Share this article: Facebooktwitterredditpinterestlinkedinmail

Image of Alan Winters8 November 2021

L. Alan Winters is Professor of Economics and Founding Director of the UKTP0 and Guillermo Larbalestier is Research Assistant in International Trade at the University of Sussex and Fellow of the UKTPO.

Key Findings:

  • To date, the UK government has signed no new trade agreements relative to what it would have had as a continuing member of the EU.
  • The Government estimates that the two agreements in principle announced this year (Australia and New Zealand) will increase UK Gross Domestic Product by between £200 and £500 million annually – that is, 0.01% to 0.02% (one to two ten-thousandths) of GDP or between £3 and £7 per head of population – and that only after they have bedded down over 15 years or so .

We were asked to sum up the economic benefits of the UK’s new post-Brexit trade agreements. Our first observation is that if we take as a starting point the trade agreements that the UK would have been party to as a member of the EU, the government has, to date, signed no new trade agreements!

There are two new trade agreements in principle. That with Australia was announced in June 2021 but has not yet been finalised, and that with New Zealand was announced last month. As the full details of these agreements have not yet been released (or perhaps determined), we do not know what their effects will be, but in June and July 2020 the government’s preliminary analysis (which is not likely to be an underestimate of the worth of the agreements) suggested that in the long run, the trade deal with Australia will raise UK GDP by between £200 and £500 million annually – that is, 0.01% to 0.02% (one to two ten-thousandths) of GDP or between £3 and £7 per head of population. The trade agreement with New Zealand is estimated to raise GDP by basically zero – minus 0.01% to plus 0.01%. (The long run ‘is generally assumed to represent around 15 years from implementation of the agreement.’)

Government analysis also treats the agreement with Japan (the Comprehensive Economic Partnership Agreement – CEPA) as new because the EU’s agreement was not fully operational when it was undertaken. In fact, however, CEPA is modelled extremely closely on the EU-Japan agreement, with a few small differences.  These include extensions on, e.g., digital trade, and one major loss – the inability of UK exporters to the EU to be able to count Japanese parts towards meeting the EU’s requirements for getting zero-tariff entry to the EU. Relative to having no agreement, the government estimated that CEPA would raise UK GDP by £1.5 billion (0.07% or £22 per head), but relative to what the UK would have had without Brexit the gains will be negligible or negative.

Parts (A) and (B) of the table below give the numbers in more detail, along with estimates from the same sources of the effects of these agreements on UK exports and imports with the partners. Of course, all modelling results are uncertain, but the ranges quoted reflect the different trade liberalisation scenarios assumed in the government’s analysis. Thus, for example, the UK-Australia deal would raise UK exports by 3.6% under a ‘substantial’ trade liberalisation and by 7.3% under an even more ambitious one. [One can also see that increases in imports are predicted to substantially exceed those in exports, but that is a story for another occasion.]

The changes in trade with the partners look impressive, but remember two key facts:

  • A trade agreement will divert some trade to, say, Australia from other partners, so total UK trade will increase by less, and
  • Australia and New Zealand are small trading partners, accounting for 1.7% and 0.2% of UK exports and 1.0% and 0.2% of UK imports respectively (in 2019).

But the small shares of overall UK trade are not enough to explain the modest impacts. Even the proposed UK-US trade agreement, shown in part (C) and a key aspiration of the Government, was estimated to increase UK GDP by only a modest 0.07%-0.16%. Bear in mind that the US is the UK’s largest single-country trading partner accounting for 20.5% and 17.9% of total exports and imports, respectively, in 2019.

The UK has also signed nearly 70 other agreements that are new to the UK, but which are designed as closely as possible ‘to reproduce the effects of trading agreements that previously applied’ when the UK was a member of the EU. They add nothing to UK trade, and, because they are not perfect replicas, actually harm it very slightly.

The issue is not that trade agreements are worth nothing, but that they barely scratch the surface of the UK’s challenge to make up the GDP lost by leaving the EU, which the Office for Budget Responsibility puts at 4%.

Thus, one might ask ‘if the effects are so small, why bother?’ Among the possible answers are that:

a) while each agreement is small, cumulatively they could sum to a significant amount,

b) the current agreements could be stepping stones to bigger, regional agreements, such as the Comprehensive and Progressive Trans-Pacific Partnership,

c) there are other impacts beyond GDP and trade, such as on longer-run innovation, investment and productivity impacts, which the assessments are not good at capturing,

d) there may be specific sectoral access issues;

e) there may be additional non-trade objectives related to climate and the environment, for example.

None of these is convincing economically. On (a) and (b), non-EU partners account for about half of UK total trade and so, to counteract the OBR’s 4% loss from Brexit, would require agreements with each and every one of them to induce trade changes that create a 4% increment to UK GDP. That is nowhere in sight in the numbers in the table.

Answer (c) is conceivably true but there has been no such evidence and one would have to explain why agreements with non-EU countries would have this effect while a deeper agreement with the EU would not. Similarly, on (d), there certainly are disproportionate gainers from these agreements, but what are they and why do we preference them? On (e), again it is conceivably true, but the dilution of the climate clauses with Australia suggests the opposite.

The sad answer is that

f) the government is happy to accept, on our behalf, the economic losses from Brexit in return for political benefits (sovereignty), and trade agreements with other countries are merely making the best of a bad job from an economic perspective.

Table: Summary of Government Estimates of the Economic Impacts of the UK’s post-Brexit Trade Agreements – differences from no agreement scenarios

 

UK GDP

Exports to

partner

Imports from

partner

Total UK Exports

Total UK Imports

(A)  New Agreements in Principle

Australia1

0.01% – 0.02%

£200 – £500 million

3.6% – 7.3%

7.4% – 83.2%

0.1% – 0.3%

0.0% – 0.1%

New Zealand2

-0.01% – 0.01%

n/a

3.8% – 7.3%

14.9% – 40.3%

0.1% – 0.2%

0.0% – 0.1%

(B)   Agreement closely modelled on previous EU agreement (no net gain to UK relative to continued EU membership)

Japan3

0.07%

£1.5 billion

17.20%

79.94%

0.58%

0.51%

(C)   Agreement not yet negotiated

USA4

0.07% – 0.16%

£1.6 – £3.4 billion

4.3% – 7.7%

4.1% – 8.6%

0.7% – 1.3%

0.1% – 0.2%

Sources

[1] UK-Australia Free Trade Agreement The UK’s Strategic Approach (p.55)

[2] UK-New Zealand Free Trade Agreement The UK’s Strategic Approach (p52)

[3] Final Impact Assessment of the Agreement between the United Kingdom of Great Britain and Northern Ireland and Japan for a Comprehensive Economic Partnership (p.37)

[4] UK-US Free Trade Agreement (p.56)

 

Disclaimer:
The opinions expressed in this blog are those of the author alone and do not necessarily represent the opinions of the University of Sussex or UK Trade Policy Observatory.

Republishing guidelines:
The UK Trade Policy Observatory believes in the free flow of information and encourages readers to cite our materials, providing due acknowledgement. For online use, this should be a link to the original resource on our website. We do not publish under a Creative Commons license. This means you CANNOT republish our articles online or in print for free.

2 Comments

  • Christopher John Lovejoy says:

    Estimating the impact new trade deals will have on the UK is in the Public Interest and the Independent recognised this when it published your findings on 7.11.21.
    Is it possible for you to quantify any additional costs that may be incurred by leaving the EU? For instance, the EU’s consumer requirements have already caused problems to both the UK’s fishing and financial service industries. And, fewer EU citizens are now working in the UK which has increased the difficulty in providing nursing care, driving HGVs, harvesting agricultural products etc.
    If it possible to quantify some of these costs, the Independent could alert its readers that the data they published on 7.11.21 understates the losses that Brexit may cause. Achieving this would also be in the Public Interest.

    • Charlotte Humma says:

      All estimates of the costs of Brexit are imprecise, but to the extent possible I think that disruption due to differences in standards and frictions in the financial sector are taken into account. The costs of the migration component arguably have not been and it is something one might start to try to quantify. One of our colleagues is starting work on it, but it will be months before there are results because we do not really know yet how it will work out in the longer run and because labour shortages are rather specific so there are many cases to work through. In a rough quantification in late 2018 – ‘EU Exit: Long-term economic analysis’, November 2018 – the UK Government suggested that cutting immigration from the EEA to zero would increase the costs of Brexit by about 1.7% of GDP.

Leave a Reply

Your email address will not be published. Required fields are marked *