Some data science on the newly released Trump and China tariffs

War! The Trump administration announced late last night its next daring attack on China’s unfair trade practices, and it has gone straight for the jugular, announcing that it is putting tariffs on about 1330 goods including “Parts for air conditioning machines” and “Bakery ovens, including biscuit ovens.”

Including biscuit ovens! Without the heat emanating from our precious biscuit ovens, we will be able to avoid using air conditioning when all the A/C parts run out. Brilliant.

There’s even more heat packed into these tariffs though, since Chinese-made flamethrowers were included on the list. No word yet on whether Elon Musk’s Boring Company not a flamethrower flamethrower will be counted as a flamethrower (or made in China for that matter).

Of course, the Chinese have responded with their own new tariffs list (a list different from the new tariffs the country announced on Monday). The new new tariffs includes soy, cars, and other goods from America.

Okay, I admit it: tariffs are really boring. The tit-for-tat attacks are like a really boring game of Battleship. Actually, that’s redundant — it’s just a game of Battleship, so I am going to try to spice things up with some data science.

The modern economy is complicated. What exactly is a “biscuit oven” anyway? How does the government know when to apply a tariff to a good and when not to? To solve these dilemmas, the U.S. government invented a mechanism called the Harmonized Tariff Schedule which has categories for every product in the world organized into chapters numbered 1 through 99.

Some chapters, like chapter 86 (“Railway or tramway locomotives, rolling-stock and parts thereof; [blah blah blah]”) have only a couple of dozen categories, while chapter 84 (“Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof,” i.e. the cool stuff) has by my count more than 2000 categories.

I wanted to look at the 1333 categories the Trump administration picked to understand what goods seemed to draw the most attention from the trade team. Unfortunately, the list of categories was released as a PDF (thanks government!), so I had to perform several conversions to get the raw data. The entire Harmonized Tariff Schedule is available in a machine-readable format through

Using Python, I performed counts of the various HTS numbers to compare the number of categories available under the schedule with the number of categories that the Trump administration proposed for tariffs. This is a pretty rough analysis, since the categories are administrative and not economic (in other words, some categories could be worth billions of dollars while others are much less valuable). That said, this analysis can still give us a sense of where the administration focused on.

Two categories were hit hard by the tariffs – Chapter 30 pharmaceuticals and Chapter 86 locomotive parts. Both had more than a third of their categories added to the tariffs list with China, far above the percentage of other categories. From there, precision equipment and machinery (Chapters 90 and 84 respectively) were hit, covering roughly a quarter of products each. My complete analysis list is included below.

One hypothesis I would take from these tariffs is that they could actually be quite a bit more punitive than first meets the eye. While the tariffs are being applied to roughly $50 billion worth of goods, the goods appear to have been specifically chosen to encompass a range of parts and machines required in manufacturing.

For instance, imagine that your “Instrument panel clocks” suddenly got more expensive, so you have to source clocks from a new vendor in another country. If you have to find new sources for enough parts, you might just consider adjusting your entire supply chain in the process, even for parts that didn’t come under the new tariffs regime. It’s this second-order economic effect that I think will be important to pay attention to.

No one should read the Harmonized Tariffs Schedule, but these categories do matter for the economy, and companies are going to be racing to understand the decisions made by the administration and what it means for them. As typical with proposed rules, there is a public comment period, so expect heavy lobbying from companies to get certain items off the list (or even maybe on the list in order to drive away cheaper competitors).

Data on Trump tariffs

Ordered by percentage:

  • Chapter 30: 47 of 129 categories (36.43%) (“Pharmaceutical products”)
  • Chapter 86: 17 of 47 categories (36.17%) (“Railway or tramway locomotives, rolling-stock and parts thereof; railway or tramway track fixtures and fittings and parts thereof; mechanical (including electro-mechanical) traffic signalling equipment of all kinds”)
  • Chapter 90: 164 of 569 categories (28.82%) (“Optical, photographic, cinematographic, measuring, checking, precision, medical or surgical instruments and apparatus; parts and accessories thereof”)
  • Chapter 84: 537 of 2173 categories (24.71%) (“Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof”)
  • Chapter 88: 16 of 72 categories (22.22%) (“Aircraft, spacecraft, and parts thereof”)
  • Chapter 89: 11 of 54 categories (20.37%) (“Ships, boats and floating structures”)
  • Chapter 85: 241 of 1222 categories (19.72%) (“Electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles”)
  • Chapter 93: 15 of 93 categories (16.13%) (“Arms and ammunition; parts and accessories thereof”)
  • Chapter 76: 27 of 176 categories (15.34%) (“Arms and ammunition; parts and accessories thereof”)
  • Chapter 72: 108 of 731 categories (14.77%) (“Iron and steel”)
  • Chapter 87: 48 of 470 categories (10.21%) (“Vehicles other than railway or tramway rolling stock, and parts and accessories thereof”)
  • Chapter 73: 44 of 746 categories (5.90%) (“Vehicles other than railway or tramway rolling stock, and parts and accessories thereof”)
  • Chapter 40: 8 of 272 categories (2.94%) (“Vehicles other than railway or tramway rolling stock, and parts and accessories thereof”)
  • Chapter 29: 38 of 1409 categories (2.70%) (“Vehicles other than railway or tramway rolling stock, and parts and accessories thereof”)
  • Chapter 94: 5 of 310 categories (1.61%) (“Furniture; bedding, mattresses, mattress supports, cushions and similar stuffed furnishings; lamps and lighting fittings, not elsewhere specified or included; illuminated sign illuminated nameplates and the like; prefabricated buildings”)
  • Chapter 28: 4 of 430 categories (0.93%) (“Inorganic chemicals; organic or inorgani c compounds of precious metals, of rare-earth metals,of radioactive elements or of isotopes”)
  • Chapter 83: 1 of 117 categories (0.85%) (“Miscellaneous articles of base metal”)
  • Chapter 91: 1 of 230 categories (0.43%) (“Clocks and watches and parts thereof”)
  • Chapter 38: 1 of 250 categories (0.40%) (“Miscellaneous chemical products”)