Changing the terms of trade


Trade policy has been a key issue in discussions surrounding the UK’s future relationship with European Union. In this bulletin, we present a model of the UK’s trade, in order to evaluate various policy options. We consider scenarios in which the UK needs to renegotiate its relationship with the EU, and is also able to pursue its own trade policies vis a vis the rest of the world.

We find that the costs of a “Hard Brexit”, in which the EU and the UK trade on MFN terms with each other, cannot be offset even on optimistic assumptions about the UK’s ability to expand trade with the rest of the world by signing free trade agreements.

We also examine an intermediate option in which the UK exits the single market, but preserves free trade in goods and services with the EU by negotiating a free trade agreement with it. Such an agreement would require that the UK and the EU agree liberalisation commitments in services trade, that are substantially broader and deeper than any free trade agreement the EU has entered into to date with non-EU/ EEA members. Deep liberalisation commitments on the movement of natural persons could help to preserve labour market access to service suppliers, while allowing a reconfiguration of immigration policy. But even under these conditions, there are likely to be hidden costs associated with rules of origin, which could also have an impact on the extent to which the UK would benefit from free trade agreements with non-EU members.

We conclude that in principle, there is a Brexit pathway that could allow the UK to preserve its economic prospects, or to at least mitigate the damage to these. Under such a pathway, the UK would agree a bespoke agreement with the EU and over time supplement this with deals with the rest of the world. But it is an inherently fragile pathway, and one that would require a convergence of factors, economic and political, that has not been observed in the past.


Following the referendum of 23 June 2016, the Prime Minister has committed to invoking Article 50 of the Lisbon Treaty by the end of March 2017. Both the process and timeline for invoking Article 50 are subject to the Supreme Court’s review of the High Court ruling that a parliamentary vote is necessary prior to invoking Article 50. The European Commission’s Chief Negotiator, Michel Barnier, has called for the negotiations under Article 50 to be completed by November 2018.

The Article 50 negotiations focus on the terms of the UK’s departure from the European Union. They do not, in and of themselves, focus on the arrangement that will succeed the UK’s membership of the EU, though clearly the process of negotiation means that there will be links between the two. Few details have emerged to date about future arrangements.  References to a “hard Brexit”, in which trade between the UK and the EU reverts to most-favoured nation (MFN) status under WTO rules, have received much press coverage.[1] But this is to be expected: each party to the negotiation has incentives to signal their toughness in the hope that the other will relent. Moreover, the UK and the EU have, for their own specific reasons, an incentive to signal a tough position to internal political constituencies.

The purpose of this note is not to speculate on what arrangements between the UK and the EU may emerge in the future. Instead, the objective is to measure the impact on the UK’s economic prospects of different post-Brexit scenarios. To do this, we provide an empirical model of the UK’s trade based on historical data, to measure the main factors driving the UK’s trade with the EU and the rest of the world. We include the rest of the world because under many plausible Brexit scenarios, the UK would regain control of its trade policies vis a vis the rest of the world. This gives it the ability to strike Free Trade Agreements (FTAs) with non-EU partners

Of particular interest is how far any trade-creating effects of FTA’s with non-EU partners could mitigate the losses incurred by any future restrictions to market access the UK may face vis a vis the EU. We also draw on the distinction that is often made between participation in the EU’s single market and free trade with EU.[2] We examine to what extent any losses from exiting current single market arrangements could be mitigated if the UK were to sign a FTA with the EU, and a series of FTAs with the rest of the world.


A widely accepted way of modelling the trade flows is to use a gravity model of trade. Gravity models show how much trade should be expected between countries, as a function of their size and their proximity. “Proximity” in this context refers to physical distance, as well as less tangible measures of proximity, such as a common language or a shared colonial history. Standard trade theory tells us that trade between countries will be driven by differences in their resource endowments and/ or consumer tastes. The intuition behind the gravity model is to qualify this prediction by proposing that countries will trade more with larger countries, and those that are closer in any of the sense described above. Even though trade is global, the biggest immediate opportunities may be those that are nearer to home.

To this basic model we can add the effects of trade policy restrictions. We use a measure of overall trade restrictiveness that captures the effects of tariffs and non-tariff measures.[3] We can also include a variable reflecting whether pairs of trade partners are members of a FTA.  Our approach is in line with several major studies on the effects of FTAs.[4]

Figure 1 presents a first set of results. The blue and red bars represent, respectively, actual imports and actual exports between the UK and its major trading partners. The triangular data points represent the value of exports predicted by the model, and the circular data points represent predicted imports.

Figure 1. Actual and predicted trade patterns between the UK and its major partners

 Figure 1

The results generally bear out the model’s predictions that the UK trades intensively with large economies that are geographically proximate (the EU countries) or culturally proximate (the US), or that differ in terms of resource endowments (China). The results highlight the fact that the EU is UK’s the largest trade partner, and that we would expect this to be the case because of the interplay of economic size and proximity, along with policy factors.

The model does not fit perfectly. For example, actual trade with Germany, France, and Italy is lower than predicted. This may be because the trade restrictiveness measure we used does not capture all aspects of policy. For example, it does not capture policy variables that affect the ease of doing business in or between countries. If that is the case, the results above point both to the importance of the EU in UK trade, but also that there remains unfinished business between the UK and the EU – trade flows could be further deepened through more policy integration.[5]

Our model allows us to test the effects of FTA membership on trade flows.

  • We find that on average, membership of an FTA increases trade by around 30-40% compared to a counterfactual case in which there was no FTA.
  • We find that FTAs are more effective in generating trade than linear reductions in trade restrictions between trade partners; these stimulate exports by around 2% for every percentage point reduction in restrictiveness. For example, the UK reducing its restrictiveness to zero would imply only a 12% increase in imports with non-FTA trading partners.

The reason for this difference in impact is that many FTAs tend to be comprehensive in their sectoral coverage, and often incorporate many dimensions of policy and institutional integration (including trade facilitation, investment measures, regulatory cooperation and so forth). Economists sometimes refer to this as “deep integration”. The EU is probably the foremost example of deep cooperation, as evidenced by the key elements underpinning the single market, and the development of common institutions and rules.


What are the implications of our modelling for possible Brexit pathways? The first is that under a “Hard Brexit”, the losses to the UK would be substantial, and would not be fully offset by the gains in trade generated by FTAs the UK could sign on to with the rest of the world.

  • If the UK were to exit the EU, and revert to trading with EU trade partners on MFN terms, the overall loss to UK exports would be around $67 billion per year, to which we would add another U$5 billion as a result of withdrawing from FTAs to which the UK is party via its membership of the EU. The corresponding figure for imports would be around US$ 100 billion and US$ 5 billion
  • If the UK were to conclude FTAs with the rest of the world, the boost to exports would be around US$ 62 billion and to imports around US$ 80 billion. The bulk of the export effect (around US$54 billion) would be the result of 10 FTAs, with a FTA with NAFTA accounting for around US$ 25 billion. China would follow with US$7 billion.

These results are represented in in figure 2 below.

Figure 2. Predicted effects of losing free trade with the EU/ EEA and negotiating FTAs with non-EU partners

 Figure 2

The findings underscore the importance to the UK at maintaining a deep level of integration with the European Union. The results likely understate the effects of a Hard Brexit because:

  • The estimates are derived on the basis of the average effects of FTAs on trade. In reality, this average is likely to understate the trade-creating effect of the EU on the UK’s trade, because the EU involves a much deeper level of liberalisation and integration, via single market membership, than average FTAs. Hence losses from a hard Brexit are likely to be understated as well. The average FTA effect also likely overstates the gains the UK might achieve through FTAs with some non-EU partners who may be less willing to implement deep liberalisation commitments.
  • These estimates assume that the FTAs with non-EU members will come into force and be fully implemented at the point of Brexit. In practice, negotiations are likely to take time (certainly if multiple partners are concerned) and the implementation of the commitments negotiated (which is what ultimately drives trade-creation) would take even longer.

But another way of considering these results is to ask whether the point to potential benefits to the UK if the UK were to: (i) retain free trade with the EU; and (ii) conclude trade deals with other partners. From the point of view of our modelling approach, free trade between the UK and the EU does not need to take the form of participation by the UK in the single market. In principle, the gains could be generated by a more classic FTA, along the lines the EU has entered into recently with Korea or Canada. This reflects the fact that our modelling estimates are based on average FTA effects.

In practice, there will be several challenges with this approach. We consider these in the policy discussion at the conclusion of this paper. We should also emphasise that these results do not mean that there are no costs from substituting participation in the single market with a more “arms-length” FTA arrangements. As observed above, our results probably understate the effects of losing single market status, and we consider some of these effects by referring to trade in services in a subsequent section of this paper.


Before we consider services trade and enter into an in-depth policy discussions, it is worth examining how the scenarios described above play out at sectoral levels. This is important as it helps us to understand the distribution of gains and losses. Figure 3 below shows changes in exports and imports under a scenario in which the UK reverts to MFN trade with the EU while concluding free trade arrangements with other partners.

Figure 3. Sectoral effects of losing free trade with the EU/ EEA and concluding FTAs with non-EU trade partners

 Figure 3


The overall result, namely that the effect of relinquishing free trade with the EU dominates the trade create effects of FTAs with non-EU partners, holds for most, but not all sectors. Imports of textiles, import and exports of machinery, and imports and exports of manufactures are sectors in which trade creating effects of FTAs with non-EU members dominate.

If we focus simply on the effects of FTAs with non-EU members, we observe that with the sole exception of motor vehicles, these FTAs have a more pronounced effect on imports than exports. From a purely economic point of view, import liberalisation is welfare enhancing since it increases consumption possibilities in the UK. From a political economy point of view, this import-enhancing effect may pose a challenge, given that one of main factors motivating the vote in favour of leaving the EU appears to have been hostility to import-competition. In any event, these findings suggest the need for the government to undertake pro-active policy measures to ensure an even distribution of the gains from trade, including training and assistance measures for workers displaced by import competition.


The previous sections focused on goods trade. An investigation into services is warranted given that the UK is the world’s second largest exporter of services, and that services account for around 80% of GDP. Moreover, services is an area where participation in the single market has had a liberalising effect that is deeper than that achieved by standard FTAs. In particular, as reported below, services trade liberalisation has been facilitated by commitments to the free movement of labour.

Measuring the effects of various Brexit scenarios on services trade is hampered by data constraints around the measurement of services trade flows and services trade restrictions. It is also not easy to identify what aspects of actual services trade liberalisation in EU jurisdictions are specifically attributable to EU membership, and which ones reflect reforms that would have happened anyway.

Figure 4 provides an overview of the direction of UK services trade. The figure underscores the importance of the EU for the UK’s services trade. The numbers likely understate the importance of the EU to services trade since in some cases (financial services trade with the US, notably) the presence of foreign affiliates in the UK is predicated on the UK’s access to EU markets.


Figure 4 Direction of UK services trade by major partner

 figure 4 updated

We draw on recent work conducted by the OECD measuring services trade restrictiveness, and the effects of these on both services and goods trade (the reasoning being that services inputs are key inputs into the production and trade of goods).[6] The econometric research linking changes in services trade restrictiveness to changes in trade flows in still in its early stages, and there is uncertainty surrounding the strength of the effect. Taking a mid-point between different estimates, we find that a percentage point drop in services trade restrictiveness increases exports of services by 1.6% and imports by 1.1%, whereas the effects on goods trade (exports and imports) is around 1.5%

Based on this, we focus on one particular aspect of services trade: its relationship to the free movement of labour with the EU. A commitment to free movement of labour reduces the level of restrictiveness on services supplied through the movement of natural persons (mode 4) and also through commercial presence ( mode 3; for example, by doing away on restrictions associated with intra-corporate transferees). We use the OECD’s services trade restrictiveness policy simulator to work out changes in services trade restrictiveness as a result of rescinding free movement of people. The effect on the UK would be to reduce services exports by $14 billion and imports by $12 billion.

The results are significant because they highlight the following:

  • Free circulation of labour is directly connected to the trade benefits derived from access to the EU’s single market. From an economic point of view, at any rate, it should not be construed as “price” to be paid to access the single market.
  • Classic FTAs have usually addressed the movement of labour through specific commitments under services in relation to modes 3 and 4. But these commitments have usually not generated much actual liberalisation and have fallen well short of current EU treaty commitments. Any FTA negotiated between the UK and the EU would thus need to include commitments at deeper levels than under any currently existing FTA. If such deeper commitments are negotiated, they may offer a route by which the UK could retain access to the labour and skills important to its economy, while also retaining the ability to set its overall immigration policy.

These results are not a complete representation of the potential effects that exiting the single market might have on services. This is the case in sectors governed by specific regulations with a fundamental impact on market access. Examples include aviation, audio-visual and media services, and certain financial services. The situation in aviation is particularly challenging given that there are no multilateral trade rules on which the UK could fall back.[7]

A loss of market access would be detrimental to these sectors, and is liable to have knock on effects through the linkages between these sectors and the wider economy. These are not captured in the modelling results.


Econometric techniques allow us to measure the historical contribution of the EU to UK trade. They show that a “Hard Brexit”, in which the UK and EU revert to trading with each other on MFN terms, will generate losses that cannot be offset even under very optimistic assumptions about the UK’s ability to conclude FTAs with non-EU trade partners.

Exiting the single market does not necessarily mean the ending of free trade between the UK and the EU. A FTA could secure at least some of the gains associated with the single market, but there may be still some losses given that classical FTAs observed to date represent a less deep form of integration than the single market. There is then a theoretical possibility that the UK could offset the losses associated with moving to a less deep form of integration with the EU, by concluding FTA with trading partners in the rest of the world.

Various factors will have an impact on how well this this approach translates from theory into practice.

First, as already observed, the process of negotiating FTAs with partners in the rest of the world will require time and resources, especially if the UK were also to be in the process of negotiating arrangements with the EU. Moreover, once negotiated, FTA’s would need to be implemented, which may require further time. Some trading partners (notably as Canada) are understood to be ready and eager to negotiate, but for others, such as China, the USA, or ASEAN countries, the process may be more complex. These factors increase the time required for the benefits of FTAs with the rest of the world to accrue to the UK, and reduce their certainty. The UK may be able to address this to some extent if it is able to negotiate a sufficient transition period with the EU between the phasing out of single market arrangements and the entry into force of a new arrangement. A transition period may be required anyway given the likely time required by both the UK and the EU to negotiate with each other, and then ratify and implement the results.

Secondly, if the UK were to negotiate FTAs with the rest of the world, and maintain free trade with the EU, it would be necessary to agree preferential rules of origin with the EU. These rules are needed to ensure that goods originating in partners with which the UK would have a FTA, but the EU does not, do not “leak” duty free into the EU. These rules create compliance costs for firms that operate cross-border supply chains. OECD data suggest that at present, most (between 80 and 90%) of foreign value added in UK final demand and exports comes from the EU. This in turn suggests that the effects of any new rules on existing value chains may be limited. But the new rules may affect the extent to which value chains involving non-EU partners may develop, thus affecting the value of FTAs with non-EU members.

Thirdly, our results suggest that the deeper the commitments in a FTA between the UK and the EU are, the lower the losses from the UK exiting the single market. For example, the UK and EU could negotiate deep commitments on the movement of natural persons (mode 4) on services on the movement of employees associated with commercial presence. These commitments could be carried over into a chapter on investment to deal with goods. These commitments could allow the UK and EU to retain a considerable degree of freedom of movement of workers and the ability of the UK businesses to access EU labour, while preserving some autonomy over immigration policy more generally. Similarly, the UK and the EU could seek to negotiate specific chapters on financial services, aviation, and audio-visual services. Negotiating all of these matters would mean that the UK and the EU would need to agree a FTA of unprecedented depth – substantially more than, say, what the EU and Canada have recently agreed. Whether this is politically possible for both the UK and EU is subject to considerable uncertainty.

In sum, if we set aside the “corner solutions” of a Hard Brexit and the single market status quo, an intermediate pathway to Brexit that limits costs to the UK may be available. For this to happen, the UK would need to simultaneously secure a deep FTA with the EU, and over time enter into FTAs with significant trading partners. But this pathway is fragile. It is likely to contain hidden costs, and would require several factors to align in ways and to a degree not observed in the past.





[1]     We prefer the expression “reverts to MFN status” to the more commonly used expression “reverts to WTO rules” as trade between the UK and the EU is already subject to WTO rules, including, specifically, those  governing the regional trade agreements in goods and services.

[2]     Participation in the single market involves free trade in goods and services, free movement of capital, free movement of workers, and incorporation of EU regulations.  A FTA would involve commitments to liberalise trade in goods and services, without necessarily requiring that the UK undertake the same level of commitments to

[3]     See Kee, Hiau Looi, Alessandro Nicita and Marcelo Olarreaga. 2009. “Estimating trade restrictiveness indices”, Economic Journal 119: 172–199. (pdf file, 326kb)

[4]     See notably Baier, S.L., J.H. Bergstrand, P. Egger and P.A. McLaughlin (2008) ‘Do Economic Integration Agreements Actually Work? Issues in Understanding the Causes and Consequences of the Growth of Regionalism’, The World Economy 31(4): 461-97; Magee, C. (2008) ‘New Measures of Trade Creation and Trade Divergence’ Journal of International Economics 75(2), 349-62

[5]     To take a concrete example, research undertaken by Frontier Economics showed that convergence to best practice in insolvency regulation across the EU could boost trade and investment flows, generating an increase in EU28 GDP worth between 41 billion and 78 billion Euros per year, with France and Italy amongst the biggest gainers in absolute terms. See  AFME (2016), Potential economic gains from reforming insolvency law in Europe. Available at

[6] See notably Hildegunn Kyvik Nordås, Dorothée Rouzet (2015), The Impact of Services Trade Restrictiveness on Trade Flows – First  Results, OECD Trade Policy Paper.

[7]     For a fuller treatment of these issues, see the Frontier Economics bulletin on Brexit and Aviation at


Comments are closed.