Research: The Costs of Circumventing Tariffs

During the first few weeks of Donald Trump’s U.S. presidency, tariffs have been threatened or levied on China, Canada, Mexico, and other countries. One key economic question is whether companies in these targeted countries will circumvent these tariffs by rerouting goods through third-party countries; this is typically done by shipping goods to an affiliate or separate business partner before exporting them to the final destination. This practice allows exporters from the targeted country to illegally avoid the higher taxes assigned to their goods by labeling them as originating from a non-targeted location.

Our recent work using high-frequency trade data examines this issue by analyzing rerouting through Vietnam following the 2018 U.S.-China trade war. Our findings reveal that rerouting levels were lower than many market analysts and journalists initially thought, although they did increase significantly due to the tariffs. Notably, nearly two-thirds of this increase involved rerouting through Chinese-owned Vietnamese firms.

For business leaders, our research indicates that tariffs based on a country of origin can reduce competitive pressures from those tariff-targeted countries, but not completely. For policymakers, our results suggest that rerouting is less common than believed. However, the level of rerouting we observed still means higher price increases for U.S. consumers. Additionally, our analysis indicates that origin-specific tariffs are a cumbersome tool for targeting Chinese companies, suggesting that ownership-based duties or firm-specific sanctions are likely more effective for this purpose.

Measuring Trade Rerouting

Our working paper, “Exports in Disguise?: Trade Rerouting during the U.S.-China Trade War,” investigates whether and how Chinese products facing U.S. tariffs were rerouted through Vietnam, thus reducing the economic impact on China.

There are two main reasons for this focus. First, past research indicates that Chinese products are more likely to be rerouted through countries with close ties to China, such as those with geographical proximity, similar institutional frameworks, and shared cultural norms, including the presence of Chinese diaspora. Given these criteria, Vietnam emerges as a prime candidate for such rerouting activities. Second, among all suppliers of U.S. imports, Vietnam has benefited the most from the decline in U.S.-China trade relations. Research shows that from 2017 to 2022, Vietnam compensated for nearly half of the market share China lost in U.S. imports. Additionally, Vietnam saw its import share from China increase by 5.5 percentage points during the same period, the largest rise among all its trading partners.

While these facts are notable, they alone do not confirm the presence or scale of rerouting. For instance, Vietnamese companies might import raw materials from China and then manufacture different products for export to the U.S. This distinction is important, as trade between Vietnam and the U.S. was already growing before the trade war began. Additionally, the tariffs may have genuinely promoted growth in exports through new foreign investments in Vietnamese companies or increased capital in existing firms.

To get to the bottom of whether rerouting was happening, and to what extent, we used two microeconomic datasets: the Vietnam Enterprise Survey (VES) and S&P Global’s Panjiva. The VES offers insights into firm-level operations such as investment, production, and revenue in Vietnam. Panjiva provides detailed data on imports and exports at the eight-digit Harmonized System (HS) product level, which allows us to trace the precise movement of goods.

We defined rerouting as a movement of the same product from China to Vietnam and then from Vietnam to the U.S. within the same quarter. We then constructed three key measures of rerouting, all of which disaggregate global trade to the HS eight-digit level. (These product levels are extremely detailed; for example, one eight-digit product is “bicycle tires.”)

In the first measure, which we call “country-level rerouting,” we counted all product flows through Vietnam. To illustrate, if Vietnam imported $8 million in bicycle tires from China and exported $10 million in bicycle tires to U.S. in Q4 of 2019, we treat that as $8 million in rerouted bicycle tires. We repeated this process for each HS eight-digit product (10,081 in total) and added them together to obtain a total measure of country-level rerouting.

This country-level measure is probably a large overestimate of actual rerouting because it could capture legitimate trade. This would happen if some of the $8 million in imported bicycle tires from China were sold to final consumers in Vietnam and never shipped to the U.S.

We therefore propose two stricter measures of rerouting as well.

The second measure, which we call “province-level rerouting,” is defined as movements of the same product from China to the U.S. through a single Vietnamese province within a single quarter. Let’s say, in Q4 of 2019, $1 million of imported bicycle tires from China went to Hue Province and $7 million went to Hanoi Province. In the same quarter, Hue Province exported $5 million in bicycle tires to the U.S. and Hanoi Province exported $2.5 million in bicycle tires to the U.S. We would then count total province-level rerouting as $1 million plus $2.5 million, or $3.5 million. We repeated this procedure for every product and every province in Vietnam to obtain a total province-level rerouting estimate.

Finally, the strictest measure we use is “firm-level rerouting,” which only includes flows of eight-digit products from China to the U.S. through one Vietnamese firm. Revisiting our Q4 2019 example, suppose there were two firms in Hanoi Province, A and B. Firm A imported $4 million of bicycle tires from China and Firm B imported $3 million. Firm A did not export any bicycle tires to the U.S. at all, while Firm B exported $1 million in bicycle tires. In this case, the firm-level rerouting for Firm B would be $1 million. We repeat this process for all products and firms in Vietnam.

The Findings

Using these measures, we identify three key findings:

1. Finer measurement suggest much less rerouting than previously believed.

The granularity of our measure significantly influences the estimated extent of rerouting. In 2021, 16.5% of Vietnamese exports to the U.S. were classified as country-level rerouting in our research, in contrast to 6.5% at the province level and only 1.7% at the firm level. These figures translate to approximately $15.9 billion, $6.3 billion, and $1.6 billion in rerouted goods, respectively.

The firm-level measure likely underestimates rerouting because groups of companies could work together: some importing from China, others repackaging and relabeling, and others reexporting to the U.S. However, it’s unlikely that companies in completely different parts of the country would be coordinating in this way because of the additional transportation costs involved; therefore, the country-level measure likely overstates rerouting. Consequently, we favor the province-level measure, which captures the combined activities of companies located near each other in Vietnam.

2. Still, trade war tariffs increased rerouting through Vietnam more broadly.

For the average tariff increase on Chinese exports, country-level rerouting increased by 4.0 percentage points, province-level rerouting increased by 2.5 percentage points, and firm-level rerouting increased by 1.5 percentage points. While these treatment effects are relatively small in levels, they represent large increases over their pre-tariff (2018) levels: a 22.9% increase at the country level, a 20.5% increase at the province level, and a 15.3% increase at the firm level.

The bottom line is that there is clear evidence that re-routing is happening and that many operations have moved to Vietnam specifically to avoid tariffs. However, these figures also show that re-routing only accounts for a portion of the total export increase from Vietnam to the United States during this time. The rest of the growth represents value-added production that contributes both to Vietnamese livelihoods and to better quality and lower-priced goods for American buyers.

3. Chinese-owned firms in Vietnam have a disproportionately large role in firm-level rerouting.

Our results suggest that 61.4% of the increase in trade war rerouting was due to Chinese-owned firms located in Vietnam, suggesting that broad, origin-specific tariffs may only partially hamper Chinese exporters. U.S. consumers and businesses appear to still be using Chinese products, but pay a higher price as supply chains become longer and less efficient. One direct implication for business leaders is that Chinese goods are likely still accessing the U.S. market, and the trade war is not decreasing competitive pressure as much as Washington lawmakers might have hoped.

Lessons for Business Leaders and Policymakers

For U.S. business leaders sourcing from countries hit by tariffs, carefully vetting alternative suppliers in non-targeted regions will be increasingly important. With heightened scrutiny of rerouting, ensuring suppliers are not engaged in such practices will be key to avoiding supply-chain disruptions and unexpected customs duties.

For policymakers, know that firms will adapt to tariffs as best they can, which includes attempts to reroute exports through third-party countries. To prevent this, it’s important to gather more precise data about where this is happening. Aggregate country data could be misleading about the extent of this kind of rerouting, suggesting that it’s much more prevalent than it actually is. The risks of overestimating rerouting can result in overreactions, such as across-the-board punitive measures against countries like Vietnam, which will raise prices for American families while disrupting legitimate economic activity that is creating jobs and feeding Vietnamese families.

Our work suggests that stricter definitions of rerouting yield much lower, much more accurate estimates. In other words, taking an ownership-based approach may be more effective in ensuring products don’t slip past tariffs than an origin-based one. This can allow policymakers can better pinpoint locations and firms that are most likely to be engaged in rerouting behavior, which can facilitate a more targeted, and more successful, policy response if one is desired. At the same time, policymakers should remember that tariff increases — even if targeted at the firm-level — will still come at some expense of U.S. consumers and possibly other firms in friendly countries.

. . .

As new, expanded trade wars heat up between the U.S. and other countries, our research signals that two things can be true at once: Overall levels of rerouting are lower than previously understood, but rerouting still increased dramatically during the most recent trade war with China. More importantly, the vast majority of rerouting through Vietnam was through Chinese-owned firms. This means that policymakers should reconsider crude, country-wide tariffs in favor of sanctioning specific firms that engage in tariff circumvention.