Case studies
Seven original analyses built on this workbench. Each reads like a working-paper figure: a question, a chart, a reading, a method, a citation. All reproducible from the Parquet files and the SQL shown on each page.
The China shock, thirty years on
A 7-figure deep-dive on China's 1995-2024 manufacturing surge. Export shares by HS Section, RCA basket widening + PCI upgrade, country-level ADH exposure, price compression in targeted products, post-2018 reallocation to VNM/MEX/IDN/IND, ECI catch-up vs peers, and how Bangladesh picked up apparel share as China retreated.
Methods: Share decomposition, RCA (Balassa 1965), PCI (Hausmann-Hidalgo 2009), ADH exposure (Autor-Dorn-Hanson 2013), unit-value compression (Broda-Weinstein 2006).
US-China trade war 2018-2024
Section 301 tariffs: who actually paid. DiD on List-3 HS6 products vs control, unit-value pass-through tests, import-share reshuffling by source (Vietnam, Mexico, Taiwan, India), China's reorientation toward non-US markets. Built on BACI bilateral at the finest granularity available.
Methods: Difference-in-differences, pass-through estimation (Amiti-Redding-Weinstein 2019, Cavallo-Gopinath-Neiman-Tang 2021), third-country diversion (Fajgelbaum-Goldberg-Kennedy-Khandelwal 2020).
Five policy shocks, five event studies
China WTO accession (2001), Section 301 trade war (2018), Russia sanctions (2022), Brexit (2020), COVID-19. Each with treatment-control setup, synthetic counterfactual where relevant, and canonical citation to the causal paper. A single page showing how the platform handles identification.
Methods: DiD, synthetic control, pre-2022 OLS extrapolation (sanctions), indexed-series diff (Brexit), HS-section shock decomposition (COVID).
Economic complexity, replicated
Hausmann-Hidalgo ECI/PCI applied to the latest BACI vintage (2023). The ECI-income fingerprint, top 30 by complexity, 6-country trajectories (USA/CHN/DEU/KOR/VNM/BGD), mean PCI by HS Section. Links to the full 200+ year trade-complexity-growth ecosystem of the site.
Methods: Spectral decomposition of M-matrix (Hidalgo-Hausmann 2009 PNAS), Mealy-Farmer iterative refinement.
Gravity of trade: estimating and testing
Distance-decay scatter, size elasticity, 25-year time series of distance coefficients (Disdier-Head "puzzle"), CEPII-computed gravity residuals identifying over/under-traders, distance-bucket sanity checks. Closed-form OLS implemented directly in DuckDB.
Methods: Frisch-Waugh-Lovell, OLS gravity, residual analysis. Anderson-van Wincoop (2003), Silva-Tenreyro (2006), Head-Mayer (2014).
Supply-chain concentration and critical minerals
Which products are controlled by a single country. Top HHI products, >50% single-exporter products by total value, country-level import-dependence, critical-minerals (cobalt, lithium, rare earths) concentration, and the long-run rise of top-1 exporter shares.
Methods: HHI concentration, geographic-dependence (OECD GeoDep), critical-mineral case studies per IEA / USGS.
Interactive tariff welfare simulator
Small-country Armington PE: drag the sliders, see imports fall, tariff revenue rise, deadweight loss open. Three preset scenarios (standard hike, Section-301 analog, prohibitive). Client-side compute, no server round-trip.
Methods: Armington (1969), Harberger triangle, Broda-Weinstein elasticities.
Current research agenda
Four threads, each built directly on the workbench’s Parquet corpus:
- RMG exit elasticities for Bangladesh. Quantifying how apparel exporters substitute across destinations when a major buyer imposes a tariff or an NTM shock, using BACI bilateral variation at HS6 and WITS preferential tariff schedules as first-stage shifters.
- Third-country diversion after Section 301. Extending Fajgelbaum et al. (2020) past their 2019 cutoff with the 2020–2024 BACI vintage. Where did the reallocation settle, and is the pass-through wedge widening or closing?
- Commodity terms-of-trade shocks and food-security thresholds. FAOSTAT self-sufficiency ratios and import-bill shares crossed with Pink Sheet prices to identify which low- and middle-income economies cross a fiscal-stress threshold in the next price spike.
- Upstreamness, concentration, and the shape of supply-chain risk. OECD TiVA upstreamness combined with BACI HHI to distinguish products that are concentrated because they are intermediates (unavoidable) from products concentrated because of resource endowment (substitutable).
Open questions
- Is the post-2018 reallocation from China to Vietnam/Mexico a one-time level shift or a slope change, and can BACI tell the two apart with four years of post-treatment data?
- Do Bangladesh’s high-PCI exports (pharmaceuticals, leather goods) actually embed higher knowledge content, or are they riding tariff preferences? A quality-ladder decomposition across 2005 and 2024 should resolve this.
- Does the ESCAP–World Bank inverted-gravity trade-cost equivalent track the post-2020 supply-chain fragmentation narrative, or does it smooth it away?
Author
Md Deluair Hossen, PhD is a postdoctoral scholar at the University of Tennessee with ten years of applied international-trade and development-policy research. Open to senior trade economist, research economist, policy analyst, and data-science roles.