Abstract
Selected Issues
Long-Term Impact of Brexit on the EU1
The integration of EU27 countries and the United Kingdom has strengthened over time, reflecting shared gains from the EU’s single market. Conversely, the departure of the U.K. from the EU (Brexit) will inevitably represent a loss for both sides. In this paper we use two approaches to estimate these losses. First, we create a multidimensional index that captures the depth and evolution of integration between the U.K. and the rest of the EU, taking into account trade via supply chains, financial linkages, as well as migration. We then use this index to estimate the average long-term impact of several Brexit scenarios. Second, we use a standard multi-country and multi-sector computable general equilibrium (CGE) model to estimate country- and sector-specific impacts from higher trade barriers between the U.K. and the rest of the EU countries. We find that the level of output of EU27 countries falls by between 0.06 and up to 1.5 percent in the long run. The range of estimates depends on whether we assume a “soft” or “hard” Brexit, or whether trade or other transmission channels are accounted for. These are likely losses that should be interpreted with caution, given the important uncertainty characterizing the empirical estimations. Moreover, there is substantial heterogeneity in the impact of higher trade barriers: countries such as Ireland, Netherlands, and Belgium are among the most affected in the simulations.
A. Euro Area and U.K.: How Strong are the Links?
Dimensions of Integration
1. EU-U.K. trade integration has benefited both parties. For example, the euro area (EA) runs a modest trade surplus with the U.K., while the U.K. has a small surplus in financial services trade with the euro area. In recent years, the euro area’s trade surplus with the U.K. increased steadily, owing to rising exports of goods, reaching 1 percent of EA GDP in 2016. In gross terms, total trade in goods and services between the euro area and the U.K. accounts for about 6 percent of euro area GDP on average over the past two decades. Trade with the U.K. is most significant for Ireland, the Netherlands, Belgium, and Luxembourg, relative to the respective sizes of their economies. The U.K. is a net provider of financial services to the euro area, driven by its large bilateral flows with Ireland. Excluding Ireland, the trade in financial services between the euro area and the U.K. is close to balance (Figure 1).
2. Trade with the U.K. involves complex supply chain linkages. Most trade today—over 50 percent of goods and almost 70 percent of services trade—is in intermediate inputs, suggesting the presence of supply chains.2 Therefore, it is important to capture also the indirect links via these supply chains when assessing euro area countries’ trade with the U.K. Moreover, it is important to account for the value added from third countries when assessing exports to, and imports from, the U.K. From a value-added perspective, euro area trade with the U.K. is a combination of direct and indirect value-added exports transiting through third countries, suggesting that supply chains also play a role. Smaller but open economies such as Ireland, Luxembourg, and Netherlands exhibit the highest exposure in value-added terms with the U.K., though this exposure is smaller than what gross trade statistics suggest (Figure 2).
3. EA-U.K. investment positions are substantial.
Euro area total financial claims and liabilities with the U.K. amounted to about 55 percent of euro area GDP in 2016 (Figure 3). Across countries, Ireland, Netherlands, and Luxembourg have the largest financial positions relative to their own economic size. Notably, the two-way FDI stock between Netherlands and U.K. is about 120 percent of Netherland’s GDP; the two-way portfolio investments between Ireland and U.K. is slightly below 230 percent of Ireland’s GDP; and the two-way bank claims between Luxembourg and U.K. is about 220 percent of Luxembourg’s GDP.
In net terms, the euro area provides financial capital to the U.K amounting to about 9 percent of euro area GDP. However, the aggregate number hides cross-country heterogeneities. The Netherlands and Ireland contribute most to the net FDI investment position (about 2.1 percent of euro area GDP in 2016). Ireland and Malta have large net portfolio investments position with the U.K., whereas most other countries are net recipients. Finally, relative to their own GDPs, Luxembourg and Ireland are large recipients of cross-border bank lending from the U.K. (more than 170 percent of GDP in the case of Luxembourg and 58 percent of GDP in the case of Ireland).
4. Migration flows between the euro area and the U.K. are small, except for some countries with historical ties to the U.K. The number of U.K. migrants living in the euro area is small relative to the euro area population, but has increased somewhat over time. The euro area has traditionally been a net sender of migrants to the U.K. for all skill levels, with a total balance of about 0.1 percent of the euro area population as of 2010. The number of migrants from Ireland, Cyprus, and Malta living in the U.K is considerable, accounting for roughly 10 percent of these countries’ population. Regarding migration from the U.K. to the euro area, there is one U.K. migrant living in the euro area for every four to five hundred euro area citizens. However, the U.K. migrant population is larger in Ireland, Luxembourg, and Spain.3
What Do the Stylized Facts Suggest of the Potential Impacts of Brexit?
5. The strength of euro area-U.K. integration implies that there would be no Brexit winners. First, the U.K. is among the top three main trading partners of the euro area. Second, the gross trade exposure masks complex supply chain linkages. Third, cross-border capital flows between the U.K. and the euro area are large. Finally, migration flows are considerable for some countries. Higher barriers to trade, capital flows and people movements following Brexit could disrupt these links, reducing trade, investment and labor mobility. All empirical studies so far concur that economic costs on both sides would be considerable. However, the EU27 would bear a disproportionally smaller share of the total cost due to its larger size.4
6. The long-run impact of Brexit is likely to be unevenly distributed across countries, with Ireland exhibiting the highest exposure. Losses would depend on bilateral integration with the U.K., sectoral specialization, the positioning of sectors within the global supply chain and the degree of substitutability between London and euro area capitals as financial centers. Countries that are more integrated (Ireland, Luxembourg, Netherlands, Belgium, Malta, and Cyprus) will likely suffer disproportionally from Brexit. Other countries, such as Germany, could also be affected via supply chains links.5
B. A Synthetic Index of Exposure to the U.K.
7. To account for the complexity of the exposure of EU27 countries to the U.K., we build a synthetic index for integration. As discussed in the previous section, the degree of integration has several dimensions, which are often correlated. To measure the degree of integration and its evolution over time in a less complex manner, we build a single index by aggregating the subcomponents into a synthetic country-specific, time-varying index. This index captures all the components of a EU27-U.K. economic relationships and can be used in the subsequent regressions to assess effect on euro area output and employment from integration with the U.K. Being a single index, it solves the multicollinearity problem that would arise if all the components of the economic relationship were to be used in a regression.
Trade in Goods, Services, and Financial Services
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Trade in Goods, Services, and Financial Services
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Trade in Goods, Services, and Financial Services
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Trade in Value Added and Migration
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Trade in Value Added and Migration
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Trade in Value Added and Migration
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Financial Flows
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Financial Flows
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Financial Flows
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
8. To build the integration index, we use a principal component analysis. Principal component analysis (PCA) can help resolve the dimensionality problem associated with the presence of a large number of variables capturing various aspects of a common phenomenon (in this instance, bilateral economic integration with the U.K.). Moreover, reducing the dimensionality problem helps reduce the multicollinearity problem that would have otherwise arisen if each of the channels of integration was introduced additively. We therefore focus on several indicators of integration, and for which data are available for European and non-European countries and which benefit from a sufficient coverage over time (starting in early 1990s):6
Trade in domestic value added. We derive an indicator of trade openness between each country and the U.K., measured as the sum of bilateral exports and imports of domestic value added normalized by the country’s GDP. Data are based on Ignatenko et al. (2017).7
Participation in supply chains. We use the sum of “backward” and “forward” trade linkages between each country and the U.K., normalized by the country’s GDP. “Backward” linkages refer to the foreign value-added embodied in the country’s and U.K.’s bilateral gross exports. The “forward” linkages refer to the country’s and U.K.’s exports of value added further re-exported to third countries. The overall indicator therefore captures the extent to which trade between the country and the U.K. involves the exchange of foreign value added, but also respective domestic value added embodied in exports and then further reexported to third countries. Data are from Ignatenko et al. (2017).
Openness in service trade. We use the sum of each country’s exports of services to, and imports of services from, the U.K., normalized by GDP. Data are from Ignatenko et al. (2017).
Cross-border banking positions. To capture the key role of London as financial center, including for cross-border lending activities, we use the ratio of claims by international banks in the U.K. on each receiving country from BIS locational data, normalized by GDP.
Migration. We use the share of each country’s migrants residing in the U.K., normalized by the country’s total number of migrants residing in 20 OECD countries. Data are from Brücker et al. (2013) who relied on harmonized census data.8 Migrants are defined as foreign-born individuals aged 25 years and older, living in each of the 20 considered OECD destination countries. 9
The principal component analysis (PCA) identifies the relative importance, e.g., weights of the different indicators in order to build the exposure index. The exposure index is rescaled so that it ranges between 0 (minimal exposure) and 10 (highest exposure). Overall, the first principal component explains 60 percent of the total variance and is positively correlated with the seven variables used to build the exposure index. For more details regarding the PCA, see the Appendix.
9. The integration index shows that euro area-U.K. integration has strengthened over the years. The intensity of integration has increased by 40 percent over the past 25 years, split in three distinctive phases. The first one, increasing by 20 percent, in the runup to euro adoption, the second one (after euro adoption) with the index staying relatively flat, and the last phase in the aftermath of the global financial crisis when integration increases by another 20 percent. Increased integration is in large part driven by a handful of countries such as Ireland, Belgium, the Netherlands, and Malta. Other countries that exhibit considerable economic ties with the U.K. are Germany, Finland, Cyprus, and other non-euro area countries such as Denmark and Sweden.
Synthetic Index of Integration With the U.K.
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Synthetic Index of Integration With the U.K.
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Synthetic Index of Integration With the U.K.
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
C. An Econometric Investigation of the Long-Term Effects of Brexit on the EU
Empirical Design
10. In the empirical analysis, we assess the long-term effect on EU27 output and employment of Brexit, modeled as a partial reversal of EU27 integration with the U.K. First, we determine the relationship between EU27 countries’ output and employment dynamics and their integration with the U.K., by regressing output and employment on several control variables and the integration index. Second, we will assess the impact on output and employment of a decline in integration, under different scenarios of the future relationship between the U.K. and the EU27.
11. We use panel cointegration techniques to estimate the long-run effect of the bilateral integration with the U.K. on output and employment. Three econometric issues arise. First, the integration index could suffer from endogeneity arising from an omitted variable bias or other sources. Second, the degree of integration reflects structural variables that are relatively slow moving (trade openness, participation into supply chains, financial integration, migration ties) and likely to affect output and employment mainly over a longer horizon. We are therefore interested in modelling the long-run relationships. One source of bias is that the index of bilateral integration with the U.K. can be confounded with the EU Single Market on countries or the overall degree of trade openness of a country. If this bias is not controlled for, the estimated effect of the index of bilateral integration with the U.K. will be unreliable. To reduce these concerns, the models will control for the trade openness variable for each country in the sample (total exports plus imports over GDP). We also further reduce endogeneity concerns by controlling for other determinants of output and employment such as a country’s domestic investment ratio, inflation rate, and total population. The model finally controls for country fixed effects to capture the influence of time invariant or other slow-moving factors that may affect the estimations. The model is formally represented as follows:
where y is the real GDP (or total employment level), I is the index of bilateral integration with the U.K., and X is a matrix of control variables (overall trade openness, domestic investment ratio, inflation rate, and total population).10 The sub-index i stands for countries, and t for the time dimension. Our main parameter of interest is −θ, which captures the long-run effect of the bilateral integration with the U.K. on EU27 output or employment.11
12. As expected, we find a positive long-run effect of integration with the U.K. on EU27 output and employment. There is a positive long-run relationship between the degree of bilateral integration with the U.K. (our synthetic index) and EU27 output and employment with a long-run semi-elasticity around 0.11 and 0.05, respectively. These results already preview a key conclusion of our paper: a decline in the level of integration, through a departure from the current EU membership arrangement, will negatively affect output and employment in the EU27.12
13. We then calibrate the change in the integration index from post-Brexit scenarios. The main goal is to answer the following question: controlling for the traditional factors that drive the bilateral integration with the U.K. (such as distance, language, common border, size), what are the effects of EU membership, European Economic Association membership, and other Free Trade Agreements (FTAs) on the index of integration? To do this, we use a gravity model for the index of integration with the U.K. and introduce a dummy variable capturing the different economic arrangements. To ensure sufficient variability to reflect alternative trade relationships, the sample of countries is extended to non-EU countries but excludes low-income countries.13 More specifically, the equation takes the following form:
where I is the bilateral index of integration with the U.K., X is the matrix of gravitational factors (bilateral distance vis-à-vis the U.K., common border with the U.K., common language with the U.K., regional dummies, population size and GDP level). We also control for year fixed effects to capture the influence of common shocks.
14. The parameters of interest are the ones associated with each economic arrangement that exist between the U.K. and its trading partners. We have grouped them into three dummy variables: EU membership (ϕ1, which denotes the effect of EU membership on the integration index), the European Economic Area arrangement currently in force with countries such as Norway and Iceland (ϕ2, which denotes the impact of the EEA membership), and a standard free trade agreement (ϕ3, which is the effect of the FTA dummy). Data on these arrangements come from Baier and Bergstrand (2009).14 From equation 2, we derive the reduction in integration with the U.K. that is consistent with some specific scenarios:15
(ϕ1 − ϕ2): This refers to the decline in integration going from EU membership to an EEA.
(ϕ1 − ϕ3): This refers to the decline in integration going from EU membership to an FTA.
(ϕ1): This refers to the decline in integration from EU membership to the default WTO.
15. The results confirm the expected hierarchy of the impact of various arrangements on integration. First, the model estimates confirm that the EU membership produces the highest degree of integration with the U.K. of all trade arrangements considered here. Second, settling on an EEA arrangement would lead to a moderate loss of economic integration compared with other alternative arrangements (an FTA). Third, the estimates of the integration shocks are statistically significant and with a relatively good precision.16
16. Results suggest a negative, but rather small effect of Brexit on EU output and employment in the long run. These impacts are derived by multiplying the various degrees of integration losses computed from Equation 2, by the long-term effect of the index of integration on output and employment estimated in Equation 1. A scenario in which access to the single market is preserved while the custom union is sacrificed (the EEA model or ‘soft Brexit’) would imply an almost zero cost (0.06 percent) for the EU as a whole, for both output and employment. In contrast, introducing more trade frictions by reverting to a standard FTA or to a no-deal outcome (WTO default) would lead to higher losses in the order of 0.8 and 1.5 percent for output, respectively. For employment, these losses would be comprised between 0.3 and 0.7 percent. These estimates on average are higher compared to previous studies that used standard CGE trade models, but are broadly similar to new studies that have augmented CGE trade models with supply chain links (e.g., Connell et al., 2017). In contrast to these previous studies, which solely modelled the effect of Brexit through trade channels, the econometric approach used in this study incorporates additional channels of integration.
Reduction in the Index of Integration due to Trade Frictions
(Decline in units; estimated from a gravity model with various controls)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff estimatesReduction in the Index of Integration due to Trade Frictions
(Decline in units; estimated from a gravity model with various controls)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff estimatesReduction in the Index of Integration due to Trade Frictions
(Decline in units; estimated from a gravity model with various controls)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff estimates
EU27: Long-Run Output Loss Due to Brexit
(in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.EU27: Long-Run Output Loss Due to Brexit
(in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.EU27: Long-Run Output Loss Due to Brexit
(in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.
EU27: Employment Loss Due to Brexit
(in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.EU27: Employment Loss Due to Brexit
(in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.EU27: Employment Loss Due to Brexit
(in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.17. These results should be interpreted with some caution. They remain conditional on the statistical power of the tests conducted and only represent average effects for the EU.17 Despite its technical appeal, the econometric estimations remain subject to statistical uncertainty. Furthermore, these results mask inevitable cross-country and cross-sector differences that reflect different exposures to the U.K. The economic uncertainty surrounding the post-Brexit period arrangement is also a non-negligible factor, although uncertainty is most likely to have a short-run impact. Moreover, the results assumed only polar and rigid post-Brexit scenarios, and do not incorporate the possibility, for example, that the EU and U.K. agree on a hybrid arrangement.
D. A Model-Based Approach to Quantify Long-Term Effects of Brexit Due to Higher Trade Barriers
18. We rely on a CGE model to explore country-by-country and sector-by-sector effects. The econometric investigations performed so far have helped in producing average effects of post-Brexit scenarios on the euro area, but have not identified any heterogeneous effects on individual member states. Digging one step further, data reveals trade exposures to the U.K. also vary significantly across sectors (Figure 5). Against this background, this section aims to quantify the impacts from higher trade barriers on individual countries in the euro area as well as sectors within each country using a multi-country and multi-sector CGE model. In addition to the rich structure, the model is well suited to investigate ex-ante the implications of trade policies in counterfactual scenarios.
19. The core of the model is to infer changes in real income associated with changes in trade barriers.18 In the Armington model (a simple CGE model), there are n countries, with each supplying its own distinct goods. There are thus n goods, with country t being the only supplier of good t in fixed quantity, which corresponds to the country’s endowment of the good. A representative household in each country maximizes its utility by consuming a variety of goods subject to a budget constraint. This implies that total expenditure (i.e., goods imported from other countries including associated trade costs) must be no greater than income (i.e., revenues from exporting good). In this case, the demand for goods from other countries (i.e., trade flows) is determined by the preference, income, cost of trade (i.e., tariffs) and price of foreign goods. Market equilibrium conditions imply demand for any good j needs to equal to the supply. Hence, when there is a change in trade costs, we solve the model by finding the pattern of income changes that is consistent with the new set of bilateral trade costs while respecting market clearing conditions. From a single-country perspective, an increase in trade cost decreases the revenues from exports as other countries buy less, reducing income with knock-on effects to other countries even if trade costs have not changed for these countries. To maintain sustainable external balance over the long run, imports will also have to fall too. In the new equilibrium, households are worse off by having lower income and consuming less varieties of goods. The key insight from the Armington model carries into more complex frameworks.
Gross Exports to the U.K. by Sector for Selected EA Countries
(percent of sector gross output, 2011)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Note: DVAfinal stands for domestic value added of exports of final goods to the U.K. DVAint depicts domestic value added of exports of intermediate goods to the U.K. and consumed in the U.K. DVA_3rd depicts domestic value added of exports of goods to the U.K. then re-exported to a 3rd country. FVA depicts the foreign value added. The decomposition is based on Wang, Wei and Zhu (2013). Sources: World Input-Output Tables and IMF staff calculations.Gross Exports to the U.K. by Sector for Selected EA Countries
(percent of sector gross output, 2011)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Note: DVAfinal stands for domestic value added of exports of final goods to the U.K. DVAint depicts domestic value added of exports of intermediate goods to the U.K. and consumed in the U.K. DVA_3rd depicts domestic value added of exports of goods to the U.K. then re-exported to a 3rd country. FVA depicts the foreign value added. The decomposition is based on Wang, Wei and Zhu (2013). Sources: World Input-Output Tables and IMF staff calculations.Gross Exports to the U.K. by Sector for Selected EA Countries
(percent of sector gross output, 2011)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Note: DVAfinal stands for domestic value added of exports of final goods to the U.K. DVAint depicts domestic value added of exports of intermediate goods to the U.K. and consumed in the U.K. DVA_3rd depicts domestic value added of exports of goods to the U.K. then re-exported to a 3rd country. FVA depicts the foreign value added. The decomposition is based on Wang, Wei and Zhu (2013). Sources: World Input-Output Tables and IMF staff calculations.Gross Exports to the U.K. by Sector for Selected EA Countries
(percent of sector gross output, 2011)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Note: DVAfinal stands for domestic value added of exports of final goods to the U.K. DVAint depicts domestic value added of exports of intermediate goods to the U.K. and consumed in the U.K. DVA_3rd depicts domestic value added of exports of goods to the U.K. then re-exported to a 3rd country. FVA depicts the foreign value added. The decomposition is based on Wang, Wei and Zhu (2013). Sources: World Input-Output Tables and IMF staff calculations.Gross Exports to the U.K. by Sector for Selected EA Countries
(percent of sector gross output, 2011)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Note: DVAfinal stands for domestic value added of exports of final goods to the U.K. DVAint depicts domestic value added of exports of intermediate goods to the U.K. and consumed in the U.K. DVA_3rd depicts domestic value added of exports of goods to the U.K. then re-exported to a 3rd country. FVA depicts the foreign value added. The decomposition is based on Wang, Wei and Zhu (2013). Sources: World Input-Output Tables and IMF staff calculations.20. Our baseline model covers 34 countries and 31 sectors, assumes monopolistic competition among firms, and captures global supply chain linkages. We consider three versions of the CGE model as in Costinot and Rodriguez-Clare (2013), differing by the climate of competition among firms. The first model considers multiple countries and sectors (34 countries plus the rest of the world and 31 sectors) and tradable intermediate inputs for production to capture global supply chain linkages. It assumes perfect competition among the production firms which has been shown to provide a lower bound to the welfare effects of changes in trade costs. We then extend the model to incorporate monopolistic competition, as in Krugman (1980), which implies firm-level product differentiation of symmetric varieties. Finally, we allow for firm heterogeneity consistent with Melitz (2003) at the cost of focusing on a much smaller set of countries and sectors (10 countries and 16 sectors) to reduce computational burden. Geared with these models, we calculate the changes in real income (therefore consumption and welfare) after Brexit by defining distinct scenarios. The income loss from Brexit is obtained by comparing welfare in a scenario where the U.K. remains an EU member and in a scenario in which U.K. does not. In view of the benefit of having a more realistic market structure (i.e., monopolistic competition) and the advantage of covering a broader set of countries and sectors, in what follows, the discussion focuses on the results from the model with monopolistic competition.
Data
21. The model draws on data and assumptions from various sources:
Trade linkage data are based on the World Input-Output Database (WIOD) for the year 2011. This database aggregates the world into 40 countries and covers 35 sectors which we further aggregate into 34 countries, the rest of the world and 31 sectors consistent with the setup of the model.
Data on the applied most favored nation (MFN) tariff by the EU are taken from Dhingra et al. (2016), who calculated MFN tariff for the 31 sectors (consistent with the ones in the WIOD database) using information on tariffs from the World Trade Organization weighted by the EU and U.K. trade shares.
Non-tariff trade barriers are related to costs of differences in product regulations, legal barriers, and other transaction costs for both goods and services—several authors point out that such costs are higher than formal tariffs (Anderson and van Wincoop, 2004). The primary source for the non-tariff trade barriers between U.K. and EU trade is from the published
EU Exit Analysis Cross Whitehall Briefing paper . However, the paper does not present the estimated non-tariff trade costs on all the sectors of interest, thus we complement the published measures with the estimates provided by Berden et al. (2009, 2013). The authors calculated tariffs equivalent of non-tariff barriers between the U.S.A. and the EU trade, using econometric techniques and business survey.Trade elasticities (Figure 6) for agriculture and manufacturing sectors are from Caliendo and Parro (2015) as their estimation procedure is consistent with all quantitative trade models satisfying the sector-level gravity equations, while trade elasticity for service sectors is simply held equal to the aggregate trade elasticity of 5 following Costinot and Rodriguez-Clare (2013).19
Estimated Trade Elasticity by Sector
(percent, -epsilon)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff calculations and WP sources shown.Estimated Trade Elasticity by Sector
(percent, -epsilon)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff calculations and WP sources shown.Estimated Trade Elasticity by Sector
(percent, -epsilon)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff calculations and WP sources shown.Alternative Post-Brexit Scenarios
22. We model post-Brexit scenarios as increase in goods tariffs and non-tariff barriers for both goods and services trade. In particular, we consider two cases:
‘FTA’scenario: We assume that the U.K. leaves the single market and the customs union, but the U.K. and the EU agree on a broad free trade agreement. Specifically, the scenario assumes that tariffs on goods trade remain at zero, and non-tariff costs increase moderately. With respect to the financial sector, we calibrate the size of the non-tariff trade cost such that net exports of financial services from the U.K. to the EU fall by about 40 percent, which is broadly consistent with the assumption that London-based financial firms continue to provide some cross-border financial services based on regulatory equivalence. We have also assumed a higher increase of non-tariff trade costs on the transportation equipment sector than the other studies to reflect the complicated supply chain linkage.20 Figure 7 illustrates the assumed increase in non-tariff trade costs (in tariff equivalent terms) for different sector under the scenarios.
‘Hard Brexit’ scenario (WTO scenario): We assume that the U.K. is no longer part of the single market nor the customs union and will trade with the remaining EU countries on the WTO terms. The U.K. would apply the MFN tariffs (see Table 1) on goods imported from the EU, while the EU would apply the MFN tariffs on goods originating from the U.K. In addition, we assume that the non-tariff trade costs would increase by twice as much as in the FTA scenario for all sectors.21
Level of Non-Tariff Trade Costs in FTA Scenario
(percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff calculations and WP sources shown.Level of Non-Tariff Trade Costs in FTA Scenario
(percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff calculations and WP sources shown.Level of Non-Tariff Trade Costs in FTA Scenario
(percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Sources: IMF staff calculations and WP sources shown.UK MFN Tariff With Non-EU Countries
(Percent)
UK MFN Tariff With Non-EU Countries
(Percent)
| Sectors | Imports | Exports |
|---|---|---|
| Agriculture, Hunting, forestry and fishing | 5.9 | 5.63 |
| Mining and quarrying | 0 | 0 |
| Food, beverages and tobacco | 7.26 | 4.96 |
| Textiles and textile products, Leather | 9.58 | 9.7 |
| Wood and products of wood and cork | 2.35 | 3.62 |
| Pulp, paper, printing and publishing | 0.04 | 0.1 |
| Coke, refined petroleum and nuclear fuel | 2.69 | 2.81 |
| Chemicals and chemical products | 2.71 | 2.16 |
| Rubber and plastics | 5.35 | 5.05 |
| Other non-metallic minerals | 3.78 | 3.32 |
| Basic metals and fabricated metal | 2.05 | 1.89 |
| Machines, etc | 2.05 | 2.13 |
| Electrical and optical equipment | 1.97 | 1.55 |
| Transport equipment | 8.09 | 7.22 |
| Manufacturing, etc. | 1.71 | 1.69 |
| Weighted average (by EU trade) | 4.43 | 3.29 |
UK MFN Tariff With Non-EU Countries
(Percent)
| Sectors | Imports | Exports |
|---|---|---|
| Agriculture, Hunting, forestry and fishing | 5.9 | 5.63 |
| Mining and quarrying | 0 | 0 |
| Food, beverages and tobacco | 7.26 | 4.96 |
| Textiles and textile products, Leather | 9.58 | 9.7 |
| Wood and products of wood and cork | 2.35 | 3.62 |
| Pulp, paper, printing and publishing | 0.04 | 0.1 |
| Coke, refined petroleum and nuclear fuel | 2.69 | 2.81 |
| Chemicals and chemical products | 2.71 | 2.16 |
| Rubber and plastics | 5.35 | 5.05 |
| Other non-metallic minerals | 3.78 | 3.32 |
| Basic metals and fabricated metal | 2.05 | 1.89 |
| Machines, etc | 2.05 | 2.13 |
| Electrical and optical equipment | 1.97 | 1.55 |
| Transport equipment | 8.09 | 7.22 |
| Manufacturing, etc. | 1.71 | 1.69 |
| Weighted average (by EU trade) | 4.43 | 3.29 |
Results
23. Calculations from our baseline model show EU output losses of 0.2 to 0.5 percent in the “FTA” and “hard Brexit” scenario, respectively. However, the effects vary significantly across country: Ireland’s real income is estimated to fall by about 2.5 to 4 percent similar to the estimated impact on the U.K.; Netherland’s and Belgium’s real income is estimated to fall by about 0.7 and 0.5 percent, respectively, in the FTA scenario and by about 1 percent in the WTO scenario (see Figures 8 and 9).
Long-Term Impact of Brexit: FTA Scenario
(Decline in the level of output compared to a non-Brexit scenario; in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.Long-Term Impact of Brexit: FTA Scenario
(Decline in the level of output compared to a non-Brexit scenario; in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.Long-Term Impact of Brexit: FTA Scenario
(Decline in the level of output compared to a non-Brexit scenario; in percent)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.
Long-Term Impact of Brexit: WTO Scenario
(Decline in the level of output compared to a non-Brexit scenario; in percent
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.Long-Term Impact of Brexit: WTO Scenario
(Decline in the level of output compared to a non-Brexit scenario; in percent
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.Long-Term Impact of Brexit: WTO Scenario
(Decline in the level of output compared to a non-Brexit scenario; in percent
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
Source: IMF staff estimates.24. However, it is important to note that the quantitative results rest on important assumptions in the model as well as the estimated trade elasticities in the literature. For example, the model assumes linear cost function, and Dixit-Stiglitz preferences. Although these assumptions are common in macro models, they may be too restrictive to give a full representation of the reality. Moreover, the quantitative estimates hinge on the assumed trade elasticities. But as pointed out by Hummels and Hillberry (2012) it is econometrically very challenging to estimate trade elasticities, and the existing estimates can vary quite significantly across paper (McDaniel and Balistreris, 2003). Furthermore, the model does not capture some important channels through which Brexit would affect the euro area. For example, the potential relocation of U.K. subsidiaries of multinational firms is not considered. That said, Caliendo and Parro (2015) show that the CGE model in their paper does a reasonably good job in capturing the impact of tariffs changes caused by NAFTA between 1993 and 2005. And CGE model remains a cornerstone of trade policy evaluation (Baldwin and Venable, 1995; Piermartini and Teh, 2005).
E. Summary of the Results
25. The estimated impacts in both empirical approaches used in this paper fall within the range of the estimates in the literature. In the pessimistic scenario, staff estimates suggest a range of output loss of between of 0.5 and 1.5 percent over the long run and with the econometric model pointing to larger impacts than CGE model-based estimates as it considers broader channels beyond trade (Figure 10).
Comparison of Estimated Impact From Brexit for the EU27 1/
(percent deviation of real GDP from no-Brexit scenario)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
1/ Staff estimates correspond to the average effect of the euro area countries.2/ Optimistic scenario corresponds to an EEA arrangement as discussed in paragraph 16.3/ Optimistic scenario corresponds to a “FTA” type of arrangement as discussed in paragraph 22.Comparison of Estimated Impact From Brexit for the EU27 1/
(percent deviation of real GDP from no-Brexit scenario)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
1/ Staff estimates correspond to the average effect of the euro area countries.2/ Optimistic scenario corresponds to an EEA arrangement as discussed in paragraph 16.3/ Optimistic scenario corresponds to a “FTA” type of arrangement as discussed in paragraph 22.Comparison of Estimated Impact From Brexit for the EU27 1/
(percent deviation of real GDP from no-Brexit scenario)
Citation: IMF Staff Country Reports 2018, 224; 10.5089/9781484368961.002.A001
1/ Staff estimates correspond to the average effect of the euro area countries.2/ Optimistic scenario corresponds to an EEA arrangement as discussed in paragraph 16.3/ Optimistic scenario corresponds to a “FTA” type of arrangement as discussed in paragraph 22.26. The range falls within the estimates in the literature, which partly reflects uncertainty around the estimates. The CGE model used in this paper delivers estimates that are broadly similar to the results in the literature which has focused on the trade effects of Brexit (Dhingra et al., 2016; Aichele and Felbermayr, 2015; OECD, 2016; Roja-Romagosa, 2016; and Booth et al., 2016). In contrast, we find higher impacts from the econometric model which takes into account the multiplicity of possible transmission channels beyond trade. Nevertheless, the study by Connell et al. (2017) which uses a deeper CGE model with complex supply chain linkages give similar results for the EU27 as those derived from the econometric model.
F. Conclusion
27. This paper has examined the consequences of Brexit on the EU27 under various post-Brexit scenarios and using two different, complementary, approaches. Our results, which are broadly in line with recent findings in the literature, are two-fold.
28. First, Brexit would have negative effects on the EU27 as well, given the depth and the complexity of the EU-U.K. integration. Similar to various empirical studies, we find that the estimated long-term output and employment losses (in percent) for the EU27 in our study are on average lower than the corresponding losses for the U.K. estimated in the literature (Dhingra et al., 2016; Aichele and Felbermayr, 2015; OECD, 2016; Roja-Romagosa, 2016; and Booth et al., 2016). The level of output and employment are estimated to fall at most by up to 1.5 percent and 0.7 percent in the long run in the event of a ‘hard’ Brexit scenario, respectively. A ‘soft’ Brexit outcome would lead to much lower losses.
29. Second, there is significant cross-country heterogeneity. For example, very open economies such as Ireland, the Netherlands, and Belgium are among the most exposed economies to Brexit-related adverse shocks. Ireland is the only EU27 country exhibiting Brexit-related output losses of similar magnitude to those estimated for the U.K. in the literature.
Appendix. Principal Component Analysis Results
Eigen Value and Cumulative Relative Frequencies
Eigen Value and Cumulative Relative Frequencies
| Principal component | Eigen Values | Proportion | Cumulative relative frequencies |
|---|---|---|---|
| 1 | 2.99 | 0.6 | 0.6 |
| 2 | 0.98 | 0.2 | 0.8 |
| 3 | 0.88 | 0.17 | 0.97 |
| 4 | 0.10 | 0.02 | 1.0 |
| 5 | 0.04 | 0.0 | 1.0 |
Eigen Value and Cumulative Relative Frequencies
| Principal component | Eigen Values | Proportion | Cumulative relative frequencies |
|---|---|---|---|
| 1 | 2.99 | 0.6 | 0.6 |
| 2 | 0.98 | 0.2 | 0.8 |
| 3 | 0.88 | 0.17 | 0.97 |
| 4 | 0.10 | 0.02 | 1.0 |
| 5 | 0.04 | 0.0 | 1.0 |
Eigen Vectors
Eigen Vectors
| Variables | P1 |
|---|---|
| Trade in value added-to-GDP | 0.57 |
| Participation in supply chains | 0.54 |
| Service trade openness | 0.56 |
| Cross-border bank claims | 0.1 |
| Migration share | 0.27 |
Eigen Vectors
| Variables | P1 |
|---|---|
| Trade in value added-to-GDP | 0.57 |
| Participation in supply chains | 0.54 |
| Service trade openness | 0.56 |
| Cross-border bank claims | 0.1 |
| Migration share | 0.27 |
Sectoral Aggregations
Sectoral Aggregations
| WIOD sector | WIOD31 sector code | Sector’s description | Sector name |
|---|---|---|---|
| 1 | 1 | Agriculture, Hunting, Forestry and Fishing | Agriculture |
| 2 | 2 | Mining and Quarrying | Mining |
| 3 | 3 | Food, Beverages and Tobacco | Food |
| 4 | 4 | Textiles and Textile Products | Textiles |
| 5 | Leather, Leather and Footwear | ||
| 6 | 5 | Wood and Products of Wood and Cork | Wood |
| 7 | 6 | Pulp, Paper, Printing and Publishing | Paper |
| 8 | 7 | Coke, Refined Petroleum and Nuclear Fuel | Petroleum |
| 9 | 8 | Chemicals and Chemical Products | Chemicals |
| 10 | 9 | Rubber and Plastics | Plastic |
| 11 | 10 | Other Non-Metallic Mineral | Minerals |
| 12 | 11 | Basic Metals and Fabricated Metal | Basic metals |
| 13 | 12 | Machinery, Nec | Metal products |
| 14 | 13 | Electrical and Optical Equipment | Electrical |
| 15 | 14 | Transport Equipment | Transport Equipment |
| 16 | 15 | Manufacturing, Nec; Recycling | Machinery n.e.c |
| 17 | 16 | Electricity, Gas and Water Supply | Electricity |
| 18 | 17 | Construction | Construction |
| 19 | Sale, Maintenance and Repair of Motor Vehicles and Motorcycles | ||
| 20 | 18 | Retail Sale of Fuel | Wholesale trade |
| 21 | Wholesale Trade and Commission Trade, Except of Motor Vehicles and Motorcycles | ||
| 22 | 19 | Retail Trade, Except of Motor Vehicles and Motorcycles; Repair of Household Goods | Retail Trade |
| 23 | 20 | Hotels and Restaurants | Hotels and Restaurants |
| 24 | 21 | Inland Transport | Inland Transport |
| 25 | 22 | Water Transport | Water Transport |
| 26 | 23 | Air Transport | Air Transport |
| 27 | 24 | Other Supporting and Auxiliary Transport Activities; Activities of Travel Agencies | Other transport services |
| 28 | 25 | Post and Telecommunications | Postal |
| 29 | 26 | Financial Intermediation | Financial Intermediation |
| 30 | 27 | Real Estate Activities | Real Estate Activities |
| 31 | 28 | Renting of M&Eq and Other Business Activities | Other Business |
| 32 | 29 | Education | Education |
| 33 | 30 | Health and Social Work | Health |
| 34 | 31 | Other Community, Social and Personal Services | Other services |
| 35 | Private Households with Employed Persons |
Sectoral Aggregations
| WIOD sector | WIOD31 sector code | Sector’s description | Sector name |
|---|---|---|---|
| 1 | 1 | Agriculture, Hunting, Forestry and Fishing | Agriculture |
| 2 | 2 | Mining and Quarrying | Mining |
| 3 | 3 | Food, Beverages and Tobacco | Food |
| 4 | 4 | Textiles and Textile Products | Textiles |
| 5 | Leather, Leather and Footwear | ||
| 6 | 5 | Wood and Products of Wood and Cork | Wood |
| 7 | 6 | Pulp, Paper, Printing and Publishing | Paper |
| 8 | 7 | Coke, Refined Petroleum and Nuclear Fuel | Petroleum |
| 9 | 8 | Chemicals and Chemical Products | Chemicals |
| 10 | 9 | Rubber and Plastics | Plastic |
| 11 | 10 | Other Non-Metallic Mineral | Minerals |
| 12 | 11 | Basic Metals and Fabricated Metal | Basic metals |
| 13 | 12 | Machinery, Nec | Metal products |
| 14 | 13 | Electrical and Optical Equipment | Electrical |
| 15 | 14 | Transport Equipment | Transport Equipment |
| 16 | 15 | Manufacturing, Nec; Recycling | Machinery n.e.c |
| 17 | 16 | Electricity, Gas and Water Supply | Electricity |
| 18 | 17 | Construction | Construction |
| 19 | Sale, Maintenance and Repair of Motor Vehicles and Motorcycles | ||
| 20 | 18 | Retail Sale of Fuel | Wholesale trade |
| 21 | Wholesale Trade and Commission Trade, Except of Motor Vehicles and Motorcycles | ||
| 22 | 19 | Retail Trade, Except of Motor Vehicles and Motorcycles; Repair of Household Goods | Retail Trade |
| 23 | 20 | Hotels and Restaurants | Hotels and Restaurants |
| 24 | 21 | Inland Transport | Inland Transport |
| 25 | 22 | Water Transport | Water Transport |
| 26 | 23 | Air Transport | Air Transport |
| 27 | 24 | Other Supporting and Auxiliary Transport Activities; Activities of Travel Agencies | Other transport services |
| 28 | 25 | Post and Telecommunications | Postal |
| 29 | 26 | Financial Intermediation | Financial Intermediation |
| 30 | 27 | Real Estate Activities | Real Estate Activities |
| 31 | 28 | Renting of M&Eq and Other Business Activities | Other Business |
| 32 | 29 | Education | Education |
| 33 | 30 | Health and Social Work | Health |
| 34 | 31 | Other Community, Social and Personal Services | Other services |
| 35 | Private Households with Employed Persons |
References
Aichele, R., and Felbermayr, G. (2015). Costs and Benefits of a United Kingdom Exit from the European Union. Guetersloh: Bertelsmann Stiftung.
Anderson, J.E., Van Wincoop, E. (2004). Trade Costs. Journal of Economic Literature 42 (3), 691–751.
Baldwin, R.E., Venables, A.J. (1995). Regional Economic Integration. In: Grossman, G.M., Rogoff, K. (Eds.), Handbook of International Economics, Vol. 3. Elsevier.
Berden, K., J. Francois, S. Tamminen, M. Thelle, and P. Wymenga (2009): “Non-tariff Measures in EU-US Trade and Investment: An Economic Analysis,” Final report, Ecorys.
Berden, K., J. Francois, S. Tamminen, M. Thelle, and P. Wymenga (2009) (2013): “Non-tariff Measures in EU-US Trade and Investment: An Economic Analysis,” IIDE Discussion Papers 20090806, Institute for International and Development Economics.
Booth, S., Howarth, C., and Persson, M. (2015) What if...? The Consequences, Challenges and Opportunities Facing Britain Outside EU. Open Europe Report 3 (2015): 15.
Brücker H., Capuano, S. and Marfouk, A. (2013). Education, Gender and International Migration: Insights from a Panel-Dataset 1980–2010, Mimeo.
Caliendo, Lorenzo and Fernando Parro (2015) “Estimates of the Trade and Welfare Effects of NAFTA” The Review of Economic Studies, 82 (1): 1–44
Chen W, Los B, McCann P, Ortega-Argilés R, Thissen M, van Oort F. (2018). The Continental Divide? Economic Exposure to Brexit in Regions and Countries on Both Sides of the Channel. Pap Reg Sci., 97:25–54.
Connell, W, Simons, W and Vandenbussche, H. (2017). Global Value Chains, Trade Shocks and Jobs: An Application to Brexit. London, Centre for Economic Policy Research.
Constinot Arnaud and Andres Rodriguez-Clare, 2014, “Trade Theory with Numbers: Quantifying the Consequences of Globalization.” Handbook of International Economics, Vol. 4.
Dhingra Swati, Hanwei Huang, Gianmarco I. P. Ottaviano, Joao Paulo Pessoa, Thomas Sampson and John Van Reenen (2017). “The Costs and Benefits of Leaving the EU: Trade Effects,” CEP Discussion Papers dp1478, Centre for Economic Performance, LSE.
European Parliament (2017): An Assessment of the Economic Impact of Brexit on the EU27, March 2017.
Hummels, D., Hillberry, R. (2012). Trade Elasticity Parameters for a CGE Model. In: Dixon, P.B., Jorgenson, D.W. (Eds.), Handbook of Computable General Equilibrium Modeling. Elsevier.
Ignatenko, Anna., Raei, Faezeh., Mircheva, Borislava, Tulin, Volodymyr (2017). Global Supply Chains: A New Dataset and Insights for Europe, forthcoming IMF Working Paper.
OECD (2016). The Economic Consequences of Brexit: A Taxing Decision. OECD Economic Policy Papers 16 (2016): 1.
McDaniel, C., Balistreri, E.J. (2003). A Review of Armington Trade Substitution Elasticities. Integration and Trade 18 (7), 161–173.
Dhingra, S., Huang, H., Ottaviano, G., and Paulo Pessoa, G. (2016). The Costs and Benefits of Leaving the EU: Trade Effects. Centre for Economic Performance Technical Report (2016).
Rojas-Romagos, H. (2016). Trade Effects of Brexit for the Netherlands. CPB Background Document, June.
Piermartini, R., and R. Teh (2005): Demystifying Modelling Methods for Trade Policy, No. 10. WTO Discussion Paper.
Prepared by Christian Ebeke, Li Lin, Haonan Qu, Jiaqian Chen, and Jesse Siminitz (all EUR). We are grateful to Borja Gracia (EUR) for substantive comments on the paper.
Remarks by OECD Secretary General Angel Gurría at the Istanbul G20 Trade Ministers meeting, October 6, 2015.
OECD does not have a complete coverage of EA countries.
There are a number of recent estimates of the cost of Brexit on the EU27. For a thorough review, please refer to European Parliament (2017): An Assessment of the Economic Impact of Brexit on the EU27, March 2017. This review of quantitative studies suggests an average long-term impact of Brexit on EU27 output between -0.2 and -0.5 percent by 2030. Connell et al. (2017)’s new study that incorporates supply chain links between countries finds an impact of Brexit on the EU in the order of -0.4 (for the ‘soft Brexit’ scenario) and -1.4 percent (for the ‘hard Brexit’ scenario). Very recently, the consultancy groups Oliver Wyman and Clifford Chance (2018) in a
Connell et al. (2017) identified that industrial sectors such as “motor vehicles” and “machinery & equipment” could be the most affected sector in Germany in terms of value added.
We use annual data covering the period 1993–2013. Due to data availability we could not retain bilateral FDI and portfolio statistics for our index of integration. However, the remaining variables should account for bilateral integration through the financial account (e.g., cross-border banking flows statistics). The index of integration with the U.K. is computed for all countries in the world for which data on sub-components are available.
Ignatenko, Anna., Raei, Faezeh., Mircheva, Borislava, Tulin, Volodymyr (2017). Global Supply Chains: a new dataset and insights for Europe, forthcoming IMF working paper.
Brücker H., Capuano, S. and Marfouk, A. (2013). Education, gender and international migration: insights from a panel-dataset 1980–2010, mimeo.
As the database has a limited number of destination countries (20 OECD nations), it was not possible to derive reverse migration ratios from the U.K. to other destination countries beyond these 20 OECD countries, a critical piece of information needed to make the index of integration available to several countries, including non-OECD countries. Recall, the objective is to construct an index of integration with the U.K. between each country in the world and the U.K., as the index will be further used in world-wide gravity equations to derive the impacts of alternative trade arrangements.
The annual macro variables are drawn from the IMF World Economic Outlook and World Development Indicators databases.
The model is estimated for the period 1993–2013 given the availability of the index of integration. The sample is restricted to European countries.
This is consistent with recent findings by Connell et al. (2017) who showed that Brexit would adversely affect both output and employment in the EU.
Extending the regression sample to include non-EU countries helps identify the effect of variables such as the EU membership (taking 1 for EU member states and 0 in the rest of world), the European Economic Association, and other non-EU specific trade arrangements on the bilateral integration with the U.K.
In economic terms, the EEA would be close, but not identical, to the status quo for a full EU membership, with full inclusion in the single market for all four freedoms, and compliance with all EEA-relevant regulatory legislation by the EU. But it excludes membership of the EU’s custom union, as well as agricultural and fisheries policies.
Ideally, one would have also allowed for interactions between the various trade arrangements to reflect hybrid arrangements. However, there are not enough variations in the data regarding various combinations.
The Delta method is used to assess the statistical significance of the differences in coefficients.
The next section will provide detailed results at the country and sector levels using a CGE modelling approach of the effects of Brexit via the trade channel.
We defer readers to Costinot and Rodriguez-Clare (2013) on the details of the model, but focusing to illustrate the main intuitions with a simple Armington model.
There are two exceptions, we set the estimated trade elasticity on coke, refined petroleum and nuclear fuel sector to close to 0 to avoid implausible sectoral level results. In addition, we calibrated the trade elasticity for transport equipment sector to be in line with the estimates in Egger and Kaynak (2017).
We have run robustness checks using lower tariffs on the transportation equipment sector, and the results do not change significantly.
In both scenarios, we assume the U.K. and EU will transition smoothly to the new trading arrangement.
References
Albert, Cecilia and García-Serrano, Carlos and Hernanz, Virginia, 2005, “Firm-provided Training and Temporary Contracts,” Spanish Economic Review, Vol. 7, 67–88.
Albert, Cecilia and García-Serrano, Carlos and Hernanz, Virginia, 2010, “On-the-job Training in Europe: Determinants and Wage Returns,” International Labour Review, 2010, Vol. 149, issue 3, 315–341.
Banerji, A., S. Saksonovs, H. Lin, and R. Blavy, 2014, “Youth Unemployment in Advanced Economies in Europe: Searching for Solutions,” IMF Staff Discussion Note 14/11, International Monetary Fund, Washington, DC.
Boeri, T. Garibaldi, P. and E. R. Moen, 2016, “A Clash of Generations? “Increase in Retirement Age and Labor Demand for Youth,” CEPR Discussion Paper 11422. Boeri, T. and J.F. Jimeno (2016), “Learning from the Great Divergence in Unemployment in Europe during the Crisis,” Labour Economics 41, 32–46
Cazes, S., Tonin, M., 2010, “Employment Protection Legislation and Job Stability: A European Crosscountry Analysis,” International Labour Review, International Labour Organization, Vol. 149(3), pages 261–285, September.
Chen, T., J. Hallaert, A. Pitt, H. Qu, M. Queyranne, A. Rhee, A. Shabunina, J. Vandenbussche, and I. Yackovlev, 2018, “Inequality and Poverty Across Generations in Europe,” IMF Staff Discussion Note 18/01, Washington, DC.
Dolado, J.J., Ortigueira, S., and Stucchi, R. (2016), “Does Dual Employment Protection Affect TFP? Evidence from Spanish Manufacturing firms,” Journal of the Spanish Economic Association, No. 7, 421–459.
International Monetary Fund, 2017, “Inequality of Opportunity, Inequality of Income, and Long-term Growth,” IMF Country Report No. 17/236, International Monetary Fund, Washington, DC.
Leschke, J., 2015, “Recent Trends and Reforms in Unemployment Benefit Coverage in the EU,” PowerPoint presentation at the European Commission-Social Situation Monitor Seminar on Coverage of Unemployment Benefits.
OECD, 2015, “Employment Outlook.” OECD, Paris.
Vestad, O. L. (2013) “Early Retirement and Youth Employment in Norway,” Statistic Norway.
Prepared by Haonan Qu and Hanni Schoelermann. The authors are grateful for helpful collaboration and contributions from Angana Banerji at the early stage of the project. The paper has also benefited from excellent research assistance from Xiaobo Shao and Jesse Siminitz.
In this paper, the young refer to the 15–24-year age group, adults refer to the 25–74-year age group, and primeage workers are 25 to 54 years old, and older workers refer to the 55+ age group, unless otherwise specified.
The higher cyclicality reflects a number of factors, including young workers’ less job-specific skills, lower job security with a greater share of the young in temporary and part-time jobs, and even perceptions of social fairness in which the young are considered to more easily cope with unemployment than older workers who may need to support families.
The education attainment level is coded according to the International Standard Classification of Education: “Low” indicates less than primary, primary and lower secondary education; “Medium” indicates upper secondary and postsecondary non-tertiary education; and “High” indicates tertiary education.
Boeri and Jimeno (2016) Boeri et al. (2016) looked at the 2011 pension reform in Italy which increased retirement age by up to five years for some categories of workers. They found that firms that were more exposed to the increase in employment duration of senior workers significantly reduced youth hiring. Vestad (2013) also identified positive impact of early retirement of pensioners on youth employment using a micro-level dataset in Norway.
At-risk-of-poverty rate measure is from the Eurostat and presents the share of people with equivalized disposable income (after taxes and social transfers) below 60 percent of the national median.
