Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


When The Trade Data Does Not Add Up

by Brad Setser
October 31, 2016


This is not a post about China, or the various discrepancies in the Chinese data.

It is about a rather puzzling thing that I only noticed as a result of the Brexit debate.

The bulk of the UK’s surplus in services come from trade with non-EU countries (services exports to the EU are large, but so are imports—Tuscan and French vacations?). See this chart (h/t Toby Nangle).

A big part of the non-EU surplus in services comes from the United States. In 2015, the UK reported a 27 billion GBP (just over $40 billion) surplus in services trade with the U.S. and an overall surplus in goods and services with the United States.

The funny thing? The U.S. also thinks it runs a surplus in services trade with the UK. A $14 billion surplus in 2015, for example.


It is pretty hard to square those two data points. UK data is from the Office of National Statistics’ Pink Book, U.S. data is from the Bureau of Economic Analysis (BEA), table 1.3 of the “International Transactions” data set.

It turns out that the U.S. thinks it sells more services to the UK than the UK thinks it buys:


And the UK thinks it sells more services to the U.S. than the U.S. thinks it buys.


My guess is that such discrepancies are actually common in the services trade numbers. Goods trade is calculated by customs bureaus. Lots of the numbers on services trade come from surveys, estimates, and the like.


  • Posted by Non-Economist

    The obvious question is, does the US and UK use the same definition of services? There is a Harmonized Tariff Schedule for goods, but is there one for services?

  • Posted by Douglas Lee

    Usually, it is the other way. Notice China thinks they import much more from the US than we think we export to China. The problem is imports are often taxed, exports are not. Custom collectors are much more careful in tracking imports than exports. This is the reason we stopped trying to track exports to Canada and Mexico years ago — we just substitute what they say they import for our exports. The problem has grown worse as the internet has made it easier for small businesses to sell abroad.

  • Posted by Brad Setser

    non-economist; good question. i think the definitions are similar but would need to delve more. also a question of what in practice is measured in each country’s surveys on services trade.

    Douglas Lee. US/ China is issue is complicated by HK. US also thinks it imports more from China than china thinks it exports to the US; that though is likely more of a how is HK treated than a tax issue. But your point stands.

    I suspect the opposite though applies with services trade for goods deficit countries; one argument that the stats guys can make for spending more on new surveys and the like is that the new survey would show more services exports, and thus a smaller goods and services deficit. so in the UK and the US i wouldn’t be surprised if more effort in practice has gone into counting service exports than imports. A hunch.

    i do wonder how much effort has gone into reconciling the numbers across countries

  • Posted by Douglas Lee

    Sometimes accounting identities can be helpful. We know the current account deficit should equal the financial account deficit which should also equal the saving/investment deficit. When you compare these for the US, the current account deficit is substantially larger than the other two. The other two are roughly comparable although the financial data is highly volatile. If you suppose the current account is the one with the largest error, under counting exports seems quite likely. This would also help explain why GDI is consistently larger than GDP.

    Reconciling numbers across countries would be enormously helpful.