Module 05 · Trade Flows Globe

Where do US trade dollars actually flow?

A spinning globe of bilateral US trade — red for imports of goods, green for exports of services. Drag the slider to watch the China peak give way to Mexican nearshoring and Vietnamese ascent.

2025
2D · world mapTotal imports goods: $2.77T · Total exports services: $848.9B
Drill the regional dashed arcs into named individual economies
2025
BEA bilateral trade by partner country, annual. Red arcs are imports of goods (dollars leaving the US); green arcs are exports of services (dollars arriving). Arc width and apex height scale with USD magnitude.
Sharing 2D world-map view
Data table (for screen readers and reference)
Trade Globe — annual US bilateral totals ($B). As of 2025-01-01.Sum across all reporting partners per year. The interactive globe breaks these totals down by destination.
YearImports of goods ($B)Exports of services ($B)
2000974.85196.86
2001917.54187.72
2002939.17191.07
20031010.47199.34
20041169.70226.24
20051314.39246.06
20061450.68288.61
20071540.41335.87
20081634.30365.73
20091239.34349.37
20101525.01395.40
20111741.84433.91
20121832.55463.51
20131853.94480.99
20141950.64508.42
20151889.26519.81
20161829.97538.73
20171948.50577.16
20182119.15597.22
20192095.92622.66
20201953.39513.88
20212344.49560.90
20222668.05669.44
20232542.41720.05
20242695.75791.31
20252773.34848.92

How to read it

Each arc connects the United States to one of its named trading partners. Red arcs are imports of goods — dollars leaving the country. Green arcs are exports of services — dollars arriving. Width and apex height both scale with annual notional, so a fat tall arc is a big trade relationship and a thin flat arc is a small one.

By default the chart shows the 20 biggest country-level partners as solid arcs, plus four dashed regional arcs (Eurozone at Brussels, Asia-Pacific at Jakarta, Middle East at Riyadh, Africa at Nairobi) that stand in for the long tail of smaller economies. Toggle show small partners above the chart to drill those regional bundles into thinner solid arcs for ~30 named small partners — Indonesia, the UAE, Spain, Australia, Norway, Nigeria, Israel, Argentina, Poland, Hong Kong and the rest. The small-tier arcs render at half opacity so they don’t shout down the primary roster, and a shorter dot at each landing point keeps the country endpoints visible without crowding.

Why both layers? Even with the small partners revealed, the dashed regional arcs still cover dozens of named-but-not-enumerated economies (Vietnam, Cambodia, Tanzania, Bahrain, …) that we don’t plot individually. The dashed arc is the “everything else in this region” residual that always remains.

Drag the year slider to see the composition shift. Hover any arc for the partner’s annual value and its share of the US total for that year. On a phone, the globe drops to a 2D equirectangular world map — same arcs, same colour grammar, same toggle, no WebGL.

Four moments

2018 — the China peak. US imports of goods from China crested at $543B in 2018, the high-water mark of the post-WTO era. Section 301 tariffs and the trade-war tweetstorm followed, and the Chinese arc has thinned in every year since.

2019 onwards — Mexico nearshoring. The USMCA replaces NAFTA in 2020. Mexico overtakes China as the United States’ largest single source of imported goods in 2023. The Mexican arc fattens visibly across the slider.

2022 — Russia falls off the map. The invasion of Ukraine and the wholesale Western sanctions response collapse the bilateral trade relationship. The Russian arc, never large, all but disappears between the 2021 and 2023 slices.

The 2020s — Vietnam emerges. Tariff arbitrage, supply-chain redirection, and electronics assembly capacity push Vietnam from a peripheral partner to a top-10 source of US imported goods in under a decade. The arc to Hanoi grows a step every year.

Sources & caveats

  • Source: BEA International Transactions Accounts (ITA), bilateral indicators ImpGds and ExpServ at annual frequency. Run nightly via /api/cron/ingest-bea.
  • Partner coverage: The chart plots the 20 biggest country-level partners BEA ITA publishes individually as the default solid layer, ~30 long-tail small partners (also individually published — Indonesia, UAE, Spain, Australia, Norway, Nigeria, Israel, Argentina, Hong Kong, Poland, …) revealed by the show small partners toggle, plus four regional aggregates BEA also publishes (Eurozone, Asia-Pacific, Middle East, Africa). Together they cover roughly 95% of US trade. The remainder still lives inside BEA’s residual “All Other” buckets and is not surfaced anywhere on the chart.
  • Regional vs country, no double counting: Eurozone overlaps Germany / France / Italy / Netherlands / Belgium / Ireland (and now also Spain / Austria / Portugal / … from the small tier); Asia-Pacific overlaps Singapore, Vietnam, Taiwan and South Korea (and now also Indonesia / Australia / Thailand / Malaysia / Philippines); Middle East overlaps Saudi Arabia (and now also UAE / Israel / Kuwait / Qatar). The “% of US total” tooltip rows therefore stay anchored to the country-level universe (top-20 plus small) — a regional arc reads as “X% the size of the named partners’ total” rather than as another slice of the same pie. Treat regional arcs as a sense-of-scale layer, not as additive bars.
  • Annual only: BEA publishes most bilateral ITA series at quarterly frequency, but the chart uses annual to keep the year-over-year story line crisp. Sub-annual seasonality (US import surges into Q3) is invisible by design.
  • Indicators are not symmetric: we plot imports of goods against exports of services because that is the spec — the United States runs a goods deficit and a services surplus, and the asymmetry is what the chart wants to surface. Both directions for both indicators exist in the underlying data.

Related