As a member of the EU, the UK not only trades freely with the EU, but currently also has preferential access to around 70 countries through the EU’s Free Trade Agreements (FTAs) across the globe. The UK Government aimed to roll over these into so-called ‘continuity’ trade agreements by Brexit-day (the day the UK leaves the EU), but it was clear from the start that doing so was no easy task. And, in fact, the Government has so far made only very modest progress.
Unsurprisingly, some of the FTA partners want to use this opportunity to negotiate new bilateral trade deals with the UK rather than merely rolling over the existing ones. However, the Government’s roll-over objective became increasingly short-sighted and narrowly defined in terms of getting something in place for a potential ‘no-deal’ Brexit in the spring of 2019. In this Briefing Paper our focus is different. On the assumption that, somehow or other, the UK will have a transition period or a period of reflection lasting at least until the end of 2020, we ask whether a simple rolling-over is either feasible or desirable from a long-term economic point of view. And if rolling-over is not a realistic option, what would be the possible alternatives?
For several reasons, we pose this question in the context of services trade with South Korea. First, the UK is a services driven economy, with 80% of GDP and 45% of exports generated by the services sector, and with strongly growing services exports (4.5% increase in cross border services with non-EU countries, 2017-2018). Thus the UK’s ‘offensive interests’ (what it wants) are likely to lie in this area.
Second, the UK and South Korea signed a joint statement for trade ‘continuity’ on 10th June, 2019. The two governments plan to ratify a deal before the Brexit deadline of 31st October. We still do not know what would be an outcome of ‘continuity’. According to Mr. Fox, the agreement would be a ‘base for ambitious future FTA’ when the UK leaves the EU. This means that another negotiation will take place once the UK leaves the EU.
Third, there is an economic rationale for an FTA with Korea, since Korea is a strong potential market for UK services exports and investment. South Korea is a middle sized high income country (51.5 million population and US$29,800 GDP per capita). Korea’s services output has reached only just over half of GDP (52.8% in 2017) and productivity in the sector is low. UK cross-border services exports to Korea increased more than three-fold over 2000-17, and now amount to almost £1.9 billion. The UK is Korea’s 4th largest foreign investor, following Japan, the US and the Netherlands, and the services sector accounts for 61.6% of Korea’s inward Foreign Direct Investment (FDI) (2014). The potential for Korea’s services import liberalisation and regulatory cooperation seem, therefore, a critical area to explore as the UK develops its trade policy options.
Finally, Korea is one of only two developed countries which have bilateral FTAs in force with both the US and the EU. In the mid-2000s Korea strategically targeted the US and the EU, the two largest consumer markets in the world, in order to establish an advantageous position in global trade and sustain its economic growth. Negotiations for the Korea-US FTA (KORUS) started in 2006, one year earlier than negotiations for the EU-Korea FTA (KorEU), and finished almost two years earlier, but the FTAs entered into force only 8 months apart in 2011-12. The EU defined its negotiating objectives relative to KORUS and was keen to conclude its deal before KORUS took effect in order to head-off potential trade diversion.
Advocates of Brexit argue that having an independent trade policy is one of the key benefits of leaving the EU. If, in the future, the UK will indeed be responsible for its own trade policy then it needs to be able to make informed decisions on the best approach to achieve its strategic interests. Korea’s agreements with the EU and the US provide an opportunity to study two very different approaches to services liberalisation by two significant trading partners of comparable size. With this analysis, we aim to shed light on the potential options for the UK if rolling-over the current FTA with Korea is not an option, and consider whether this would be best pursued by modifying KorEU or negotiating a new FTA modeled on KORUS.
This paper looks at the services segment of KorEU and KORUS, comparing the approaches taken by KorEU and KORUS in services and the nature of Korea’s commitments in the two agreements. It is important to clarify at the outset that our approach involves two caveats. First, we do not look at the corresponding liberalisation commitments by the EU and the US. FTA signatories generally make commitments based on reciprocity and thus we have not analysed the influence that EU and US concessions will have had on the commitments made by Korea. Second, we compare the liberalisation in the two FTAs relative to Korea’s GATS commitments. We do not fully consider the consistency of the FTAs with the terms of GATS Article V: Economic Integration. We acknowledge that further research is needed in this area.
Although the WTO Uruguay Round (1986-1994) introduced the General Agreement on Trade in Services (GATS) into the multilateral trading system, subsequent services trade liberalisation has been almost exclusively bi- or pluri-lateral since 1994. KorEU and the KORUS took very different approaches for services trade liberalisation in at least four significant dimensions as discussed below. Nonetheless, since the KORUS negotiations had a head start, the EU was able to use KORUS as a benchmark for its objectives when it negotiated KorEU.
The first difference in approach concerns service trade commitments. While KorEU applies a positive list approach modelled on the GATS, KORUS, in principle, uses a negative list approach that was first introduced in NAFTA. In the positive list approach, parties list explicitly the sectors in which they commit liberalisation (a bottom-up approach). They schedule Market Access (MA) and National Treatment (NT) commitments in accordance with each sectoral (and sub-sectoral) column. In the negative list approach, on the other hand, parties commit to full liberalisation including future initiatives unless they register explicit reservations (a top-down approach).
It is generally believed that the negative list approach provides more transparency and predictability than the positive list approach. However, the two approaches could end up with the same degree of liberalisation, depending on the parties’ objectives. In the case of KORUS, Korea’s long reservation lists diminish the degree of liberalisation, and the way in which Korea’s limitations are inscribed in KORUS is more complex than in the positive list approach adopted in KorEU. A particular complexity is introduced through a Korean reservation included in Annex II of KORUS, in which Korea reserves the right to limit its level of Market Access liberalisation to its GATS level in the future. A similar reservation has also been included by the US as an Annex II reservation. This introduces a positive element into the negative list which means that, in practice, KORUS may be better thought of as taking a hybrid approach rather than a negative list approach.
The second difference is coverage of modes of supply. In the GATS, and KorEU, services trade is categorised separately into Mode 1 (cross-border supply); Mode 2 (Consumption abroad); Mode 3 (commercial presence); and Mode 4 (defined as ‘Presence of natural persons for business’). In contrast, KORUS makes no reference to Mode 4. This is because the U.S. Trade Representative does not have a mandate to negotiate provisions for the movement of natural persons. Also, in contrast to GATS and KorEU, KORUS combines Modes 1 and 2 in a chapter on ‘Cross-Border Services Trade’ without making a distinction between the two.
The third difference is in the application of the core disciplines. In the GATS, Market Access and National Treatment are the central disciplines of the liberalisation commitments. In KorEU, Market Access and National Treatment apply to all four modes of supply as does the GATS. KORUS, on the other hand, does not stipulate Market Access on investment, neither for services nor for goods. Accordingly, the reservation lists do not generally cover Market Access on investment. On the other hand, KORUS explicitly includes obligations relating to ‘Local presence’; ‘Performance requirements’; and ‘Senior Management and Boards of Directors’. This in itself may not constitute any new degree of liberalisation, as these obligations are typically covered under the auspices of National Treatment in GATS, but having them explicitly listed adds transparency for services suppliers.
The fourth difference is sectoral coverage and classification. KorEU applies the WTO (GATS) style of classification of services, defining commitments in terms of the 160 sub-sectors of the WTO’s Services sectoral classification list, which in turn uses the UN Central Product Classification System (CPC) to specify 675 detailed subsectors. In contrast, KORUS defines services purely by textual descriptions. In principle the difference should not matter, but the latter introduces an additional aspect of potential ambiguity compared with the former. Further, both KorEU and KORUS specify some sectors which are carved out from the Agreements. Similar to the GATS, neither KorEU nor KORUS applies to services supplied in the exercise of governmental authority and both exclude a range of air transport related services. In addition, KorEU excludes audiovisual services and national maritime cabotage.
Lastly, different standards of National Treatment are identified. KorEU applies the WTO (GATS) style National Treatment definition, with an obligation to provide National Treatment to ‘like services and like services suppliers’ of the other party. KORUS, on the other hand, introduced the concept of ‘like circumstances’. This means that each party shall provide National Treatment to services suppliers of the other party in ‘like circumstances’. The difference possibly leads to a different legal interpretation of National Treatment.
To inform our analysis we need some way of comparing Korea’s commitments in KorEU and KORUS. We use a methodology developed in Magntorn and Winters (2018) whereby every commitment or reservation is scored as either 0, 50 or 100 depending on the extent of liberalisation. The methodology is summarised here, but expanded upon in the Appendix.
We use the Provisional Central Product Classification (CPC) system from 1991 to map each commitment made in GATS, KorEU and KORUS to the relevant CPC sector. At the most detailed level, the relevant section of the CPC contains a total of 675 subclasses. The WTO services sectoral classification (W/120) summarises these detailed CPC lines into 160 sub-sectors within 12 broad sectors.
There are two differences between the approach taken in Magntorn and Winters (2018) and in this paper: both reflect complexities on which there is no clear consensus about the best approach. The first relates to the relationship between Market Access and National Treatment. While Market Access and National Treatment are scheduled as separate obligations, in reality there is a degree of overlap between them. The distinction becomes particularly problematic when a country schedules more restrictions on Market Access than on National Treatment in the same sector. The GATS itself provides no guidance on how such cases should be interpreted, and the topic has been the subject of debate among WTO members. In our previous work, Market Access and National Treatment were considered separately from each other, whereas here we consider Market Access to prevail over National Treatment in situations where the commitment in Market Access is more restrictive than in National Treatment.
The second difference relates to how the interplay between modes of supply is treated. Previously, to capture the potential impact that a limitation of one mode can have on the other modes of supply, we defined the restrictiveness of a sector as that of the most restricted of modes 1-3. In this paper, we instead use a weighted average across the three modes of supply. The weighted average aims to ensure that the modes of supply most important for a specific sector are given a larger weight. We base our weights on a Eurostat pilot study, which has provided experimental estimates of EU’s services exports by mode of supply broken down into 12 services sectors. More details, including the weights used, can be found in the Appendix.
In this section we start by comparing the extent of access granted by Korea in KorEU and KORUS with Korea’s GATS commitments, to see if any substantial improvements were made in the two FTAs compared with Korea’s third-party (GATS) policies. We then go on to compare the extent of liberalisation that Korea granted in the two Agreements.
Figure 1 gives a broad comparison of Korea’s Market Access commitments in GATS, KorEU and KORUS, summarised across the WTO’s 12 sectors. The scores refer only to modes 1-3, and use weighted averages across modes for each subsector, and simple averages across the relevant CPC subsectors. The commitments in mode 4 are quite different in nature from those in the other three modes, and we therefore discuss them separately in a later section.
Scores range from 0 to 100, where 0 shows that a sector is fully unbound (potentially fully closed), and 100 indicates that a sector is fully committed (fully open). While figure 1 shows only Korea’s MA commitments, the scores for NT follow a similar pattern. One caveat should be kept in mind: the structural differences between the agreements, and the limitations of the scoring method, prevent us from capturing variations in the degree of restrictiveness within each subclass (e.g. all partial restrictions are given the same score (50) irrespective of their relative degree if restrictiveness). This limits the extent to which a precise comparison can be made. The estimates should therefore be viewed as indicative rather than precise and not perfectly comparable across sectors.
Korea’s Market Access commitments exceed its GATS commitments in 7 out of 12 sectors in KorEU and 8 out of 12 sectors in KORUS. The sectors with the largest improvements in Market Access are Construction services and Recreational services in both FTAs, followed by Financial services in KorEU and Environmental services in KORUS. No substantive changes were made in Health and Social services, which remain unbound in all three agreements. Similarly, no significant improvements were made in Distribution services and Tourism services, although these services were relatively liberal to start with.
The most striking feature of Figure 1 is the similarities between the three agreements. The four most restricted sectors are the same in KorEU as in the GATS, three of which also fall within the most restricted sectors in KORUS. Similarly, three out of the four most liberal sectors in GATS fall within the most liberal sectors in KorEU and KORUS respectively.
Table 1 reports the extent of liberalisation disaggregated into 28 subsectors. Some of the sectors in figure 1 have now been broken down into several subsectors, with the numbers in brackets in column one giving the correspondence to the 12 sectors listed in Figure 1. The top and bottom quartiles (the most liberal and most restricted sectors) are highlighted in green and orange respectively. To give a sense of the sectors’ importance to the UK, the final column in table 1 reports, for groups of sectors defined in the ONS Pink Book (2018), their share in the UK’s total cross-border services exports to Korea. It is important to note that this share captures only cross-border services trade. Thus, it does not capture trade taking place through establishing a local presence (mode 3), which in 2015 accounted for 63% of all EU’s extra-EU services exports. Those sectors in which trade takes place predominantly through mode 3, such as in construction and distribution, are thus substantially understated in these figures.
Although the level of liberalisation granted by Korea in the two FTAs is similar for most sectors, the US appears to have achieved slightly more than the EU. Considering only differences of three or more points (out of 100), KORUS has a higher score than KorEU in six out of the 28 sectors, and KorEU has a higher score in only one.
While the focus of this study is on the GATS-plus achievements that the EU and US have achieved, there are also occurrences of ‘GATS-minus’ commitments in the two FTAs. A commitment is ‘GATS-minus’ if it falls short of the country’s obligations under the GATS. It has been shown that such GATS-minus commitments are relatively common in FTAs, and that agreements adopting a negative-list approach tend to have relatively more GATS-minus elements than those using a positive list.
From the Korean perspective, one such GATS-minus commitment appears to exist for marine scientific research, where both KORUS and KorEU contain reservations requiring foreign persons to obtain prior authorisation or consent before conducting research in the territorial waters or exclusive economic zones of Korea. However, no such reservation exists in Korea’s GATS schedule. Further details on other GATS-minus commitments can be found in the Appendix.
Turning to details that might be particularly relevant for the UK:
Business services: Business Services account for the largest share (37.9%) of the UK’s total cross-border services exports to Korea in 2017. A number of sub-sectors under this heading were already relatively liberal in GATS, but trade in legal services is subject to a number of restrictions. For example, only Korean-licensed lawyers registered with the Korean Bar Association can supply legal services, and only such lawyers may establish or invest in law companies. Notwithstanding this, Korea did grant some concessions to both the EU and the US equally, such as allowing EU/US law firms to establish representative offices in Korea and for EU/US lawyers to provide legal advisory services for public international law and laws of the jurisdiction in which they are licensed. In addition, under KorEU, EU lawyers are permitted to use their home country titles in their home country language instead of the more restrictive “Foreign Legal Consultant” title. According to an EU report “only a few countries in the world grant the concessions obtained by the EU to foreign law firms.”
The category ‘other business services’ includes a wide range of services such as advertising and market research, management consulting services, building cleaning services, agricultural services, printing and publishing. In some areas, such as agricultural services and building cleaning services, KorEU appears to have gone further than KORUS.
Financial Services: Another key area for the UK is Financial Services, where the UK is highly competitive internationally. This sector is heavily restricted under Korea’s GATS commitments, particularly for cross-border trade where only a few insurance services and no banking services are committed. While the two FTAs go some way towards reducing these barriers, and improving transparency, mode 1 remains highly restricted and a number of restrictions for mode 3 also remain. Although Korea’s commitments for Financial Services in KorEU and KORUS are largely identical, KorEU adopts a wider definition of financial services than KORUS, covering banks, insurers, asset managers and rating agencies.
Audio-visual Services and Recreational services: The UK has the largest audio-visual sector in Europe, and it is growing each year. Thus, while this sector currently accounts for only a small share of the UK’s exports to Korea there is potential for future growth. Similar to other EU agreements, the KorEU agreement explicitly excludes audio-visual services from the scope of the chapters on cross-border trade and investment, although the agreement contains a Protocol on Cultural Cooperation covering some aspects of audio-visual services. In contrast, KORUS includes a number of GATS+ provisions in this sector, such as relaxed content requirements for animation and movie productions by US programme providers, and permission for US persons to hold up to 100% of the equity interest in some program providers. In addition, KORUS goes further than KorEU in liberalising the recreational services sector.
We can also break trade restrictiveness down by mode of supply – see Appendix Table A1. Commitments are pretty similar in Modes 1 and 2 (‘cross-border trade’) and Mode 3 (establishment) for most sectors, but there are a few notable differences.
For Mode 1 and 2, KORUS goes considerably further than KorEU in sectors such as Audiovisual services (largely excluded in KorEU), Road Transport services (deeper commitments in KORUS) and Recreational Services (a wider range of services). KorEU scores notably higher than KORUS in Financial Services, due to a more liberal scheduling of mode 2 which, as previously discussed, is not committed as a separate mode in KORUS.
Korea’s current restrictions on FDI are high relative to other OECD countries and its FDI regulations are much more restrictive in services (e.g. media related services and telecommunication) than in manufacturing. Korea’s liberalisation in Mode 3 is generally very modest except for a few specific sectors, such as architectural and engineering related services; tourism services; and computer related services. KORUS goes considerably further than KorEU in sectors such as Air and Space transport, Recreational Services and Environmental services, while KorEU exceeds KORUS in ‘Other Business Services’ where services such as ‘building cleaning services’ have been fully committed in KorEU but not in KORUS. From the UK’s point of view, facilitating Mode 3 is a potentially important route to increasing the UK’s services inputs to Korea’s manufacturing exports (in which Korea’s comparative advantage lies).
Presence of natural persons for business: Mode 4 requires that natural persons of one party are allowed entry into, and stay in, the other party. It grants rights to travel to a country only to supply a service, usually subject to a contract negotiated prior to arrival; it does not grant access to the other party’s labour market. This is traditionally the most restricted of the four modes of service delivery. In GATS, Korea defines its mode 4 access rules in terms of ‘Intra-Corporate Transferees (ICT)’, ‘Business Visitors (BV)’, ‘Service Salespersons (SS)’ and Contractual Service Suppliers (CSS). KorEU uses largely the same definitions but denotes Intra-Corporate Transferees and Business Visitors as ‘Key Personnel’, and also includes an additional category, ‘Graduate Trainees’, which was not committed in the GATS. Since, as previously described, the KORUS does not refer to mode 4, the GATS commitments prevail in this dimension of trade between Korea and the US.
The extent of liberalisation of Mode 4 varies across sectors as shown in Appendix Table A1, while Appendix Table A2 reports the basic structure of the commitments by class of labour. Overall, ‘Contractual Service Suppliers’ is the most restricted category, committed in only a very narrow range of sectors, such as certain consultancy services, installation and repair services and architectural services. For this category, KorEU simply reaffirms the GATS commitments. Intra-Corporate Trainees, Business Visitors and Graduate Trainees are conditioned on the right of establishment, and are thus committed only where mode 3 is also committed. Entry and stay tends to be more restricted for Graduate Trainees than for ICTs and BVs.
While KorEU does improve on the coverage of mode 4 in some sectors such as ‘Legal, Accounting, Auditing, Book-keeping and Taxation services’, ‘News Agency services’, and other business services such as veterinary services, the main improvement of KorEU is in added transparency. The FTA contains expanded definitions of each category, and clarifications as to the measures which shall not be maintained or adopted (no limitations on the total number of natural persons in the form of numerical quotas or economic needs test, and no discriminatory limitations).
Taken on their own, liberalisation commitments on MA and NT are not sufficient to ensure the liberalisation of services trade. Domestic regulation is also critical for services activities. Indeed, the GATS recognised that a lack of regulatory transparency, and differences in the types of regulatory regimes, could each constitute de facto trade barriers, but it was not able to develop disciplines to address them. As a result, FTAs have come to be used to develop GATS-plus functions in reducing unnecessary regulatory incompatibilities between parties. This is especially true of recently concluded FTAs, such as the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP); the EU-Canada Comprehensive Economic and Trade Agreement (CETA); and the EU-Japan Economic Partnership Agreement, which pay much attention to promoting regulatory transparency, coherence and cooperation.
Albeit of a previous generation of FTAs, KorEU and KORUS also tackle regulatory incompatibilities. Both FTAs have stand-alone chapters on establishing mechanisms to ensure regulatory transparency, covering the whole agreement. Both KorEU and KORUS provide GATS-plus (higher level) and GATS-extra (beyond the scope of the GATS) disciplines on regulatory coherence and cooperation. However, they take different approaches to regulatory transparency disciplines, coherence and cooperation, reflecting the different regulatory culture, and regimes, in the EU and the US.
KorEU has a stand-alone section which addresses regulatory frameworks in services. A ‘Working Group on MRAs (Mutual Recognition Agreements)’ was established to promote MRAs between the EU and Korea. In addition to general provisions, disciplines were developed in detail in five selected sectors (computer services; postal and courier services; telecommunication services; financial services; and international maritime transport services). For example, a subsection of postal courier services defines the principles of the regulatory framework necessary to ensure competition in the sector.
In KORUS, general disciplines on domestic regulation and recognition apply only to cross-border services. Here professional services and express delivery have detailed provisions on domestic regulations. For example, the Working Group on Professional Services was established to promote mutual recognition and licensing for engineering services, architectural services and veterinary services. The stand-alone chapters on financial services and telecommunication services separately develop disciplines and mechanisms for regulatory coherence and cooperation in detail.
Both KorEU and KORUS have a stand-alone chapter on electronic commerce, but in KORUS it is more ambitious. KorEU’s chapter is relatively simple, merely confirming the absence of customs duties on ‘deliveries by electronic means’ and identifying the areas where future dialogue on regulatory cooperation is needed (e.g. electronic signature, consumer protection and paperless trading). Issues such as the non-discrimination principle; cross-border data flow; and personal information protection, are not covered in KorEU, not least because KorEU predated the General Data Protection Regulation (GDPR) by ten years. In comparison, the coverage of electronic commerce provisions in KORUS is wider and the disciplines provided there are deeper. For example, KORUS stipulates how to cooperate in the areas of electronic signature, consumer protection and paperless trading.
Although it is technically difficult to compare the achievements of the EU and the US in terms of services trade liberalisation with Korea, on the whole, the US appears to have targeted what it wanted well, by taking a sectoral approach and extending the disciplines for liberalisation through a largely negative list approach. The US also focused on maximising Korea’s liberalisation through regulatory cooperation in areas of its specific interest, such as professional services, financial services and electronic commerce.
In contrast, KorEU could not get much beyond KORUS despite the fact that the EU negotiated KorEU using KORUS as a benchmark. At least part of the explanation could be that the EU had to negotiate on behalf of 28 different countries, with differing priorities, each of which had competence for some aspects of services policy. In an FTA, therefore, each EU Member State determines some of its own services commitments, which makes it more difficult for the European Commission to target EU offensive interests and make concessions and trade-offs on Korea’s.
These observations imply two possible options for a future Korea-UK trade deal if rolling-over is not an option: to negotiate a new FTA building on KorEU or to negotiate a new FTA from scratch, based on KORUS. In the first option, the UK would apply the positive list approach and try to fill a liberalisation gap between KorEU and KORUS to reflect its own export interests, such as including audiovisual services. In addition, the UK might be able to deepen regulatory cooperation.
In the second option, the UK would apply the negative list approach and essentially try to replicate what the US has achieved. A benefit of the second option would be more flexibility to focus on the sectors where the UK’s interests lie, since the UK would not be directly bound by the existing KorEU. Also, applying the negative list approach may be welcomed by Korea since its recent FTAs with developed countries, such as Australia (2014), Canada (2015) and New Zealand (2015), all use this approach.
If the UK adopts the KORUS model, two things should be kept in mind. First, in all of the above mentioned FTAs Korea has included a comprehensive Market Access exemption through an entry in Annex II, as briefly discussed above. This constitutes a considerable departure from the negative list approach, since it gives Korea the right to revert back to its Market Access commitments in GATS for any sectors not listed in its Annexes. Interestingly, some of Korea’s earlier FTAs (pre-KORUS) such as Chile (2004) and Singapore (2006), also adopting negative lists, do not include this Market Access reservation. If possible the UK should try to avoid such a reservation being included as it has potentially large implications on Korea’s Market Access commitments. Second, KORUS’s derogations from the GATS Article V, such as exclusion of Mode 4 and the introduction of a different notion for National Treatment requirement; ‘like circumstances’ should be very carefully considered, not least for the possible violence they do to the rule-based trading system in general.
As far as liberalisation is concerned, two stumbling blocks lie ahead. The first is the UK’s lack of negotiating power. The EU and the US were the world’s two largest markets, and as such had a great deal of economic power and influence in trade negotiations. On its own, the UK is a much smaller player, and would likely struggle to extract the same depth and breadth of commitments as were achieved by the EU and the US, at least without having to give additional concessions in return. Also, the political pressures on the UK Government to make a trade deal of some kind quickly significantly weaken its hand.
The second stumbling block is the Most Favoured Nation (MFN) clauses included in both KorEU and KORUS. These MFN clauses act as safeguards to ensure that preferences granted in these trade agreements are not eroded by one of the parties subsequently granting better treatment to another country in a future trade agreement. In such cases, the MFN clauses stipulate that any such further preferences must also be extended to the parties of the original agreement. Thus, where MFN applies, if Korea granted the UK better treatment than it had previously granted to either the EU or the US, it would have to extend that same treatment to those, substantially larger, parties ‘for free’. This would make it much less attractive for Korea to grant the UK any such further concessions.
The MFN clauses are not wholly ubiquitous, however. In terms of coverage, MFN applies to both cross-border trade (modes 1 and 2) and investment (mode 3), but not to mode 4. Further, since Market Access for mode 3 is not committed in KORUS, MFN is not applied in this dimension in this agreement. Moreover, Korea has made some very specific sectoral exemptions to MFN, for example with respect to fisheries, aviation, broadcasting and audio-visual services and a range of transportation services. These sectors offer some opportunity for the UK to seek deeper liberalisation.
KorEU and KORUS have taken different approaches for services liberalisation. Although KORUS looks slightly more ambitious than KorEU, its approach is more complicated and less transparent than KorEU’s. KORUS also contains some WTO-inconsistent aspects. Our comparative analysis suggests some potential areas of negotiations. For the UK, if rolling-over is not an option, alternative options would be either to negotiate a new FTA modeled on KorEU or to negotiate a new FTA modeled on KORUS. In reality, UK negotiators would face two hurdles, limited negotiating power and MFN clauses. However, the alternative options would arguably pave a way for more potential opportunities than a simple copy and paste of the existing KorEU in terms of reflecting bilateral economic relations and in achieving the UK’s political objective of pursuing an independent trade and investment policy. If rolling-over is not possible, this should be taken as a chance for the UK to be pragmatic and far-sighted, and seek areas of mutual interest to both the UK and Korea. Whatever the scenario, negotiations require a detailed understanding of the domestic services environment in order to identify what limitations exist and which should be protected (and thus listed) in any agreement. The UK Government needs to start by urgently building its trade negotiating capacities to cope with these challenges.
 At the time of writing, those continuity trade agreements signed are: Chile Association Agreement; Comprehensive Trade Agreement with Colombia, Ecuador and Peru; Eastern and Southern African States (interim) Economic Partnership Agreement; Faroe Islands FTA; Iceland and Norway Trading Arrangements, Israel Trade and Partnership Agreement, Pacific (interim) Economic Partnership Agreement: Palestinian Authority (interim political) Trade and Partnership Agreement: Switzerland Trade Agreement: and Switzerland-Lichtenstein Trade Agreement.
 Morita-Jaeger, M. (2018). New economic partnership between the UK and Japan – Does rolling over the EU-Japan EPA make sense?, UK Trade Policy Observatory Blog https://blogs.sussex.ac.uk/uktpo/2018/07/06/new-economic-partnership-between-the-uk-and-japan/
 Office of National Statistics
 World Bank World Development Indicators.
 Calculated from the ONS Pink Book 2018
 Calculated from the Data provided in WTO Trade Policy Review Korea 2017, Report by secretariat.
 Canada has the Comprehensive Economic and Trade Agreement (CETA) with the EU and the NAFTA with the US. Strictly speaking, the NAFTA is a plurilateral agreement of the three parties: Canada, Mexico and the US, not a bilateral agreement. Also the EU has Association Agreements with some countries which have FTAs with the US (e.g. Chile, Israel, Jordan and Morocco). But these do not include comprehensive agreements on services.
 Korea exercised a “Simultaneous and Multilateral FTA” strategy in the middle of 2000s. See Liou, To-Hai, L. (2008). South Korea’s FTA Strategy, Foreign Trade Review, 2008, Vol.43 (1), pp.42-65.
 Korea started negotiations with the US from June 2006 and negotiations with the EU from May 2007. The KorEU was provisionally applied in July 2011 (ratified in December 2015). The KORUS entered into force in March 2012 after long domestic political battles on both sides.
 Ho-Jin Lee (2010). The EU-Korea FTA: A Boost to Economic Recovery and a Challenge to the US, Brookings, https://www.brookings.edu/opinions/the-eu-korea-fta-a-boost-to-economic-recovery-and-a-challenge-to-the-u-s/
 Marchetti, J and Roy, M. (2012). Is there Reciprocity in Preferential Trade Agreements on Services?, WTO Staff Working Paper SRSD-2-12-16, WTO.
 Ho-Jin Lee (2010).
 Marchetti, J. and Roy, M. (2008). Services liberalization in the WTO and in PTAs, in J, Marchetti and Roy, M. (Ed.), Opening Markets for Trade in Services, Countries and Sectors in Bilateral and WTO Negotiations, Cambridge University Press, pp. 61-112.
 Adlung R. and Mamdouh, H. (2014). How to Design Trade Agreements in Services: Top Down or Bottom-Up?, Journal of World Trade, Vol.48 (2), pp.191-218.
 Reservation in Korea’s Annex II, p.10 and in US’s Annex II p. 8. See the Appendix for more on this.
 In fact, this casts doubt on KORUS’s WTO compatibility, because the GATS Article V (Economic Integration): 1 (a) stipulates that agreements should not provide for the a priori exclusion of any mode of supply.
 The EU also combined Mode 1 and 2 as ‘cross-border services trade’ in the Canada-EU FTA and EU-Japan FTA. The EU applied the negative approach for these FTAs.
 See Chapter 11: Investment in the Korea US FTA. KORUS is not unique in excluding Market Access for mode 3, this is relatively common in modern trade agreements. It could be argued that granting full National Treatment to foreign investors is largely equivalent to extending Market Access under mode 3, as quantitative restrictions affecting investment that are discriminatory in nature should also be captured by the National Treatment obligation. In the GATS, FTA signatories are called upon to eliminate substantially all discrimination “in the sense of National Treatment (Article V. 1. (b))”.
 See The Korea-EU FTA, Section B: Article 7.4. and Korea-US FTA, Article 12.1. GATS does not apply to traffic rights (Annex on Air Transport Services).
 KorEU does, however, promote cooperation in a stand-alone protocol on cultural cooperation.
 The Korea-EU FTA: Article 7.6. See Rudolf Adlung “WTO/GATS-Alien Framework Provisions in RTAs – A Closer Look”, forthcoming World Trade Review, for more on these unheralded changes to GATS terminology in a variety of FTAs.
 The Korea-US FTA, Article 11.3 and Article 12.2.
 Magntorn, J, Winters, L.A (2018). Can CETA-Plus Solve the UK’s Services Problem? UK Trade Policy Observatory Briefing Paper 18. Available at: http://blogs.sussex.ac.uk/uktpo/files/2018/03/BP18-10.209199781912044559.pdf
 United Nations, (1991). Statistical Paper series M, no. 77. Available at: https://unstats.un.org/unsd/classifications/Family/Detail/9
 Subject to Article XX:2 of the GATS
 For example: if MA is scheduled as unbound and NT as none, this could either mean that the country retains its right to discriminate regarding measures covered by MA (i.e. unbound prevails) or it could mean that the country commits not to discriminate for either MA or NT (i.e. none prevails).
 For further on this, see Muller, G (2017). Troubled relationships under the GATS: Tensions between MA (Article XVI), NT (Article XVII) and domestic regulations (Article VI), World Trade Review, 16: 4, 449-474.
 Eurostat, ‘Services trade statistics by modes of supply’, March 2019
 For this comparison we use Korea’s updated draft GATS schedule from 2005 rather than the original schedule from 1994. Because the Doha Round was never completed, the 2005 schedule never came into operation, but we can take it as a fair representation of Korea’s intentions on binding services trade at that time.
 KORUS does include a slight concession for certain health services in Free Economic Zones and the Jeju Special Self- Governing Province.
 The exception is ‘Recreational Services’ which does not fall within the 4 most restricted sector in KorUS.
 Eurostat, Services trade statistics by modes of supply, March 2019 https://ec.europa.eu/eurostat/statistics-explained/index.php/Services_trade_statistics_by_modes_of_supply
 Adlung, R., Miroudot, S., (2012), “Poison in the Wine? Tracing GATS-Minus Commitments in Regional Trade Agreements”, 46 Journal of World Trade, Issue 5, pp. 1045-1082
 UK Trade & Investment (2010), “South Korea open for Business – 100 opportunities for UK companies in South Korea following EU-South Korea Free Trade Agreement”
 Directorate-General for External Policies of the Union (2010). An Assessment of the EU-South Korea FTA. pp. 75 http://www.europarl.europa.eu/RegData/etudes/etudes/join/2010/133875/EXPO-INTA_ET(2010)133875_EN.pdf
 Due to the structural differences between the two FTAs it is sometimes difficult to compare commitments. For example, KorEU permits consulting services related to agriculture and animal husbandry, while KORUS makes a seemingly unbound reservation (Annex II p.36) against ‘services incidental to agriculture, hunting, forestry and fishing’. It is not entirely clear to us whether this should be taken to include consulting services or not.
 Directorate-General for External Policies of the Union (2010). An Assessment of the EU-South Korea FTA, pp. 76 http://www.europarl.europa.eu/RegData/etudes/etudes/join/2010/133875/EXPO-INTA_ET(2010)133875_EN.pdf
 Oliver & Ohlbaum Associates Ltd (2018). The contribution of the UK-based film, TV and TV-related industries to the UK economy, and growth prospects to 2025, https://www.mpaeurope.org/wp-content/uploads/2018/09/OO-UK-AV-sector-economic-contribution-report-FINAL-2018.09.21.pdf
 Article 7.4.1 and 7.10 of the KorEU agreement. For the purposes of scoring it has been assumed that Korea’s GATS commitments prevail in this sector in KorEU.
 Program providers not engaged in multi-genre programming, news reporting or home shopping.
 The OECD FDI Regulatory Restrictiveness Index.
 Article 7.20.1 in KorEU
 This is explicitly stated in KorEU (Art, 7.18) but not in GATS. However, it has been assumed that the same condition applies in both cases.
 For example education/experience requirements; qualification/ license requirements and procedures; and technical standards and licensing requirements
 GATS Article VI (Domestic Regulation) and Article VII (Recognition).
 Lazo, R. P. and Sauve, P. (2018). The Treatment of Regulatory Convergence in Preferential Trade Agreements. World Trade Review, 17: 4, pp. 575-607. Also, Gari, G. (2018). Recent Preferential Trade Agreements’ Disciplines for Tackling Regulatory Divergence in Services: How Far beyond GATS?. World Trade Review, 11/12/2018, pp.1-29
 The KorEU FTA, Chapter Seven, Section E: Regulatory framework.
 Kim, E. Y. (2019). E-commerce in South Korean FTAs: Policy Priorities and Provisional Inconsistencies, World Trade Review, 11/02/2019, pp. 1-14.
 Services are examples of so-called shared competences. See ‘Divisions of competences in the EU’: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=LEGISSUM%3Aai0020
 The reservation is found in Annex II p.10 in KORUS, see discussion in Appendix for more on this.
 It should be noted that KORUS uses ‘Like circumstances’ approach in MFN provisions (See Article 11.4 and Article 12.3).
 See Art. 7.8 and 7.14 in KorEU, and Art. 12.3 and 11.4 in KORUS. KORUS also includes a separate MFN clause for financial services (Art. 13.3).
 Details on all MFN reservations in KORUS and KorEU can be found in the Appendix.
Click here for the Online Appendix.