China & Japan
What is the bilateral trade relationship between China and Japan? Flows in both directions over thirty years, the composition of each side’s export basket, how similar the two countries’ baskets are, and whether the pair over- or under-trades relative to what size alone would predict.
Bilateral merchandise trade, 1996-2024
Query
SELECT year,
SUM(CASE WHEN exporter_code = 156 AND importer_code = 392 THEN total_value END) * 1000 AS ex_to_im,
SUM(CASE WHEN exporter_code = 392 AND importer_code = 156 THEN total_value END) * 1000 AS im_to_ex
FROM 'data/parquet/bilateral_year/**/*.parquet'
WHERE (exporter_code = 156 AND importer_code = 392)
OR (exporter_code = 392 AND importer_code = 156)
GROUP BY year ORDER BY year;Trade balance from CHN's perspective, 1996-2024
Each side’s top products
BACI bilateral aggregates are not disaggregated at the HS6-level in the public product file, so we show each country’s top 15 HS6 exports to the world in 2024 as a proxy for the composition each side brings to this bilateral relationship. This reveals the structural overlap (or lack of it) between the two export baskets.
China: top 15 HS6 products (world exports, 2024)
Japan: top 15 HS6 products (world exports, 2024)
Export similarity
The Finger-Kreinin (1979) export-similarity index measures the overlap between two countries’ export baskets. Defined as FK(a,b) = 100 × Σᵢ min(sᵢᵃ, sᵢᵇ) over HS6 shares, it ranges from 0 (disjoint baskets) to 100 (identical). High FK indicates direct competition in third markets; low FK indicates complementarity and suggests gains from bilateral trade driven by comparative advantage rather than horizontal differentiation.
Finger-Kreinin export similarity between CHN and JPN, 1996-2024
Query
WITH shares AS (
SELECT year, country_code, product_code,
SUM(export_value) * 1.0 / NULLIF(SUM(SUM(export_value)), 0)
OVER (PARTITION BY year, country_code) AS s
FROM 'data/parquet/country_year_product/**/*.parquet'
WHERE country_code IN (156, 392)
GROUP BY year, country_code, product_code
)
SELECT year, 100 * SUM(LEAST(sa, sb)) AS fk FROM (
SELECT year, product_code,
MAX(CASE WHEN country_code = 156 THEN s ELSE 0 END) AS sa,
MAX(CASE WHEN country_code = 392 THEN s ELSE 0 END) AS sb
FROM shares GROUP BY year, product_code
) GROUP BY year ORDER BY year;Over- or under-trading?
The trade intensity index (Brown 1947, Kojima 1964) asks a simple question: does country A send a larger share of its exports to country B than B’s share of the rest of the world’s imports would predict? Formally, I = (Xᵢⱼ / Xᵢ·) / (M·ⱼ⁻ⁱ / X··⁻ⁱ). Values above 1 mean over-trading relative to size; below 1, under-trading. This is the Balassa (1965) revealed- comparative-advantage logic applied to a bilateral partner instead of a product.
Bilateral trade intensity, CHN→JPN, 1996-2024
The last decade with crisis markers
Zooming into the most recent ten years of combined bilateral flows surfaces the shape of the relationship through the two largest global shocks of the BACI window: the 2008-09 Global Financial Crisis (Baldwin 2009, The Great Trade Collapse) and the 2020 COVID-19 contraction (Espitia et al. 2022, Economic Policy). Marker dots flag each crisis year when present in the window , a visual cue for whether the pair contracted in sync with the world or decoupled from it.
Combined bilateral flow, 2015-2024, with 2008/2020 crisis markers
Intensive vs extensive margin of export growth
Hummels & Klenow (2005) decompose export growth into two margins: the intensive margin (growth in existing product lines) and the extensive margin (the net effect of adding new HS6 lines minus dropping old ones). Trade surges driven by the intensive margin reflect scale-up of established production; extensive-margin growth reflects diversification into new goods. BACI bilateral aggregates do not carry HS6, so the decomposition here is applied to China’s HS6 world-export basket , a close proxy for what shows up in the CHN→JPNflow, but not the bilateral flow directly.
China: year-over-year growth decomposition, 1997-2024
How big is the partner, from each side?
A bilateral relationship has two shares, not one: the slice of CHN’s total exports going to JPN, and the slice of JPN’s total imports coming from CHN. Size-mismatched pairs make these diverge sharply. The wedge between the two lines is a structural feature of the relationship and a first-order political-economy variable in bilateral negotiations (Krugman 1991, “The Move Toward Free Trade Zones”).