Globalization holds firm at a record level

The phrase "deglobalisation" has been in circulation for most of a decade now, usually accompanied by a photograph of an idle port. It is a satisfying story: the world knitted itself together, then thought better of it, and trade is quietly unwinding. It is also, on the evidence, not what is happening.

What is happening is harder to photograph. Goods are still moving in enormous volume — but they are moving between different partners, over longer distances than at any point on record. Aggregate trade holds up. The map underneath it does not.

Reconfiguration is not decline

When a manufacturer shifts production from one country to a neighbour, the headlines call it deglobalisation. From a freight desk it looks like something else entirely: the same goods, a new lane, a different customs regime and a completely different set of things that can go wrong. The trade did not disappear. It changed address.

The nearshoring story runs into the same problem. Everyone expected a shift from global to regional, and the aggregate data has not shown it: in 2025 traded goods travelled further on average than in any year on record.2 Moving one supplier closer shortens one leg, but the components that supplier needs are still coming from where they always did. A shorter final leg on top of an unchanged upstream chain is not less globalisation. It is more links in the same chain.

Why this matters more than the headline number

Look at two of the numbers above together and the shape becomes obvious. Trade between the United States and China fell from 2.7% of world trade to 2.0%3 — a genuine, substantial decoupling between the two largest economies. And the global level of connectedness did not move.1

Both of those are true at once, which tells you where that trade went. It did not stop. It rerouted through third countries, over longer distances, through more intermediaries and more customs jurisdictions than the direct lane it replaced. That is the single most important sentence on this page for anyone planning freight.

If globalisation were genuinely contracting, the right response would be to shrink with it — fewer lanes, shorter chains, less exposure. That is not the situation. Volume is holding while the routing churns underneath, which calls for the opposite response: more optionality, not less.

A business with one origin, one lane and one customs position is efficient right up until the moment that lane is the one affected by a tariff change, a congested corridor or a supplier relocation. The efficiency was always borrowed against a stable map. The map is what moved.

What we see from the desk

Enquiries for lanes that barely existed a few years ago. Customers asking us to price a second origin they have no intention of using yet, purely so the option exists. More classification work, because moving production changes the legal origin of goods and therefore the duty payable — a consequence that regularly surprises the people who made the sourcing decision.

None of that looks like a world trading less. It looks like a world trading over longer, less direct routes — with a lot of freight professionals re-drawing lanes they had not touched in years. Longer routes are not automatically worse. They are simply harder to plan, and considerably less forgiving of anyone who has not planned them.

Where the numbers come from

Every figure on this page is referenced. If we quote research we did not carry out, it is named here with its publisher and year — you should be able to go and read the original rather than take our summary of it.

A note on other people's research. Where a finding comes from a published index or report — including reports produced by other logistics companies — we name it. Restating someone else's research as your own market view is how a commentary page loses the credibility it was written to build.