Investigation into Customs Practices
One of the pressing questions in trade economics right now is: who is responsible for customs duties?
There’s a lot at stake, with hundreds of billions in tariffs being collected. These customs revenues certainly help to reduce the budget deficit for the U.S. However, it’s the importers who pay these duties directly. But this raises a key question: Who actually feels the financial burden? Is it the importers who cut prices to keep their market share, or could it be foreign manufacturers? Are importers absorbing these taxes, or are they simply passing them on to consumers through higher prices?
Research from the Kiel Institute in Germany suggests that the resolution of the tariff dispute reveals “clear evidence” indicating that U.S. importers and consumers will shoulder an overwhelming 96 percent of tariff costs by 2025, while ore exporters only take on 4 percent. The authors label these tariffs as an “own goal.”
But, there’s a catch. The study doesn’t convincingly prove its point.
Perception vs. Reality
The impressive figure of 96% shift to U.S. buyers sounds convincing, but it’s built on shaky statistical grounds, contrary to the authors’ claims.
Notably, their significant results are only significant at the 10 percent level, with a coefficient of -0.039 and a standard error of 0.024. For those not versed in statistics, this implies that the estimates are rather imprecise. Their dataset includes 25.6 million observations, but they barely reach statistical significance, indicating that the data carries a lot of variability.
The absorption rate for foreign exporters could realistically be anywhere from zero to nine percent based on their findings. Presenting that range as an exact “4 percent” number demands more confidence than the data allows. A 10% significance level doesn’t support a claim of 96% certainty.
Shifting Patterns in Trade
A deeper flaw emerges with how tariffs impact imports. The authors indicate that tariffs led to a decrease in both import value and volume, suggesting an increase of 28 to 33 percent. This is where their methodology falters.
Imagine a simplified scenario. The U.S. imports widgets from China, categorized into three quality levels: budget at $10 per kilogram, mid-tier at $15, and premium at $25.
Now apply a 50 percent tariff. Budget-friendly suppliers suddenly raise their prices to $15 due to customs duties, thereby competing with mid-tier products. Unable to lower prices enough to remain viable, they exit the market. Mid-tier suppliers might drop their prices to $13 but still cede market share, while premium suppliers can afford to lower their prices from $25 to $22 to hold on to less price-sensitive buyers.
Once the situation stabilized, the import mix shifted heavily towards premium products, with average unit prices advancing from $15 to $20.
The Kiel authors observe this trend and conclude that “prices have gone up—proof that tariffs have passed through!”
However, that’s misleading. Trading prices could actually fall at various levels even if the average increases. That’s because lower-cost suppliers vanished from the data. Now, we are comparing premium widgets to what previously included a mix of budget, mid-tier, and premium options. Once tariffs eliminate low-margin products, the remaining goods no longer represent the same quality mix.
Is there considerable underlying variation in their statistical results? This variance arises precisely because different items within a category react differently as lower-cost options disappear while higher-priced ones prevail.
Measurement Challenges
There are multiple issues with this study that compound the central flaw. Tariffs are applied at a very detailed product level, such as HS8 or HS10 classifications, differentiating between frozen boneless beef cuts and bone-in cuts. (HS refers to the Harmonized System for classifying traded items—more detailed classifications mean higher numbers.) The Kiel data only uses HS6, which might encompass just “frozen beef.” In simpler terms, they are measuring prices without accounting for the nuances of tariffs.
This leads to a situation where, if the authors assign a tariff rate to their dataset, they might be inadvertently averaging products facing varying tariffs. Some frozen beef products may incur a 50 percent duty, while others just a 10 percent. This discrepancy errors the price measurement.
The outcome is unsurprising. If the central independent variable is imprecise, the statistical results will likely lean toward showing no effect. And the perceived “no effect” of tariffs on unit prices is interpreted by the authors as a “complete shift to Americans.” Their regression may have been constructed in a manner that downplays the very effects they’re measuring.
Misleading Control Group
The study’s control group is flawed as well. Tariffs lead to trade diversion, with U.S. importers moving from China to places like Vietnam and Mexico. This increase in demand allows non-tariffed exporters to raise their prices or switch to more profitable products. If prices for both tariffed and non-tariffed items rise, the observed difference appears smaller—not because exporters aren’t absorbing costs, but because the comparison is fundamentally skewed.
Consumer Burden Assumption
The authors argue that even if we entertain their findings on import prices, it would be an even larger leap to claim American consumers will ultimately bear the burden of these costs. However, they fail to provide empirical analysis of retail prices, company profit margins, or how these costs are passed down the supply chain.
This occurrence can influence profits, compress margins for wholesalers and retailers, or impact various stakeholders. Claiming it’s a “consumption tax” is more speculation than substantiated evidence.
Questionable Economic Impacts
Beyond their findings, the Kiel authors also noted that imposed tariffs create “dead weight loss,” describing economic waste due to disrupted consumption patterns and supply chain issues. They assert these costs are “pure economic waste, burdens the American populace with no compensatory benefits.” Though alarming, this is based on the premise that production and trade were efficiently allocated before tariffs were implemented.
However, this assumption falls short. If decades of subsidies, state financing, forced tech transfers, and restricted market access from China have already skewed global production, tariffs could potentially reduce these aggregates rather than exacerbate them. Standard welfare economics refers to this as “Second best theory”: rectifying one distortion among many may worsen the overall situation. Conversely, introducing tariffs to counteract foreign subsidies could bring the economy closer to efficiency.
The authors of the Kiel study never clarify whether their “dead weight loss” is indeed valid. If it’s simply the cost of mitigating prior distortions, it could represent an investment toward a more efficient economy rather than pure waste.
Finally, while their assessment shows a narrower focus, it also loses rhetorical confidence. In sea freight transport, even as trade volumes decrease, average unit prices haven’t significantly dropped.
This points to exporters likely maintaining their prices. This could also indicate that price drops are concealed by compositional changes. Or there could be measurement errors distorting the true effect—likely a mix of all three.
What stands is that it hasn’t been established that 96 percent of the cost will fall on the American people. It doesn’t confirm that consumers will face higher prices, nor does it establish that these tariffs represent an “own goal.”
