The Harmonized System, maintained by the World Customs Organization, is the shared language of trade statistics. Products are grouped into chapters, headings, and subheadings, and the first six digits of a code mean the same thing in every country that uses the system. A researcher who knows the six-digit code for a product can compare trade flows across countries on a like-for-like basis — at least in principle.
In practice, two complications matter. The first is depth: countries extend the six-digit international code with their own digits, often to eight or ten, for tariffs and detailed statistics. Two countries can therefore publish very detailed data that does not line up exactly, and a product that sits neatly in one national extension may straddle two in another. Comparing national figures without checking the code structure underneath is a common cause of confident errors.
The second is change. The classification is revised every few years to keep up with new products and retired ones, and codes move, split, or merge. A time series that crosses a revision without adjustment will show jumps and drops that never happened in the real world. Any report that compares years across a revision should say so, and show what was adjusted.
There is also a simpler trap: a code describes what a shipment was declared as, not what it truly was. Misclassification happens, deliberately and accidentally. That is why a careful report treats HS-based figures as evidence about declared trade, cross-checks them against a second record where the decision matters, and writes down the code it used — so anyone can re-run the same query and get the same answer.