"Splendid Isolation" vs. "Common Insularity": Quantifying the trade effects of Brexit

As the deadline gets closer, the United Kingdom (UK) public debate is heating up on an event that, one way or another, could change the identity of the European Union (EU) and its economic relations with Japan and the rest of the world: the referendum on the UK's membership of the EU in 2017 (or possibly even earlier) to which the UK Prime Minister and his Conservative party committed since 2013.

The decision by the UK of staying or going can be described as a decision on the trade-off between "exit" and "voice". If the UK decides to stay in the EU, it will happen because its demands for a lighter approach to European integration (based on more national independence in several policy domains) will be expected to receive positive answers from Brussels, leading to a more flexible EU with variable geometry across different policy dimensions and a rethinking of the original commitment to ever deeper integration as currently envisaged by most continental members. The UK will also decide to stay because its bargaining position with respect to non-European countries such as Japan will be much weaker as a standalone negotiator than as a member of the EU. Finally, it will decide to stay because its internal market will become less attractive to international companies and investors if it is not an integrated part of the much bigger EU market.

From this Europhile perspective, the UK should stay in order to make its "voice" heard inside and outside Europe. This would change the EU from the inside injecting some British common sense into a sometime too abstract continental view of what the EU should be.

Differently, if the UK decides to leave the EU, deeper integration will likely gain prominence at the top of the EU policy agenda, with remaining members strengthening their economic, political and social ties. In this respect, it will give them an unprecedented opportunity to further develop the European model of "compassionate capitalism" as opposed to the perceived Anglo-Saxon model of "ruthless capitalism". On the other hand, also the UK would be free to pursue its alternative self-determined path to efficiency and competitiveness in the global economy without the burdening red tape imposed by bureaucrats in Brussels under the perceived pressure of rent-seeking special interests from the Continent.

From this Eurosceptic perspective, the "exit" of the UK from the EU would be good for both sides as "voice" is just wasted breath when people speak too different languages. This would change the EU from the outside as future decisions by remaining members would be also shaped, to some extent, by some degree of contraposition with the UK.

One way or another, after the referendum in the UK the EU could become a very different partner and market for companies and investors from Japan and the rest of the world.

From an economic point of view the implications of all those effects are hard to quantify. There is, however, an economic area in which quantification is not impossible: international trade. This is an important and arguably dominant area for Brexit. The reason is that trade between the UK and the EU has grown significantly since the UK joined the European Economic Community around forty years ago. In 1973 the share of UK exports to the EU was slightly more than 30%. By the time of the recent crisis that share had reached its climax at almost 55%, remaining still over 50% in 2011. This is a disproportionately large number as the EU (without the UK) accounts for 20% of world GDP (without the UK). From the reverse angle, EU expenditures on UK exports account for roughly 15% of UK GDP, showing that the EU is not only a key customer but also a key supplier for the UK. Could Brexit jeopardize these intense trade relations? If so, how large would the associated welfare losses for UK citizens be?

These questions have been addressed by Ottaviano et al (2014a,b) using a standard but state-of-the-art quantitative model of international trade (Note 1). Their model considers trade in 35 sectors (including intermediate inputs) among the 40 major countries in the world. An important feature of the analysis is that the model is static and that the effects of trade on welfare are primarily driven by the specialization of countries in industries in which they enjoy comparative advantage. Hence, the analysis does not account for either the dynamic effects of trade on productivity growth or other static effects of trade on competition, selection, scale of production and product variety. In this respect, Ottaviano et al (2014a,b) underestimate the welfare gains from remaining in the EU.

Most naturally, in order for the model to quantify the effects of Brexit, one still has to make assumptions on how trade costs would change between the UK and EU if the former separated from the latter. Ottaviano et al (2014a,b) consider two scenarios. In the "optimistic" scenario, after Brexit the UK keeps on enjoying the same degree of access to the EU common market as it does now. While this is what Switzerland or Norway currently enjoy by paying a sort of "admission fee", in the optimistic scenario the EU is assumed to waive such fee for the UK.

Differently, in the "pessimistic" scenario, after Brexit the UK cannot negotiate such favourable terms given its "renegade" status of former member. Hence, in this scenario trade costs between the UK and the EU increase for three main reasons. First, tariff barriers may go up. Second, also non-tariff barriers (due to regulations, border controls, etc.) may rise. Third, the UK may not be allowed to take advantage of future deeper EU integration (including any possible reduction in non-tariff barriers).

In their simulations, Ottaviano et al (2014a,b) define the two scenarios as follows. In the pessimistic scenario, after Brexit UK-EU trade takes place under MFN tariffs on goods (Note 2). While this looks reasonable immediately after Brexit, it is not unthinkable that after a while the UK will be able to negotiate a better tariff deal with the EU, similar to those involving Norway and Switzerland. Hence, the optimistic scenario considers that tariffs continue to be zero between the two parts.

Turning to non-tariff barriers related to regulations and other legal obstacles, they concern both goods and services. The optimistic scenario assumes that the UK has to face a fourth of the non-tariff barriers currently imposed by the EU on US imports. The pessimistic scenario assumes, instead, that the UK would have to cope with two thirds of those barriers.

But leaving the EU would also prevent the UK from enjoying the future reductions in non-tariff barriers that could come with deeper EU integration. As trade costs for intra-EU flows have fallen roughly 40% faster than in any other OECD country, a decade from now the gap in non-tariff barriers between intra-EU trade and extra-EU could become larger than its current value. Accordingly, the pessimistic scenario assumes that intra-EU non-tariff barriers will keep on falling 40% faster than in the rest of the world, implying a cumulated differential decrease in trade costs of 10% that the UK would forego with Brexit. The optimistic scenario assumes that intra-EU non-tariff barriers will fall only 20% faster than in other countries, thus leading to a differential decrease in trade costs inside the EU of only 5.7%.

Table 1 reports the implications of Brexit for the UK as derived from simulating the quantitative trade model in the two scenarios. The table shows that even in the optimistic scenario the costs associated with current and future non-tariff barriers determine an overall welfare loss of 1.66% of GDP. This is explained by the fact that non-tariff barriers are particularly relevant for services, in which Britain enjoys a remarkable comparative advantage. The overall loss increases to 3.62% of GDP in the pessimistic scenario and it is again mostly driven by higher non-tariff barriers.

These losses are much bigger than any fiscal saving the UK can attain from cancelling its net transfers to the EU, which are estimated to account for 0.53% of its GDP. All in all, in cash terms, the loss (net of any fiscal saving) for the UK amounts to £50 billion in the pessimistic scenario and still a substantial £18 billion in the optimistic scenario.

As the numbers reported in Table 1 quantify the losses due to foregone specialisation according to comparative advantage, they should be seen as lower bounds on the total costs of Brexit. A first reason for this is that the comparative advantage model misses other sources of static losses such as poorer variety of imported goods and services, reduced economies of scale, weaker competition and increased misallocation due to more limited exit of less productive firms (Corcos et al, 2012).

Another, and possibly more important, reason is that the numbers in Table 1 do not account for the impact of trade on growth. Econometric studies of important episodes of trade liberalisation (such as the EU's single market programme of the early 1990s) tend to uncover much larger effects of trade on output than those simulated in static models. This is consistent with the idea that trade can increase productivity via innovation and the adoption of better technologies triggered by tougher competition. When static quantitative models are enriched with these dynamic effects, they predict that technology adoption doubles (Bloom et al, 2014) or even triples (Sampson, 2014) the static effects. These predictions are in line with the findings of econometric studies of the effects of EU membership (Baier et al, 2008; Feyrer, 2009).

Hence, considering all these additional effects of trade should at least double the static losses from Brexit reported in Table 1, leading to an expected loss of 2.2% of GDP in the optimistic scenario and nearly 10% in the pessimistic one. To put these numbers in perspective, during the global financial crisis between 2008 and 2009, the UK's GDP fell by around 7%.

Quantifying the costs and benefits for the UK of leaving the EU is complex but quantitative analysis of the losses associated with foreign trade alone shows that costs could be quite large.

What about the effects of Brexit on other countries including Japan? This question is even more difficult to answer given the higher dimensionality of the quantification exercise when the focus is on several countries. A promising way to circumvent this problem is to focus on specific industries. This is what Head and Mayer (2015) do for the car industry with an emphasis on multinational production. Using a quantitative model that shares some basic ingredients with the one used by Ottaviano et al (2014b), in a pessimistic scenario Head and Mayer (2015) find substantial reallocation of car production away from the UK (-12.1%) and to a lesser extent from other big EU car producers (-3.5 for France and -2.2% for Germany) towards extra-EU producers (+1.6% for Japan and +1.7% for South Korea).

This shows that there are important industries in which Brexit could have substantial effects around the world. Companies and investors in Japan and elsewhere should scale up their contingency planning.

* This article draws from Ottaviano, Pessoa, Sampson and Van Reenen (2014a,b) and Dhingra, Ottaviano and Sampson (2015).