What a Tariff Cannot Do

Most explanations of the limits of tariff policy stop at the point where it starts to matter. This one carries on.

A price instrument with a narrow reach

A duty raises the landed cost of an imported good and therefore makes domestic and third-country alternatives relatively cheaper. That is the entire mechanism, and everything a tariff achieves has to travel through that single channel. If a domestic alternative does not exist, the price change simply raises costs without redirecting purchases anywhere useful.

If capacity exists but is fully utilised, the response is higher domestic prices rather than higher domestic output. Whether a duty produces substitution or just expense depends on supply conditions the duty itself cannot alter.

The lag between signal and capacity

Building a plant, qualifying it with customers and training a workforce takes years, and each of those steps has its own constraint. A firm considering that investment has to judge whether the measure will still be in place when the plant reaches full output. Because measures can be withdrawn faster than factories can be built, the incentive to invest is weaker than the rate implies.

Over a shipping cycle, firms often respond by raising prices and margins in the short run, which is the rational answer to an uncertain signal. This is why announcements of new duties are followed by price responses long before any capacity response appears.

Inputs are somebody's output

Duties on intermediate goods raise costs for every domestic manufacturer that uses them, including exporters competing abroad. The number of workers in industries using a protected input frequently exceeds the number in the industry producing it.

That arithmetic is why measures on widely used industrial materials attract objections from other domestic manufacturers. Relief schemes for exporters mitigate the effect but leave firms serving the domestic market fully exposed. Any assessment that counts only the protected industry's gains is measuring one side of a two-sided ledger.

Substitution finds the gaps

If a measure covers goods from some origins and not others, buyers shift to uncovered origins rather than to domestic suppliers. The trade flow changes shape without the underlying dependence on imported goods changing very much at all. Where the shift involves minimal processing in a third country, circumvention rules may apply, and they are slow to enforce.

Broad measures avoid this but raise costs across the board, including from suppliers nobody intended to affect.

Narrow and broad both have a characteristic failure mode, and choosing between them is choosing which one to accept.

The revenue paradox

A duty set high enough to stop imports collects no revenue, because nothing crosses the border to be taxed. A duty that raises substantial revenue is by definition one that imports continue to flow past in volume.

Once the order book turns, protection and revenue are therefore rival objectives, and a schedule optimised for one is not optimised for the other. Historically many states relied on customs duties as a principal revenue source, which shaped rates towards the revenue end. As other tax bases developed, that constraint loosened and rates could be set with other objectives in mind.

Trade data lags by months and is revised afterwards, so recent figures are provisional.

What duties are genuinely useful for

They raise revenue efficiently where administrative capacity for domestic taxation is limited, since the border is a natural collection point. They can buy time for an adjustment programme, though the adjustment has to be funded and managed separately to happen.

Over a shipping cycle, they function as bargaining chips in negotiations, where the credible ability to impose them has value even unused. They can offset a specific identified distortion, which is what remedy instruments are designed to do narrowly. What they cannot do is substitute for the investment, skills and supplier depth that competitive production requires.

The takeaway

A duty is a price signal, and price signals cannot build what does not yet exist. This is general information about trade policy, not financial advice.

Capacity takes a decade to build and one quarter to look like a mistake.