier 2 pwt:rgdpetier 3 pwt:avhtier 3 | log_level |
real_gdp_pc outcome | world_bank_wdi:NY.GDP.PCAP.KDtier 2 pwt:rgdpetier 3 | log_level |
ict_services_export_share outcome | world_bank_wdi:BX.GSR.CCIS.ZStier 2 oecd:trade_in_servicestier 2 | level |
high_tech_employment_share outcome | oecd:employment_high_techtier 2 ilostat:EAR_INEE_SEX_AGE_OCU_NBtier 2 | level |
high_tech_pivot_indicator treatment | constructed:1 for ISR from 2000 onwardtier 5 | binary |
tertiary_enrolment_share control | world_bank_wdi:SE.TER.ENRRtier 2 oecd:educationtier 2 | level |
rd_share_gdp control | world_bank_wdi:GB.XPD.RSDV.GD.ZStier 2 oecd:rd_expendituretier 2 | level |
us_policy_rate control | fred:FEDFUNDStier 1 | level |
us_nasdaq_index control | fred:NASDAQCOMtier 1 | log_level |
● ready · ● pending · ● reconstruct-needed
Detailed result card
Result card — mena_israel_high_tech_economy_2000_2024
Verdict: PARTIAL — mean_gap=-1.21e+04, |gap|/pre_sd=17, p_perm=1 (gap below 0.5×pre_sd or placebo p≥0.10)
Pre-registration
- Claim: Israel's 2000-2024 transformation into a high-tech-export-led economy ("Start-Up Nation") produced a sustained labour-productivity acceleration, services-export composition shift, and high-skilled-employment expansion exceeding OECD peers without comparable defence-tech / venture-capital ecosystems. The pre-registered claim is that, in a synthetic-control design with small-open-economy OECD peers (Ireland, Denmark, Finland, New Zealand, Singapore is-non-OECD-but-comparable), Israel's cumulative labour-productivity (real GDP per hour worked) growth 2000-2024 exceeds the synthetic counterfactual by at least 25 log-points AND ICT services export share rises by at least 10 percentage points more than counterfactual. The null counter-claim is that Israel's headline outperformance is driven primarily by Russian/Soviet-Jewish-immigrant human-capital absorption (1989-1995 cohort) rather than continuing 2000-2024 policy content.
- Falsification rule: Not supported if EITHER (a) labour-productivity (real GDP per hour) cumulative gap 2000-2024 vs synthetic counterfactual is less than +25 log-points at p_perm < 0.10, OR (b) ICT services export share gap is less than +10 ppts, OR (c) the cumulative gap controlling for 1995-2000 Russian-Jewish-immigrant-cohort human- capital effect (proxied by post-1989 net-migration share) is reduced by more than 50% — i.e. the immigration channel dominates and the policy-content channel is not separable.
Synthetic-control estimate
- shape: synth_did
- treated_country: ISR
- event_year: 2000
- n_donors: 5
- donor_weights (top): {'DNK': 0.4641, 'NZL': 0.396, 'SGP': 0.1399, 'IRL': 0.0, 'FIN': 0.0}
- pre_rmse: 9871.132423419527
- pre_period_sd: 696.6729510569214
- mean_post_gap: -12101.727363285501
- end_period_gap: -12625.55370504361
- post_period_years: [2000, 2024]
- placebo_p_value: 1.0
- n_placebos: 5
- method: synthetic-control via NNLS, permutation inference
Variables resolved
world_bank_wdi:NY.GDP.PCAP.KD; pwt:rgdpe→ real_gdp_pc (outcome, n=14131)world_bank_wdi:SE.TER.ENRR; oecd:education→ tertiary_enrolment_share (controls, n=7272)
Generated by scripts/run_synth_did.py at 2026-04-30T10:15:31+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
Designed to be orthogonal to demo_israel_soviet_absorption_1989_1994 (which tests the 1989-1994 immigration absorption); this hypothesis tests the post-2000 high- tech-led growth model. The post-October-7-2023 war effect on the high-tech sector is too recent to be fully reflected in 2024 data; v1.1 spec will revisit.