The lie that travelled around the world

Mark Twain once said that a lie can travel halfway round the world before the truth has got its boots on. This is certainly true of the Aston Business School’s latest ‘Research paper’ entitled: Unbound: UK Trade post-Brexit.

Incredibly the BBC[1], The Guardian, the Financial Times, the Independent and even CNN decided to greatly inflate the reach of a paper by a little known Birmingham university by reporting on effectively one paragraph of the report and imagining, or at least encouraging their readers to imagine, that this paragraph fairly represented the totality of the report. Had the journalists at these once reputable media institutions actually read the report, they would have known that the report’s conclusions were not what they assumed.

Had they even bothered to read the report’s executive summary, they would have noticed that one of its two conclusions was that the UK should: Enhance its trade infrastructure; strengthen its domestic supply chains; diversify its global sourcing; and foster strategic partnerships with countries outside the EU. None of this would be necessary if the report had concluded, as the journalists’ headlines implied, that the UK should rejoin the EU. Most of the UK’s trade with the EU is adequately handled by the port and tunnel infrastructure that we have now and by the customs automation that HMRC has developed.

But much as I agree with this second conclusionof the report, unfortunately, having read the whole report, I cannot say that it proves any of its policy implications especially its first conclusion that the UK needs to ‘mitigate the adverse effects of the TCA with sector specific negotiations’.  No amount of sector specific negotiations will turn almonds imported from California or cotton yarn imported from India into a UK export to the EU in compliance with the Rules of Origin in the UK EU Trade and Cooperation Agreement (TCA).

The report is full of what can only be described as rookie errors as well as some serious mistakes. The authors appear to be unaware that the Harmonised System (HS) of tariff codes were changed for several of the UK’s largest export sectors, so their results about 1645 fewer traded goods varieties is what used to be known as GIGO (Garbage In, Garbage Out). I could only find 650 6 digit tariff codes that recorded less than £1000 worth of exports to the EU in 2023. I reviewed them all and found almost all of them were caused by either a change in the HS code, the TCA rules of origin, or were simply intermittently traded goods. There is more detail on this in section 5 of this paper. But in short, you cannot analyse trade using statistics without also understanding all of the nuances, changes and possible flaws in those statistics.

Reading the Aston paper, it is also apparent that the authors, including a professor and an associate professor, know nothing about: UK trade, manufacturing supply chains, the rules of origin of the TCA, the lack of accuracy in Intrastat data or even how goods are distributed around the EU. They couldn’t even find a full list of trade figures for all countries even though the International Trade Centre publishes such a list for all countries including British Overseas Territories and has 20 years of data. Instead, they used something called Trade Data Monitor for 91 countries between 2017 and 2023.

So, what else did Aston get wrong?

The report opted to evaluate UK post Brexit trade in two ways: by value and by the variety of goods traded between 2019 and 2023. This is an unusual choice; most economists insist on using deflated trade data such as Chained Volume Measures or, when possible, actual volume or units traded.  Aston decided to measure the varieties of goods traded using 6 digit HS code data (see above), but this requires a great deal of product knowledge that the Aston team don’t appear to have.

1. Size matters

Intriguingly, Aston assumes that the UK is able to trade all products, at all times and in all quantities. So, they reach some hilarious conclusions that the BBC or the Financial Times should have spotted. For example, the report is more impressed by the UK’s increase in export varieties of Tobacco products than it is by the smaller increases in UK exports of Railway and Aircraft goods. And the report mourns the demise of the UK’s export varieties of Silk, Cork, Wood pulp and Vegetable plaiting material.

I am sure that readers of Briefings for Britain know that the UK has never grown tobacco, despite being home to two of the world’s largest tobacco companies, and that HS88 Aircraft is one of the UK’s largest export sectors, and it doesn’t even include jet engines, another massive UK export, which are counted in HS84 Machinery. Together aircraft related exports make up about  20% of all UK goods exports by value. A small increase in this sector is worth a lot to the UK economy even if it doesn’t impress Aston.

Tobacco product exports, on the other hand, were worth 0.03% of UK goods exports in 2023, as they were in 2017. If you are wondering how we even managed to export that much of tobacco products: these exports were almost all vaping products (HS2404: products intended for inhalation without combustion). But before you imagine that this is an actual UK export, vaping products are also the UK’s largest tobacco product import. We imported 33,000 tonnes of vaping products in 2023, mainly from China, and we exported just 1,835 tonnes, of which about a third went to Ireland.

If the Aston professors had thought to check UK imports of any particular good as well as UK exports, before making their ridiculous pronouncements, they might have avoided some of these embarrassing mistakes and not have been at all surprised about lower silk, cork or wood pulp exports.

They should also have checked if the products they were so concerned about even matter to the UK economy. They claim that HS14, which they call ‘Vegetable products’ when it is actually Vegetable plaiting material, has ‘experienced the most significant decline in exported varieties… dropping by 68.5%’[2]. In 2023 the UK exported a mere £2.5 million pounds worth of Vegetable plaiting material, up from £2 million in 2019. There are only 11 six digit tariff codes in this group and the UK has never exported goods in 7 of them. There has not been a 68% drop in export varieties in this sector. Aston appears to have made that up. The UK does not export kapok, eelgrass, broomcorn or piassava. While the small amount of rattan, bamboo, raffia and cotton lint plaiting material that the UK did export would have been imported first. We imported about £19 million worth of Vegetable plaiting material in 2023, of which £2.5 million was exported – it mostly to Ireland.

2. Trade distribution and re-exports

This brings me to another point: Ireland doesn’t have any ports that can unload large container ships. The UK’s ports provide this facility for Ireland, so a disproportionate amount of trade between the UK and Ireland is simply redistribution of goods manufactured in China and landed in the UK with some repackaging or other value added in the UK to make it a UK export.

In general,  some additional treatment must be added to imports in the UK, for them to be counted as a UK export under the Rules of Origin in the UK EU Trade and Cooperation Agreement (TCA). The TCA Rules of Origin vary by 4 digit tariff code: raw agricultural produce must be grown in the country of origin, while clothing must be manufactured in the country of origin but can use imported materials. So, while the Aston Professors were worried by the 73% fall in UK exports of (HS08) Fruit and Nuts, or the fall in UK exports of shoes and textiles, they shouldn’t be. Many of these goods were never UK exports, they were simply re-exports. But also, the beneficiaries of the way the EU counts intra EU trade.

The professors at Aston also misunderstand the Rotterdam effect. It doesn’t just refer to goods landed in Antwerp and Rotterdam: goods landed in any large EU port and then re-exported within the EU could cause it, including goods landed in UK ports that were then reexported to Ireland and miscounted as UK exports.

Since Brexit we have seen an increase in exports to both Belgium and the Netherlands not because of the Rotterdam effect but because many UK goods from SME producers are being distributed within the EU by Dutch or Belgian companies. This is especially obvious in UK pharmaceutical exports that are distributed from Belgium.

EU road transport also distorted UK trade data with smaller EU countries before Brexit. For example, Luxembourg has doubled its UK vehicle imports since Brexit. Before Brexit, consignments of UK vehicle exports destined for Luxembourg were probably forwarded from larger consignments of UK vehicles destined for France, Belgium or the Netherlands and not counted by Intrastat as UK exports to Luxembourg.

3. Data accuracy changes everything

Even though the paper discussed the data the authors used, the paper only focused on the limited number of countries covered by their data and appeared to assume that the data itself was accurate. UK trade data is now collected by HMRC for UK exports and imports to all destinations including EU countries. But before Brexit, HMRC collected data for goods landed in the UK from outside the EU but UK trade with other EU nations was gathered by Intrastat surveys and was not accurate. Therefore, comparing pre-Brexit trade with post Brexit trade requires a cautious approach and careful sector analysis.

The EU collects very accurate data on its non-EU imports, because tariffs on imports are a large source of revenue for Brussels. But once the goods have been landed, without border checks or paperwork, it is hard for anyone to know exactly where the goods came from or where they ended up. Since Brexit, the UK has begun to see the many inaccuracies in Intrastat’s data. From pineapple juice from Costa Rica previously recorded as an import from the Netherlands to sports shoes from Vietnam previously recorded as imports from Germany.

I suspect it was in the EU’s interest to ignore Intrastat’s double counting of goods in transit as this exaggerates the importance of intra EU trade. It also made EU politicians and the world’s academics believe that the EU was ‘working’. But the TCA’s Rules of Origin have exposed Intrastat’s miscounting of UK trade, and we should assume there is a similar level of miscounting for all other EU countries.

4. The Rules of Origin exclude whole product lines

All trade agreements have rules of origin in order to define which goods can be counted as being legitimate products of the countries that are party to the trade agreement. The Aston Professors believe that the TCA Rules of Origin are simply a cost to trade as importers must be able to prove that the good complies with the rules and can be imported tariff and quota free. But for many goods, the rules of origin disallow them completely from being counted as UK or EU goods. UK companies still sell goods to the EU, but if no value is added in the UK, then the goods will either be shipped directly from their country of manufacture to a distributor in the EU or via a free port or bonded warehouse in the UK.  This it what has happened to UK ‘exports’ of fruit and nuts, shoes, clothing, textiles and many other goods. Goods no longer counted as UK goods under the TCA appear to lower both the total value of UK exports to the EU and the number of varieties of goods exported.

But it is not just the professors at Aston that don’t know this. Incredibly the BBC ran an article on their website written by their Economics Editor, Faisal Islam, claiming that the UK’s nut exports had fallen by 70%. He would appear to know less about UK Agriculture than he knows about trade. The UK can’t grow almonds, Brazil nuts, cashews, coconuts, pistachios or walnuts. Worse, he doubled down the next day, replacing his ‘nut’ article with another article highlighting the Aston paper, but this time using UK popcorn exports as an example. For the record, the UK grows very little corn, I can’t tell you how little because it is so insignificant that DEFRA doesn’t even bother to count it. But the UK does import over 2 million tonnes of corn each year. Its biggest supplier is now Canada, but it used to be Ukraine before the Russian invasion. Importers of corn from either country would have had to fill in trade forms. But according to the BBC, a UK popcorn manufacturer is upset that he needs to fill in forms to export his products to the EU. Lucky for him, the suppliers of his basic ingredient, corn, were not deterred by trade compliance.

As for the UK’s footwear exports: Yes, they have fallen by over 60% since Brexit but this is also due to the TCA Rules of Origin, that require shoes to be produced in the UK or EU, but they can use imported materials. As most of the UK’s high street brand shoes are made outside the UK, the largest falls in UK shoe exports have been to EU countries such as Germany (-86%), France (-80%), Spain (-74%), and Ireland(-59%). I assume weak economies and lower consumer spending in most EU countries will have also played its part. However, UK shoe exports increased to many countries outside the EU such as South Korea (+35%), Switzerland (+69%), the UAE (+32%) and Macao, China (+149%).

European clothing and shoe brands moved much of their manufacturing to Asia in the 1990s. This is especially true for high-street brands producing affordable clothing. This has been a great benefit for UK consumers and for European fashion companies, but it has now created very misleading post-Brexit trade statistics for clothing, footwear and textiles. The effect the Rules of Origin had on UK EU trade statistics was apparent in the first year of Post Brexit trade data. It is strange that the Ashton paper still doesn’t understand it.

5. Changes in the HS code numbers does not mean trade stopped

But this was not the Aston Professors greatest and most unacceptable mistake. Incredibly they claim that the UK no longer exports 1,645 product varieties measured by 6 digit HS code, a fall of 33.5%[3]. However using full ITC trade data, I cannot verify this claim, so I have trouble believing that their smaller TDM data set shows this decrease. But either way, they have misinterpreted what they thought they were seeing.

According to ITC data, the UK exports about 5,800 product varieties measured at 6-digit HS code  and in 2023 the UK did register zero exports in 650 of these product codes. However, most of these were because the tariff code had been changed. For example, the UK still exports Under carriage gear for aircraft but now it is recorded as  HS880720  rather than as HS880320. Exports of the old product line, HS 880320, were zero not just for UK exports to the EU, but also zero for EU imports from any country and for UK exports to non-EU countries. Did the professors at Aston not think to investigate whether one of the UK’s larger export goods had really suddenly disappeared? The world’s planes still need landing gear. But the professors didn’t check, instead they claimed that the UK was being cut out of the EU’s supply chains.

In 2021, the HS codes were changed for some products in: HS88 Aircraft, spacecraft and parts; HS85 Electrical machinery and equipment; HS30 Pharmaceutical products (including vaccines); HS38 Miscellaneous chemical products, HS90 Precision equipment; HS94 Furniture and lighting; HS97 Art and antiques; HS57 Floor coverings; HS89 Ships and boats; HS87 Vehicles; HS 29 Organic chemicals; HS62 Clothing; HS 68 carbon fibres; HS 70 glass. I could go on, but you get the picture and hopefully recognise that these are some of the UK’s largest export sectors.

More than half of the UK’s exports come from just 5 out of the 98 2 digit HS code product sectors[4],  while 75% of all UK exports came from just 10 product sectors[5].   As six of these 10 product sectors have had changes to the 6-digit HS codes used to identify goods in the sector, no wonder the Aston professors thought they had identified a large fall in the UK’s export varieties. A modicum of common sense should have made them question their results.

A modicum of common sense should also have alerted the BBC and the Financial Times that there was a problem with Aston’s data. Instead, still suffering from Brexit Derangement Syndrome, the usual suspects seized on this information and amplified it, without questioning whether it was true.

6. Intermittent goods exports

Of course, not all of the 650 goods varieties showing exports of zero or less than $1000 in 2023 were due to a change in HS code. Many were simply goods that the UK usually exports in very small quantities and often intermittently. By setting their bar as low as $1000 of exports, the Aston professors managed to catch many goods that the UK has either never exported to the EU or only exports occasionally. My favourite occasional example, was a stray £1000 worth of frozen camel meat, exported to the EU in 2021. I have no idea where this camel meat came from, but I can imagine the excitement  at Aston’s econometrics department when yet another example of a ‘lost export variety’ appeared in their data set. Inexplicably they assumed that all HS codes are equally important to UK trade, and that any product line that was not exported to the EU in 2023 was due to the ‘adverse effects of the TCA’. This just isn’t true.

I have reviewed the 650 6 digit tariff codes that recorded less than £1000 worth of exports to the EU in 2023 line by line and found very few that were NOT caused by either the TCA rules of origin, a change in the HS code or simply have always been intermittently traded goods. I found only 34 export varieties that could require further investigation, and they were all either fish or chemicals.

Most of these specific chemical product lines were not large UK exports to the EU before Brexit so our trade authorities should not be rushing to ‘fix’ this. While some chemical exports, such as Acrylonitrile, used in the manufacture of acrylic fibres, have been falling since 2010. This is not due to Brexit. We now import acrylonitrile from Germany. A quick search on the internet tells me that INEOS announced it would close the UK’s acrylonitrile plant in October 2019 due to safety concerns. So, this is another example of a ‘lost export variety’ that should be removed from the Aston data. The UK can’t export goods it doesn’t make and no amount of tinkering with the TCA agreement or the REACH regulations will change this.

As for UK fish exports, that is more difficult. There are some obvious UK re-exports of fish imported from non-EU countries such as abalone, catfish, and Pacific salmon. While other ‘lost export varieties’ may simply reflect a change of taste. For example, UK oyster exports to the EU in 2023 were back to their 2019 tonnage. But the EU bought no HS 030719 smoked, dried or salted oysters but it did buy more HS 030731 fresh oysters. Smoked oysters haven’t been a large export for the UK for years, yet this ‘zero exports’ of smoked oysters in 2023 is one of the lost export varieties being used by Aston to pretend that the UK is exporting fewer product varieties to the EU due to the ‘adverse effects of the TCA’. Again, tinkering with the TCA will not force consumers in the EU to eat more smoked oysters.

It is also worth noting that Spain is the world’s largest exporter of smoked oysters, and their average export price was cheaper than the UK’s smoked oyster export price last time we exported any. So even if there is a revival of the fashion for smoked oysters, it is unlikely that the UK will be supplying them to the EU. Surprisingly, the Aston paper was not interested in researching the effect that relative price has on export quantities of any commodity.

7. Comparing reality with expectations

Aston claims that UK export values to the EU were 27% lower and import values from the EU were 32% lower between 2021 and 2023, compared to a counterfactual scenario without Brexit. This was of course, the paragraph that attracted the attention of the world’s media. Aston gives no information about its counterfactual, but we do know that Aston is comparing HMRC exact data for UK trade post-Brexit with projections made using inexact Intrastat pre-Brexit trade data for the limited number of countries covered by its data base.

But using actual export values rather than comparisons to a counterfactual, total UK export values to the EU increased by 19% between 2021 and 2023[6], compared to a 23% increase in UK exports with non-EU countries. I have trouble believing that UK exports to  EU countries would have been 27% higher had the UK remained in the EU. The EU’s proportion of UK goods exports has been falling steadily from 58% in 2004, to 47% in 2014, and only 42.5% in 2023. It is absurd to imagine that this trend would have changed if the UK had remained an EU member.

The EU is becoming a smaller and smaller part of the global economy. Both the US and the developing world are growing much faster than the EU. UK exports to the EU have increased by 57% in the 20 years since 2004, but UK exports to the US increased by 99% over the same period and increased by more than 1000% to China. No amount of tinkering with the TCA will change this regardless of the econometric contortions produced by Aston.

Trade is determined by the buyer. Importers initiate trade and pay the costs, insurance and freight. Exporters don’t export goods on spec in the hope someone will pay for them after they have been landed. Germany was always the UKs largest market in the EU, buying 22% of UK exports to the EU in 2019. This has now dropped to only 18% of UK exports to the EU. This is probably because the German economy has been suffering since it lost its access to cheap Russian gas. When Germany sneezes the rest of Europe catches  cold. The Aston paper notes the lower levels of trade in consumer goods between the UK and the EU but doesn’t associate this with struggling economies, higher interest rates or heating costs. Instead, lower exports of UK consumer goods are assumed to be more proof that the TCA needs adjustment.

8. Outliers and unusual data

The Aston paper seems to misinterpret any data that supports its case without looking into the details. For example, the Aston professors chose to leave Luxembourg out of their graphs, claiming it was a small country outlier even though they happily included Malta whose populations is about the same as Luxembourg. However, unlike Luxembourg, Malta appeared to conform with their expectations of falling UK trade, while UK exports to Luxembourg have increased since Brexit. These exports include about £1 billion of HS 841112 turbojets as well as the doubling of vehicle exports noted earlier.

On the other hand, the Aston paper claims Malta’s lower UK imports since 2019 are due to distance, when in fact Malta’s UK imports were greatly inflated in 2019 by €900 million spent on imports of HS 890392 Motor yachts, for pleasure or sports. This Billionaire’s Folly more than doubled the UK’s usual exports to Malta, which reverted to their pre-Brexit levels in 2023. This is not proof that the TCA needs to be renegotiated. It is also not proof that the UK trades less with countries that are further away – the UK’s largest export market for Motor boats and yachts is the United States. In fact, the US is the UK largest single export market.

By the way, Motor yachts is another HS code that has been replaced by three codes depending on the length of the yacht. So besides distorting the trade data for Malta, it will have also added another ‘lost export variety’ to the Aston data base.

9. Fake checking, peer reviews and misinformation

I could go on and on about this report. I have found an error, misinterpretation or false assumption in almost every paragraph. I hope that Graham Gudgin will be able to check Aston’s econometric calculations and perhaps write part 2 of this report. I don’t have the computer power to do that so I must rely on my common sense and knowledge of international trade, the TCA Rules of Origin, HS codes and UK agriculture.

But there is an important point to make before I finish: How could any university publish such a report without having it peer reviewed? And if it was peer reviewed, then Aston needs to find more knowledgeable reviewers.  Secondly why did so many media companies pick up this report and run with it without either reading it or making even the slightest checks on its findings? Where was BBC Verify? Why did the BBC publish not just one, but two articles on this report, without even apologizing for the mistakes in the first?

[1] This link goes to an update of the original story by Faisal Islam. The first was so full of errors the BBC took it down, but the replacement is also full of different errors.

[2] Page 55

[3] Unbound: UK trade post-Brexit, Du, Liu, Shepotylo and Shi, Page 12

[4] HS84 Machinery, HS71 Precious Metals, HS87 Vehicles, HS27 Mineral Fuels, HS85 Electrical machinery.

[5] HS84 Machinery, HS71 Precious Metals, HS87 Vehicles, HS27 Mineral Fuels, HS85 Electrical machinery, HS 30 Pharmaceuticals, HS 99 commodities not elsewhere specified, HS 90 precision equipment, HS 88 Aircraft, spacecraft and parts and HS29 Organic Chemicals.

[6] Annual data, GBP, from ITC