UK Trade figures are subject to a lot more doubt than we are usually told

by Shaun Richards

Today is a cleat example of the morning after the night before so let me open by congratulating Italy on their victory in the European Championships football last night. However I wish to look back to Friday lunchtime when the UK released some new details on trade some 5 hours after they were supposed to be released. A good day to bury bad news or has the issue of there being rather different numbers on European trade being produced by the UK and EU come to a head?

What were the numbers for May?

They started well as we note a rise in exports.

Total exports of goods, excluding precious metals, increased by £1.3 billion (4.9%) in May 2021, driven by a £1.0 billion (8.0%) increase in exports to EU countries.

The first part is no great surprise as we are looking at a period where economies picked up and the second part of a rise to the EU is hopeful for the post Brexit era. As it happens the other side of the balance sheet was good for the trade balance too.

Total imports of goods, excluding precious metals, fell by £0.5 billion (1.4%) in May 2021 because of a £0.7 billion (3.4%) fall in imports from non-EU countries, which offset a slight increase of £0.1 billion (0.8%) to EU countries.

So we imported less in spite of the economy growing by 0.8% in May and it was a non-EU issue. Indeed by our admittedly poor standards we seem to be in a better run for the trade data.

In the three months to May 2021, the total trade deficit, excluding precious metals, narrowed by £2.2 billion to £3.5 billion.

If we now switch to the Brexit issue we see that there has been a change.

Monthly goods imports from non-EU countries, excluding precious metals, continues to be higher than the EU for the fifth consecutive month, but the gap is narrowing.

But that was it and as someone who has formally asked for more detail on services trade it is hard not to have a wry smile. Because we have ended up with it being nearly the same as the detail on goods trade where we learn very little about either.

In the three months to May 2021, the trade in services surplus fell by £0.2 billion to £28.1 billion.

If you were hoping to find out if it was exports or imports changing I am afraid that was it.

Services Trade

I thought I would scan the data and I return to the issue I raised with the Bean Review. In each of the last 3 sets of 3 months data we have a surplus of around £28 billion. So in a world with the changes we have seen the 3 months to May gives the same answer as the 3 months to February and the 3 months to November 2020. Does anyone actually believe that?

There is one change in that services trade has fallen compared to the peaks in 2019 with exports some £15 billion per quarter lower and imports £18 billion lower.

Why did we get so little May data?

This was singing along with Lyndsey Buckingham.

I think I’m in trouble
I think I’m in trouble

Here are the details and nice effort to blame HMRC ( the tax body).

We identified an error in the UK trade data prior to the planned release on 9 July 2021. The error occurred as we opened up the 2020 year to take on board corrections to HMRC data for non-EU trade which has caused a processing issue for the EU data.

With all the debate over Brexit it was in fact trade outside of it which saw the big move with exports in 2020 revised down by £4.5 billion.

Ironically the numbers improved the view of post full Brexit trade. The moves were more minor but January and February saw a downwards imports revision of £500 million. Then March and April saw an upwards revision of £600 million to exports.

We are primarily funded by readers. Please subscribe and donate to support us!

Oh and the HMRC changes were published on the 29th of June so in plenty of time…

Brexit Trade Number Crunching

Regular readers will be aware that some time back I looked at the figures for trade between the US and China. The problem was that you git rather different answers depending on whose figures you looked at. This has been repeated in 2021 as we look for signs as to what a full Brexit has done to UK trade with the EU. Here is the Office for National Statistics.

We have been exploring several possible reasons for the growing disparities between EU and UK trade statistics.

So what is it?

A key reason for the differences we have found is that Trade statistics for imports can be reported on a country of dispatch and country of origin basis. Whereas for exports only country of destination is recorded.

Seems a bit odd that you would count exports and imports differently. So who does what?

For the UK, our main publications for trade statistics shows imports based on country of dispatch and exports on country of destination. This is both for EU and non-EU trade.
The European Union and its Member States will record the country of dispatch and country of origin for non-EU import but will publish their non-EU imports trade based on country of origin.

The finger is pointed at Eurostat as this pre Brexit.

Eurostat statistics record this transaction as £200 imports from UK to NL.

Now becomes this for the same transaction.

Eurostat statistics record this transaction as £200 imports from China to NL.

What does this mean in terms of numbers? Well this was the state of play.

 Just looking at 2020 as an example, the difference between the ONS and Eurostat non-seasonally adjusted datasets were as little as under £100 million in some months, while reaching around £1.5 billion in others.

Which has been replaced by this.

Revisiting the differences now between the UK and Eurostat non-seasonally adjusted datasets, the divergences are much larger, up to £2.7 billion in some months

The areas especially affected are these.

The commodities driving this increased divergence are ‘machinery & transport equipment’ and ‘miscellaneous manufactures’. While a large volume of goods for these commodities, for example cars being finished, pass through the UK, the importing European country receiving the finished good will report country of origin and not country of dispatch.


This has been a humbling period for official economic statistics. Indeed with the way the the Markit PMIs have got manufacturing so wrong recently we can widen the troubled list. Trade figures have long been amongst the worst and I recall the UK version losing their National Statistics status back in 2014.

This adds to the fact that there has been trouble elsewhere like in the Labour Force Survey.

In addition, a change to the non-response bias means that we appear to have a poorer response from some subsets of the population, in particular those with a non-UK country of birth.

which means this.

Another a distinct but related issue was the fact that the population estimates that feed into published the LFS statistics predate the pandemic, and so do not show its demographic and structural impacts.

Which reminds me of this problem.

Our latest data, using information from the Annual Population Survey (APS), shows that in mid-2020 there were around 3.5 million EU citizens living in the UK, a lot smaller than the 6 million applications for the EU Settlement Scheme.

Are there 6 million? No I do not think so but the 3.5 million is probably wrong too.



Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.