Lost in tabulation: Diverging US-China trade data point to tariff evasion

The United States and China share one of the world's largest and most closely monitored bilateral trading relationships. Yet the official statistics used to assess that relationship are sending sharply different signals. China reports exporting more goods to America than the United States says it imports—particularly since early 2025, when the Trump administration sharply increased tariffs on China. Among possible reasons for the gap, the evasion of US tariffs appears to be an important contributor. As a consequence, US bilateral import data for China appear to have become less reliable. The reported decline in US imports from China should not be treated as a clean measure of either economic decoupling or the effectiveness of US tariffs.

From January through July 2026, the US Census Bureau recorded $158.3 billion in merchandise imports from China (including Hong Kong). Over the same period, China reported $258.1 billion in merchandise exports to the United States. The $99.8 billion difference was equivalent to 63 percent of reported US imports from China. This share was nearly five times its 2021–24 average. If the average monthly gap observed from January through July persists for the rest of the year, the cumulative gap for 2026 will exceed $170 billion.

Figure 1 shows how China's reported goods exports to the United States compare with America's reported goods imports from China since early 2024: Both declined sharply starting in February 2025 following US tariff increases.[1] Their subsequent trajectories diverged. By July 2026, China reported monthly exports to the United States about 4 percent below their 2024 average. By contrast, US-reported imports from China were 26 percent below their 2024 average, indicating a more persistent contraction in bilateral trade.

Why have US-China trade statistics diverged?

US-China bilateral trade statistics have never fully aligned. Before 2020, the gap was negative: The United States generally reported more imports than China reported as exports to the United States. The pattern reversed in 2020. The gap became positive and averaged roughly 13 percent of US imports during 2021–24. Federal Reserve researchers estimated that tariff evasion accounted for about $55 billion of the $88 billion shift in the reporting gap between 2018 and 2020. The tariffs imposed in 2025 were substantially higher and covered more goods than the 2018–20 measures.

Tariff evasion can take several forms. Firms may misstate a shipment's country of origin to take advantage of lower bilateral tariff rates or reclassify goods when rates differ across product categories. They may also understate the reported value or quantity of imports. Because import and export declarations are filed separately, underinvoicing can create a gap between reported exports and imports. A recent New York Times analysis found that the average value of merchandise per shipping container from China fell by around 40 percent between January 2025 and February 2026. In some cases, Chinese exporters may be responsible for the US customs entries. For example, a Nikkei Asia investigation describes allegations that logistics providers used shell companies as importers of record and manipulated invoices to reduce the duties owed. Nikkei separately reports that some Chinese companies have been reluctant to claim US tariff refunds because renewed scrutiny could expose irregularities in the underlying entries.

The effects of tariff evasion are not confined to US-China trade data. False declarations of origin create two gaps in bilateral trade data: a positive gap for the true country of origin, as observed with China, and an offsetting negative gap for the false country of origin. Relabeling of Chinese exports through Vietnam appears substantial. As New York Federal Reserve researchers note, its reporting gap runs in the opposite direction from China's (figure 2). From January through July 2026, the United States recorded $149.6 billion in imports from Vietnam, while Vietnam's customs agency reported $105.0 billion in exports to the United States. The resulting $44.6 billion discrepancy is consistent with some Chinese goods being relabeled as Vietnamese, although aggregate data cannot establish exactly how much relabeling occurred.[2]

Do other factors explain the current US-China gap?

Other developments have contributed to the gap, but they appear much less significant than tariff evasion. First, the emergence of a positive gap in 2020 coincided with rapid growth in low-value shipments entering under the US de minimis exemption, which allowed eligible imports valued at $800 or less to enter duty free with reduced documentation. China and Hong Kong accounted for roughly two-thirds of US de minimis shipments around this time, according to US Customs and Border Protection (CBP). The Congressional Research Service estimates that Chinese low-value e-commerce exports to the United States rose from $1.4 billion in 2018 to $22.9 billion in 2024. Since these exports were more likely to appear in Chinese than US reporting, they contributed to the gap.[3]

However, the United States ended duty-free de minimis treatment for shipments from China in May 2025 and suspended it globally that August. China's Ministry of Commerce also issued guidance in June 2024 encouraging e-commerce retailers to use overseas warehouses for fulfillment, shifting goods from individual parcels into bulk shipments subject to ordinary customs reporting. The value of global imports entering the United States under the de minimis exemption fell 26 percent in fiscal year 2025. This should have narrowed the reporting gap.

Chinese customs data through 2026 reinforce this conclusion. Although the categories shown in figure 3 cover a narrower range of goods than the CBP estimates, their trend remains informative. Chinese exports to the United States in these categories peaked in 2024 and declined thereafter. The US series for low-value imports from China requires a different interpretation. It shows a clear structural break in May 2025, when the de minimis exemption ended and additional entry and reporting requirements took effect. Its subsequent rise likely reflects improved coverage rather than growth in low-value trade. Convergence with the Chinese series should have further narrowed the portion of the overall gap attributable to low-value shipments.

A second possible cause of the US-China reporting gap stems from Chinese exporters artificially inflating export values to claim tax rebates from the Chinese government. The Fed analysis cited earlier estimated that changes in Chinese tax incentives accounted for about $12 billion of the $88 billion shift in the reporting gap between 2018 and 2020. There are two reasons to think this was not a major contributor to the gap's recent growth. First, China eliminated or reduced export rebates for selected products beginning in December 2024 and announced further cuts in January 2026. This contrasts with the 2018 episode, when it increased rebates to offset US tariffs. Second, Chinese authorities have intensified tax enforcement since 2024, inspecting more than 130,000 firms suspected of fraudulent invoicing.

Conclusion

The exact causes of the reporting gap remain uncertain. Still, tariff evasion appears the most likely explanation for its recent growth. To better counter tariff evasion and tax fraud, more systematic sharing and matching of transaction-level customs data could benefit both the United States and China. A framework for such cooperation already exists, but implementation appears limited. More broadly, the episode illustrates the limits of using discriminatory tariffs to address bilateral trade imbalances. As argued in a recent PIIE Policy Brief by Mary E. Lovely and Christine Y. Wan, such tariffs can redirect trade through third countries without substantially reducing US reliance on Chinese content or China's role in global production. Addressing trade imbalances instead requires domestic adjustment on both sides: The United States needs to rein in its large fiscal deficits, which contribute to its trade deficit, while China needs to stimulate domestic consumption and reduce industrial overcapacity.

Notes

1. In 2025, the average effective tariff rate on US merchandise imports from China and Hong Kong was 30 percent. It remained elevated at 24 percent in 2026, through July, compared with a 6 percent average effective rate for imports from other countries.

2. Subtracting the 2021–24 average monthly gap from subsequent observations (through July 2026) leaves a cumulative gap of $109.7 billion for China and Hong Kong and a negative $62.5 billion gap for Vietnam. If Vietnam's gap entirely reflected relabeled Chinese exports, such relabeling would account for 57 percent of the adjusted US-China gap.

3. China records low-value shipments directly, whereas US bilateral totals use Census estimates that rely on historical trade patterns and may miss recent changes.