Readers, it is difficult to exaggerate how incredibly bad the “Day of Liberation” scheme is. Top to bottom, it’s inconsistent. When the plan was approved on April 2nd, when Donald Trump was lifting a committee of seemingly random numbers, economists had to schoolbull to figure out exactly what these numbers came from. They appeared to be completely divorced from reality and had nothing to do with reciprocity. People thought that the number was that the bilateral trade deficit was divided by imports into the country. This led the USTR to deny re-arsease by asserting the current calculation. Sumhhow, that was worse than what people thought.
You would think that the “mutual” tariff model includes tariffs from other countries. Certainly, there is already established methodology in US law, and there is an economic theory in determining whether there should be mutual tariffs that face unfair trading practices. Section 301 of the Trade Act of 1974 provides for various remedies, similar to the Mutual Customs Act. The offset obligation calculation already exists. The service of this new model is not the goal of reciprocity or even amending “unfair” trade practices.
Rather, this model treats the bilateral trade deficit as evidence of unfair practices in itself. It can be argued that excessive trade deficits are bad (but still, because such arguments are conditional and not because they are themselves, but because they have become even weaker if the country is an international reserve currency). Bilateral though? It’s definitely not. There is no reason to think that trade between any two partners is equally balanced. We don’t live in a barter economy. The key to money is to promote imbalances in bilateral trade. I have an surplus with my employer. So it’s not evidence of myself that I’m torn them apart. Similarly, I have a trade deficit with GRCER, butchers, brewers and more. That’s not evidence that Lusa’s supermarket, bourgeois meat market, and Abitia Brewing are torn me apart. If that was about to seek money, we were to balance bilateral trade. I had to find exactly what excitement wanted in my daily meals (I am happy that I don’t want to study economics). Therefore, the overall modeling premise misinterprets the VRY foundation of financial and economic exchange.
But for discussion, let’s assume that the model’s premise is valid. Let’s take a look at the model itself. USTR reports the model as the balance of trade with a particular country divided by imports with a pass-through (φ) times the imports. This Mel means nothing. It hides this meaninglessness behind the Greek letters, but this model has no interpretable meaning. It is not clear that even doing what the author wants it to do. Readers, you will not find this model in economics textbooks or papers. No different reports on model logic have been released, at least at the time of this writing. Therefore, there is no apparent reason for Cousa to think that the model itself has something to do with the premise, even if it is valid.
But for discussion, let’s assume that the model is valid. The selected model parameters are illogical. For ε and φ, USTR was chosen the same value for each country. However, ASE is not a reason to assume that ε and φ are the same for each country or for each country’s interests. Both ε and φ depend on exchange-specific factors. For example, goods with many alternatives, ε is higher (or lower if there are most of the alternatives in the goods). Similarly, this prevents the calculated tariff rates are likely to be incorrect. It may be too expensive and too low.
But for the sake of discussion, let’s assume that all countries in the world have exactly the same ε and φ. The number selected for the first parameter is incorrect. The author states:
Recent evidence suggests that elasticity is near two in the long run (Boehm et al., 2023), but estimates of elasticity differ. To be conservative, studies that found high elasticity near 3-4 (e.g., Broda and Weinstein 2006; Simonovska and Waugh 2014; Soderbery 2018) were drawn.
These are the estimation of the subject. However, the statement ε of 4 is “conservative” and not news. Many studies find that ε is above 5-7, especially after the trade shock (see, for example, here). Furthermore, they simply set φ to 0.25. *No quotes. The only justification given is:
Recent experience with US tariffs in China has demonstrated that Taiff Passthrowr is lower from retail prices (Cavallo et al, 2021). [link added]
However, retail prices are not a relief measure here. A total passthrough is required. Here’s what Alberto Cavallo and others are actually saying (additional emphasis):
At the border, the passthrough of import duties is more Heiger than the passthrough of exchange rates. Chinese exporters did not significantly lower the dollar’s prices. In contrast, the US has cut prices significantly, unhappy with foreign retaliatory tariffs. Popularity suggests that the price impact is more limited, reducing retail margins. Our results mean that incheon of tariffs, so far, you have fallen in the long part of our company.
The author of the report accurately retreats Cavallo et al. Rather than indicating a low pass-through, it actually shows almost complete pass-through, indicating that tariff thru is being borne by Americans. This is a scary case of cherry picking by USTR. Oh, by the way, studies show that φ is closer to 0.8 instead of 0.25. Both the brain and φ are underestimated, indicating that tariff calculations are systematically too high.
But for discussion, let’s assume that their ε and φ chews are accurate. You’ll reach the real kicker. The author writes:
“Assuming it offsets the effects of exchange rates and general balance, it’s a small engaging thing to ignore…”
This assumption is hage. If the assumption is not kept, the entire model will fall apart. The problem is: The overall point of tariffs is that they have a general balancing effect with the exchange rate. They have said it quite multiple times in their report, and so has said it, too, as well as the Trump administration’s economic adviser. Therefore, the assumptions that require the model’s core are never held. In other words, the whole thing is a two-stage ab initio. Certainly, not 24 hours after the Triff scheme was released, the stock market was a two-day record and a dollar decommissioned. I don’t think you’ve seen the model wrong. 24 hours must be a sub-sum of the record. It took months for the Covid lockdown model to explode, even the bad model just like them.
Repeat again. It’s hard to exaggerate how bad this model is. Starting from the end, it is inconsistent, with no bad assumptions, bad parameters, and logic. This is not the result of rational thinking. It is a shameful exhibition of science.
*By the way, this is why people initially thought that the model was merely divided into imports. For imports with ε=4 and φ=0.25 and denominator ε*φ*, ε*φ=1, and the entire denominator decreases to imports only. The fact that UST had not seen it is a problem.
