2.1 Introduction
Preferential trade agreements (PTAs) featuring disciplines on trade and investment in services (hitherto referred to as ‘services PTAs’Footnote 1) have expanded at breakneck speed since the early 2000s, a trend that shows no sign of abating. Over that period, services PTAs have increased at a faster pace than PTAs without services provisions. While only a few services PTAs chiefly involving a small number of more economically advanced World Trade Organization (WTO) Members had been notified during the first five years of the WTO’s existence, services PTAs have increased almost forty-fold from 2000 to 2022 (Figure 2.1). A large majority of Members at all levels of development are today parties to at least one services PTA. A number of deep PTAs featuring services disciplines are furthermore under negotiation or were recently completed but not yet notified under the WTO’s General Agreement on Trade in Services (GATS). These include the African Continental Free Trade Area (AfCFTA), the Regional Comprehensive Economic Partnership (RCEP), and the EU–Mercosur PTA.Footnote 2 Services provisions have become a standard feature of latest-generation PTAs, including within South–South agreements linking developing and least developed countries.
Number of services PTAs notified under GATS Article V, 1995–2022.
Note: New PTAs notified each year are highlighted in the upper part of the bars, while those notified in previous years are represented in the lower part of the bars, in black.
The sharp recent rise in services agreements conducted along preferential lines stands in stark contrast to the frozen state of multilateral services commitments and negotiations at the WTO, where, aside from those supplied by acceding Members, no new market opening commitments have been undertaken by WTO Members since the curtain fell on the Uruguay Round and its extended negotiations.Footnote 3
The growing embrace of services within preferential agreements reflects mounting recognition of the central importance of services to global production, employment, innovation, foreign investment, and trade. Accounting for half of global trade in value-added terms, services present increasing opportunities for developing country exports and global value chain (GVC) insertion, driven in part by the increasing digitalisation of cross-border transactions (WTO 2020; World Bank and WTO 2023).
This chapter reviews the main innovations brought about by preferential agreements in the global governance of services trade while pointing to some limitations. The contribution of services PTAs is assessed against the background of existing multilateral norms across a threefold prism: (1) architecture and liberalisation modalities, (2) rulemaking, and (3) market opening commitments.
Our analysis underscores that, while WTO+ levels of market opening commitments have been an important feature of services PTAs, progress on rulemaking has been more modest overall, though advances in some related cross-cutting areas carry potentially important implications for services trade. This is most notably the case of advances registered in the areas of e-commerce/digital trade, investment, and procurement. As regards the architecture of services agreements, services PTAs – including many concluded most recently – have produced major advances, exhibiting the role that PTAs can play as policy laboratories. Such advances are most prominent with regard to liberalisation modalities, with agreements predicated on a negative-list approach to market opening today outnumbering the positive-list approach first pioneered in the GATS, even as the latter tends to be favoured in many South–South services PTAs.Footnote 4
In recent years, services PTAs between larger economies have progressively moved away from the binary choice between the traditional positive versus negative listing models. Overall, this has resulted in the extension of some of the key features of the negative-list approach to a growing number of countries and served to further emphasise, sometimes through innovative means, the importance of sowing greater transparency on the multiplicity of measures governing market access conditions at and behind borders, across sectors and modes of supply.Footnote 5 Looking ahead, services PTAs can be expected to further consolidate this trend and move beyond the traditional dichotomy between positive- and negative-list approaches, focusing instead on maximising transparency and predictability, promoting improved regulatory governance, while also taking due account of the policy sensitivities of countries at different levels of service sector development. Ensuring that services PTAs feature scaled-up Aid for Trade provisions and resources will also prove necessary to increase the participation of lower-income countries in services trade and help them reap the full development dividends of an ever more service-centric world economy.
2.2 Key Trends in Services PTAs
The evolving practice of preferential negotiations in services reveals a range of salient recent trends. For one, an increasing number of new WTO Members have initiated, or intensified, their participation in services PTAs. This includes the United Kingdom (UK), which has not only concluded continuity agreements with trading partners to which it was previously bound through PTAs of the European Union (EU), but has also concluded preferential pacts with new preferential partners. The UK’s exit from the EU accounts for a significant share of the new services PTAs notified since the end of 2020.
Recent years have also seen Türkiye, the Pacific Islands, and countries from Eastern Europe and the Caucasus become involved in services PTAs for the first time, while others, including India, China, and Mauritius, have become more active in recent years. While the heightened involvement of developing countries in services PTAs has reduced cross-regional gaps in participation levels, imbalances nevertheless persist to a greater degree than in goods trade, as some regions, notably the Middle East, Africa, and South Asia (aside from India), are home to few notified services PTAs. Such a situation is expected to change in the coming years, particularly in Africa, where the services Protocol of the Southern African Development Community (SADC) was notified to the WTO in 2022, and where schedules of services commitments have been negotiated among the fifty-four parties to the AfCFTA. Meanwhile, countries of the Gulf region, where far-reaching unilateral reforms have been enacted in services markets, have recently notified new services PTAs, either as a regional grouping (Gulf Cooperation Council with the European Free Trade Association (EFTA) in 2022) or individually (United Arab Emirates with India in 2022).
A major new trend is the increasing involvement of developing and least developed countries (LDCs) in services PTAs. Services agreements among countries long hesitant to make commitments in the sector account today for a significant share (41%) of all notified PTAs. Reflecting a marked shift in policy attitudes among countries, agreements between developing countries have, since 2010, accounted for a growing share of services PTAs (49%), compared to developing–developed country PTAs (45%) and developed–developed country agreements (6%) (Figure 2.2).
Services PTAs by development level of parties, 2010–2022.
Note: Not including notifications of agreements to ensure the maintenance of PTAs of the UK that had been earlier notified by the EU.
Despite marked changes in the geography of services PTAs, the fact remains that, other than the AfCFTA negotiations encompassing all thirty-three of the continent’s least developed members, lower-income countries are, on the whole, party to few services PTAs in comparison with goods-only PTAs. Currently, LDC participation is mostly through agreements signed by LDCs that are part of the Association of Southeast Asian Nations (ASEAN) (Lao PDR, Cambodia, Myanmar), Pacific Island countries that are parties to the Pacific Agreement on Closer Economic Relations Plus (PACER Plus) Agreement (Kiribati, Solomon Islands, Tuvalu, Vanuatu), the Member States of the East African Community (Burundi, Democratic Republic of Congo, Rwanda, South Sudan, Tanzania, and Uganda) and, more recently, those of the SADC (Lesotho, Malawi, Mozambique).Footnote 6
Recent years have witnessed a continuation of the trend towards multi-party PTAs, sometimes dubbed mega-regional when spanning regional lines. Such agreements tend to consolidate pre-existing bilateral links or can create new links among parties. Examples include the RCEP (fifteen parties), the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) (eleven parties), PACER Plus (ten parties), SADC (eleven parties applying the Protocol on Trade in Services), as well as the AfCFTA (fifty-four parties). All of these large-membership PTAs are mostly composed of developing countries.
In the fifteen years that followed the creation of the WTO, the largest economies (developing and developed) had few preferential agreements on services linking them, with the obvious exceptions of the EU and the Members of the North American Free Trade Agreement (NAFTA). Another notable recent trend, with consequential implications for the scope and depth of agreed outcomes, is the greater number of agreements linking large economies and important services exporters. These include the India–Japan and EU–Korea Agreements notified in 2011, the China–Australia Agreement brokered in 2015, the Canada–EU Comprehensive Economic and Trade Agreement (CETA) of 2017, the CPTPP (which includes Canada, Japan, and Australia) in 2018, the EU–Japan Agreement of 2019, the EU–UK Agreement in 2021, the RCEP signed in 2022 (linking, among others, the economies of China, Japan, Korea, and Australia), as well as the Australia–India Agreement also signed in 2022.Footnote 7 Figure 2.3 illustrates the marked uptake in preferential services agreements among larger economies since 2010.
Notified services PTAs linking large developing and developed countries, before and after 2010.
This latest development holds important implications. As supported by earlier studies (Roy et al. Reference Roy, Marchetti and Lim2007; Marchetti and Roy Reference Marchetti, Roy, Marchetti and Roy2008), larger countries are likely to extract from smaller trading partners greater market access commitments than smaller economies might. Large economies can also be expected to provide greater market access when negotiating with trading partners of similar size. As will be examined in greater detail in the next sections, agreements among larger economies also tend to influence the architecture and liberalisation modalities of subsequent PTAs. Greater balance and the resulting search for compromise have favoured the emergence of innovative negotiating models and approaches.
Even though services trade relations between economies such as the United States, the EU, and China are not governed by PTAs but rather by the GATS, the question arises whether deep preferential arrangements among larger economies may actually lessen their interest in pursuing market opening bargains at the multilateral level. This may occur as the value of WTO commitments from countries with which commitments have already been secured in PTAs is reduced.
In the context of trade in goods, another consideration is the impact of multilateral negotiations on the erosion of preferences granted earlier. Indeed, one may ponder the reactions of Members when seeing their PTA partners dilute hard-fought tariff preferences by extending them on a most-favoured-nation (MFN) basis at the WTO. However, given the far greater administrative burden of maintaining parallel regulatory regimes between preferential and non-preferential trading partners, the market opening commitments that are bound in PTAs tend to be applied in most instances on a de facto MFN basis. Accordingly, the question of preference erosion should not be as significant in services as in goods trade. Seen this way, one may well conclude that the proliferation of WTO+ advances registered by PTAs and the proclivity for implementing such commitments on a non-discriminatory basis should, in principle, facilitate their multilateralisation under the GATS. Important qualitative gains in policy predictability and effective access to markets could be reaped if WTO Members were to inscribe their best PTA commitments in GATS schedules (World Bank and WTO 2023).
A final important recent trend is the progressive disengagement of the United States – the world’s leading services exporter along with the EU – from preferential market opening in goods and services trade alike. With the exception of the United States–Mexico–Canada Agreement (USMCA) that replaced the NAFTA in 2020, the last services PTA the United States notified to the WTO dates back to 2012. Since then, the United States, who was an original signatory of the Trans-Pacific Partnership (TPP) Agreement whose substantive provisions it had largely shaped, informed its trading partners in January 2017 that it would not become a party to the agreement (USTR 2017). Around the same time, negotiations on a plurilateral Trade in Services Agreement (TiSA) fizzled out, with the last meeting of participating economies taking place in December 2016 (Global Affairs Canada 2022). Recently, the United States has focused its efforts on soft law approaches to facilitating trade, boosting the resilience of supply chains and promoting regulatory convergence through the Indo-Pacific Economic Framework and the Americas Partnership for Economic Prosperity, among other similar initiatives. These initiatives do not seek to reduce tariff barriers nor pursue enhanced market access opportunities in areas such as trade in services or investment.
2.3 Innovations and Value Added in Services PTAs
Services PTAs can be analysed by assessing how they differ from – and potentially add value to – the multilateral framework of obligations and market opening commitments found in the WTO’s GATS. This can be done along three dimensions: (1) negotiating architecture and liberalisation modalities, (2) rulemaking, and (3) liberalisation commitments.
While this chapter cannot do justice to a full assessment of services PTAs under all three dimensions, the discussion that follows focuses greater attention on the issue of architecture and liberalisation modalities while touching more briefly on the other two dimensions. This chapter’s focus on the latest-generation PTAs suggests that the greatest scope for policy innovation in services PTAs rests with the choice of architecture and liberalisation modalities. With few exceptions, services PTAs have not, on the whole, revealed a marked appetite for extending the frontiers of services rulemaking. Meanwhile, continued multilateral stasis favours an ever-increasing gap in the degree of preferential versus GATS-induced market access commitments. However, as noted above, where preferential market opening advances appear possible, their tendency to be implemented on an MFN basis and the limited evidence of policy reversal once services reforms are enacted and commitments made both suggest that the latest generation of services PTAs likely produces less trade distortion than PTAs governing goods trade.
2.3.1 Architecture and Liberalisation Modalities
From the outset, services PTAs have followed two main and well-delineated approaches: positive and negative listing approaches to liberalisation commitments. The positive list – or bottom-up approach – is modelled on the GATS and provides, in essence, that market opening obligations – market access and national treatment – only apply to the sectors, subsectors, and modes of supply expressly inscribed in each party’s schedule of commitments. As a result, if a sector is not mentioned in a schedule, the obligations of market access and national treatment do not apply, and the Member retains full discretion to introduce new discriminatory or market access-impeding measures.
For sectors and subsectors that are inscribed in schedules, parties can, as under the GATS, list limitations for particular modes of supply, indicating which measures inconsistent with market access or national treatment they wish to retain the right to apply. Parties may also indicate, for a committed sector, that market access or national treatment does not apply to a particular mode of supply (i.e. the inscription ‘unbound’).
This approach to scheduling liberalisation commitments was initially followed by the EU in its PTAs with non-European partners, in the PTAs signed by EFTA and China, as well as within Mercosur and ASEAN. In traditional positive-list agreements, all elements relating to trade in services, including all modes of supply, are typically addressed in the same chapter in a manner analogous to the GATS.
The main difference between negative- and positive-list approaches is that, in the former, all sectors and subsectors are deemed open unless otherwise stated through the (negative) listing of reservations of non-confirming measures. Reversing the logic of positive listing, measures and sectors not found in reservation lists are deemed open, that is, fully compliant with relevant obligations. Obligations subject to negative listing modalities involve not only national treatment – and, in more recent agreements, market access – obligations, but typically also feature obligations on performance requirements, local presence requirements, nationality of senior management and boards of directors, as well as MFN treatment.
Reservations for non-conforming measures are typically contained in two annexes. The first annex – and default scheduling option – lists reservations for existing non-conforming measures, that is, restrictive measures currently applied. A second annex contains reservations listing the sectors or policy areas where parties can maintain and adopt new restrictive measures. The transparent listing of existing non-conforming measures under negative-list modalities marks an important departure from the GATS and positive-list PTAs, where limitations inscribed in schedules do not necessarily reflect existing measures. Because the starting point is that all sectors are covered, and because it binds and lists the bulk of existing non-conforming measures, negative-list agreements yield commitments that largely bind the regulatory status quo, thereby better ensuring against the introduction of new trade-restrictive measures and providing a higher level of transparency and predictability.
Another distinctive feature of the negative-list approach is the so-called ratchet mechanism, which ensures that any subsequent liberalisation of existing non-conforming measures listed in the first annex of reservations is automatically bound. Through such a mechanism, the PTAs’ services commitments can harvest autonomous liberalisation measures enacted after an agreement’s entry into force, preventing the kind of situation occurring under the GATS, for example, where the gap between the level of openness bound in commitments negotiated in the 1990s and the level of applied openness has gradually increased over the following decades. The gap between levels of commitments bound in PTAs and future applied levels of openness, which fuelled recourse to a ratchet, is linked to the nature of services liberalisation and services agreements. Indeed, unlike for goods, where PTAs will typically bring down customs duties to zero within agreed timelines for most products, services agreements tend to result in little de novo liberalisation,Footnote 8 arguably because services barriers are behind-the-border measures embedded within broader regulatory frameworks that are most often not the responsibility of trade authorities and whose reform is more complex (e.g. a need for new regulations to accompany liberalisation) and more challenging to implement than a modification of duty rates. In addition, since most services reforms tend to be implemented on an MFN basis, the urge to achieve them through PTAs is more contained.
A negative listing implies important architectural implications. Contrasting positive-list agreements, where all services elements and modes are addressed within the same chapter, negative-list agreements typically address different modes or areas of services trade across multiple complementary chapters dealing with cross-border trade in services, investment, movement of business persons, alongside dedicated chapters on financial services, telecommunication services, and digital trade.
Negative listing was pioneered in the NAFTA and subsequently used in agreements signed by the United States, Canada, and Mexico with various other trading partners. It is commonly found in several agreements brokered by Latin American countries. Nowadays, negative-list agreements account for a majority of notified services PTAs. Studies have highlighted that such agreements show a higher incidence of GATS+ commitments than do positive-list agreements (Roy et al. Reference Roy, Marchetti and Lim2007; Stephenson and Robert Reference Stephenson, Robert, Sauvé and Shingal2014; Stephenson Reference Stephenson2015). Other studies have found that countries whose PTAs feature a negative-list approach are more likely to engage in agreements with other countries using such a modality (Kim and Manger Reference Kim and Manger2017).
2.3.2 Recent Innovations in Negotiating Architecture and Modalities
In recent years, various innovations in negotiating architectures and modalities have been introduced, resulting in variations in the traditional positive- and negative-list approaches and the emergence of models combining elements of the two approaches in novel ways. Some of these innovations have evolved organically through experience with the traditional approaches, while others are more closely associated with the necessities of finding negotiating compromises between parties with differing interests.
Earlier on, one novel departure from the pure negative-list approach was to include in the chapter on cross-border trade in services a market access obligation modelled on Article XVI of the GATS. In other negative-list PTAs, which typically did not cover mode 3 in the services chapter but rather addressed investment in services through a chapter governing investment in goods and services alike, another improvement has been to ensure that relevant obligations of the chapter on cross-border trade in services (e.g. on domestic regulation) also apply to services supplied through foreign-owned/foreign-controlled investments.
More recently, more significant architectural innovations, focusing on liberalisation modalities, have emerged from PTAs among larger economies at the global and regional levels. A chronology of notable advances includes the following examples:
Trade in Services Agreement (TiSA, 2013–2016): The negotiation of the TiSA – a plurilateral and non-WTO-anchored negotiation among twenty-three economies, including many of the top services exporters (e.g. US, EU, Japan) – was suspended in 2016.Footnote 9 While some issues in relation to rulemaking and national schedules of commitments remained outstanding when the talks were stopped, participants had agreed on scheduling modalities involving a mix of the elements of positive-list and negative-list elements. A negative listing approach was thus to be used for national treatment commitments (i.e. applying fully to all sectors, unless otherwise indicated). Meanwhile, a positive listing approach was used for commitments on market access. This information was all contained in a schedule format resembling that of the GATS (Fefer Reference Fefer2017). As in most negative-list agreements, a standstill (i.e. binding of existing non-conforming measures) would apply for national treatment purposes, unless otherwise specified, along with a ratchet to capture autonomously decreed liberalisation measures (Villup Reference Villup2015).
The TiSA scheduling modalities were motivated, in part, by the desire of some participants to use a format that would have similarities to the GATS so as to facilitate subsequent multilateral integration. No other PTA has used such an approach, likely because there was no similar interest in possible future incorporation into the WTO. However, subsequent PTAs have, in different ways, tried to combine elements of positive and negative listing approaches.
Australia–China PTA (2015): An early and significant effort to combine elements of positive and negative listing modalities was found in the Australia–China Agreement. China had until then solely pursued positive listing for trade in services, and Australia, while having used different approaches in different contexts, had more often opted for negative listing.
The agreement’s chapter on trade in services innovates by laying out two alternative scheduling approaches, where China opted to follow a positive listing approach while Australia pursued a negative-list approach, with a standstill but without a ratchet. While Australia also uses the negative-list approach for MFN, the MFN obligation for China applies, in a positive-list manner, only to certain services listed in an annex, and, notably, without an exception for preferential treatment granted in future PTAs.
A similarly innovative approach was used for the Agreement’s investment chapter. The chapter features a broad, asset-based definition of investment (covering, e.g., portfolio investment and not only ownership/control of enterprises) but does not apply to measures covered by the trade in services chapter, which covers supply through commercial presence. In the investment chapter, both parties use a negative-list approach – with both a standstill and a ratchet – though there is still asymmetry between the two parties in that the scope of the key obligations differ. The national treatment obligation applies to both the pre- and post-establishment phases of investment in the case of Australia, but only to post-establishment for China. Meanwhile, unlike in the services chapter, the MFN obligation applies to both parties on a negative-list basis and covers pre-establishment. For both national treatment and MFN, the main difference between the two parties is that Australia has to list its non-conforming measures, while China’s existing non-conforming measures are grandfathered (i.e. bound, but not listed).
Canada–EU (2017): The CETA marked the first time that the EU used a negative-list modality for trade in services in a PTA. The chapter on cross-border trade in services, covering modes 1 and 2, uses a negative scheduling technique for the obligations of market access, national treatment, and MFN, with a standstill and a ratchet. The investment chapter, which includes coverage of investment in services,Footnote 10 uses the same approach for provisions on market access, performance requirements, national treatment, MFN treatment, and nationality of senior management and boards of directors.
The CETA also marked the first time that Canada listed reservations for non-conforming measures maintained by provincial governments, which have significant responsibilities in various service sectors under the country’s federal system of governance. While, in past agreements, including the NAFTA, liberalisation obligations applied to measures at the sub-federal level of government, existing non-conforming measures of provinces did not have to be listed and were simply grandfathered with no information provided on them. In the CETA, existing non-conforming measures of provinces are not only bound, but listed in detail in Annex I.
Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP, 2018): The CPTPP featured important innovations in services-related areas, notably by including comprehensive disciplines on state-owned enterprises (SOEs), including in services, as well as deep disciplines on digital trade in its e-commerce chapter. It also secured higher levels of market opening commitments relative to what many of the parties had agreed in their earlier agreements (Elms Reference Elms2018).
In terms of negotiating modalities, the CPTPP, like the TPP on which it was based and which was negotiated earlier, followed negative-list modalities for both cross-border trade in services and investment, as per the PTA practice of various parties. However, an interesting novelty concerns liberalisation commitments on the temporary entry of business persons, as a number of parties undertook commitments on a variable geometry or less than an MFN basis, with commitments for particular categories of business persons extended solely to parties making reciprocal commitments in similar categories.
European Union–Japan PTA (2018), UK–Japan PTA (2020), and EU–UK PTA (2021): The EU’s turn to negative listing modalities in the CETA was reiterated in the subsequently concluded EU–Japan PTA through chapters on investment liberalisation (covering service sectors) and on cross-border trade in services (for modes 1 and 2), together with a ratchet and a standstill.
The EU–UK Agreement, of significance given its commercial relevance and the broad scope of disciplines it contains, followed a similar approach, with negative listing modalities for the chapters on cross-border trade in services (modes 1 and 2) and investment. The UK’s PTA with Japan, from 2021, also pursues a negative-list approach, with a ratchet and a standstill.
UK and EU PTAs with Armenia, Moldova, and Ukraine (2019–2020): In recent PTAs that the EU and UK each concluded with certain countries of Eastern Europe and the Caucasus, a hybrid model was used, whereby a positive listing approach was followed for modes 1 and 2 (chapter on Cross-Border Trade), while a negative-list was used for mode 3 (chapter on Investment/Establishment).
Australia–Indonesia PTA (2020): The particular relevance of the liberalisation modalities of the Australia–Indonesia PTA stems from the fact that this Agreement was concluded before the RCEP, where discussions surrounding the choice of negotiating modalities had proven difficult, given that various ASEAN Members had so far stuck with a positive listing in their individual PTAs, and that PTAs negotiated by ASEAN as a whole had until then always used GATS-type approaches.
The Agreement’s chapter on trade in services, which covers all four modes of supply, uses a negative-list approach, with a standstill and a ratchet. The investment chapter, which applies to investments in services, also uses a negative-list approach but does not feature a ratchet mechanism. These aspects are novel for Indonesia, ASEAN’s largest economy.
Even though negative-list modalities are referred to in the chapters and are used for Annex I on existing non-conforming measures, Indonesia’s reservations under its Annex II use a positive listing of commitments, where a broad right to maintain and adopt non-conforming measures is spelt out.Footnote 11
Regional Comprehensive Economic Partnership (RCEP, 2022): The RCEP is the largest PTA, as it links ASEAN Members with countries of the region with which ASEAN had already concluded PTAs, namely Australia, China, Japan, Korea (Rep. of), and New Zealand. The Agreement’s commercial significance was, however, negatively impacted by India’s decision to withdraw from the negotiations prior to their conclusion.
The negotiations pitted countries that had frequently used negative listing in their earlier PTAs (Australia, Japan, Korea, New Zealand), with others that had traditionally favoured positive listing modalities: China, India, and a number of ASEAN Members. While the ASEAN Framework Agreement on Services and the PTAs that ASEAN collectively negotiated with third parties followed a GATS-type approach, Singapore has often used negative lists in its own PTAs, while Malaysia, Viet Nam, and Brunei Darussalam had used negative lists in the context of the CPTPP. Indonesia, for its part, had, as noted above, experimented with some elements of a negative-list architecture in its PTA with Australia.
For the trade in services chapter, which covered the four modes of supply, the negotiated solution to diverging positions on liberalisation modalities was akin to the approach taken in the Australia–China PTA, where each party can choose between two liberalisation modalities.Footnote 12 Those choosing positive-list modalities would make commitments on market access and national treatment (as well as on any additional commitments). Those making commitments in accordance with negative-list modalities would list non-conforming measures in relation to national treatment and market access as well as with regard to local presence requirements.
The modalities for negative listing provided for a standstill (list of existing non-conforming measures in a first annex) and a ratchet. The novelty of the approach taken in the RCEP relates to the conditions attached to the use of the positive listing modality. A first deviation from the traditional, or GATS-type, approach is that a provision provides that parties choosing to schedule on a positive-list basis could identify sectors for future liberalisation through the annotation ‘FL’. In these sectors, the limitations scheduled are to relate to existing measures (Article 8.7:3), and a ratchet applies (Article 8.7:4).Footnote 13
Another innovation within the positive-list modality concerns the inclusion of non-binding transparency lists of existing measures that are inconsistent with national treatment and market access. The Agreement (Article 8.3) provides that parties choosing to use the positive-list modalities shall either make commitments on MFN treatment (Article 8.6) or with respect to transparency lists (Article 8.10). parties choosing the ‘transparency list’ option are to make publicly available a non-binding list of existing measures that are inconsistent with market access and national treatment. The list would cover sectors in which the party had undertaken specific commitments, and the measures listed were limited to those of the central government. As with reservation lists under negative-list agreements, the elements of information in the list are to include the sector and subsector, the obligation with which the measure is inconsistent, the legal source of the measure, and a succinct description. The Agreement specifies that the sole purpose of the lists is to promote regulatory transparency, that the measures can be amended by the relevant party, and that they are not subject to dispute settlement. At the time of writing, no party had made its transparency list publicly available, in lieu of MFN commitments.
The ‘transparency list’ approach, while non-binding, can be seen as aiming to offset a perceived shortcoming of GATS-type agreements, namely that commitments undertaken – and limitations inscribed – do not necessarily reflect prevailing levels of openness and therefore that the value added of commitments in terms of transparency is limited. Indeed, even where GATS commitments may well reflect existing conditions of access at the time of entry into force, nothing indicates so.
A further significant architectural innovation of RCEP’s trade in services chapter is the binding obligation for those who used positive-list modalities to transition to a negative-list schedule within a specified time frame. A draft negative-list schedule of non-conforming measures has to be submitted to other parties no later than three years after entry into force and adopted no later than six years after entry into force. For Cambodia, Lao PDR, and Myanmar, which are LDCs, longer compliance time frames were agreed upon (up to fifteen years). In addition to the three LDCs, China, New Zealand, the Philippines, Thailand, and Viet Nam used the Agreement’s positive-list modalities for trade in services, meaning they will transition to negative listing. The other parties, including Indonesia, used the negative-list scheduling approach. As a result, at the end of the transition period, the top ten countries in the Asia-Pacific (by GDP size), with the exception of India, but with the notable inclusion of China, will all have scheduled commitments on a negative-list basis, with sectoral coverage that could, in theory, far exceed that of the same countries’ commitments in positive-list PTAs and in the GATS, and, through increased sector coverage, standstill and ratchet, mark a pivotal shift towards negative listing, given that such a model was already being used by the main developed country economies and was most common in Latin America (with the exception, for the time being, of Mercosur, though the group is now negotiating negative-list PTAs).
The RCEP’s investment chapter, which adopts a broad definition of investment but does not apply to measures covered by the chapter on trade in services, also uses negative-list modalities. All parties list non-conforming measures in this way for liberalisation obligations, including national treatment for pre-establishment. One particularity is that the ratchet mechanism for existing non-conforming measures does not apply to some parties (Cambodia, Lao PDR, Myanmar, but also Indonesia and the Philippines). For the other parties, the ratchet starts to apply five years after the Agreement’s entry into force.
Australia–India (2022): While India did not, in the end, join the fifteen-country RCEP Agreement, the bilateral PTA that it recently concluded with Australia largely replicates the main liberalisation modalities of the RCEP for services trade, in particular the obligation to transition towards negative listing.
As in the RCEP, each party can choose whether to use the positive-list or the negative-list modalities, the latter with a ratchet. A draft negative-list schedule of non-conforming measures has to be submitted to the other party no later than five years after the PTA’s entry into force, and an agreed schedule should be submitted no later than six years after entry into force. This PTA has no investment chapter.
Japan–ASEAN (2022): The Japan–ASEAN PTA was signed in 2019 and entered into force in 2020 before being notified to the WTO in 2022. It contains, along similar lines as in the RCEP, an obligation (Article 50.4) for parties to produce a non-legally binding list (central government only) of measures inconsistent with the obligations of market access and national treatment, as well as MFN treatment. The transparency lists cover sectors where specific commitments are undertaken in the PTA and in other PTAs in force, as well as, where possible, other sectors.
The obligation to produce the transparency lists has a broader scope of application than in the RCEP since it here applies to all eleven parties, which all use a positive-list approach for the listing of specific commitments. In the RCEP, the use of such lists was subject to conditions and did not apply to parties using negative listing modalities. Parties have to exchange the lists and make them public four years after entry into force. The lists are not to form an integral part of the Agreement and shall not be subject to dispute settlement.
2.3.3 Advances in Services Rulemaking
2.3.3.1 Progressing the GATS’ Unfinished Agenda
Contrasting the changes observed in the architecture and negotiating modalities used to open services markets, PTAs have, on the whole, proven less transformative as rulemaking laboratories in the services field, notably with regard to the GATS’ unfinished rulemaking agenda (Adlung Reference Adlung2006). Trade officials involved in PTA talks in services have not moved the needle markedly on key elements that the first generation of GATS negotiators confronted but failed to resolve three decades ago. As a result, and with obvious exceptions noted below, substantive provisions found in the services chapters of PTAs tend to mirror provisions found in the GATS. Thus, PTAs pursuing a GATS-type approach have tended to largely cut and paste the WTO framework into their services instruments and focus negotiations almost exclusively on scheduling matters. Meanwhile, PTAs following the negative-list approach first pioneered in the NAFTA have also revealed a marked degree of path dependency by largely replicating that original template into their different PTAs.
Emergency safeguard measures and subsidy disciplines: Regardless of the architectural choices made, PTA negotiations have for the most part not tackled the key elements of the GATS’ unfinished agenda. This is particularly so with regard to emergency safeguard measures (Article X of the GATS), where scepticism over the very need for – and the operational means of – deploying safeguard provisions in services trade has arguably deepened since the end of the Uruguay Round (Gauthier et al. Reference Gauthier, O’Brien, Spencer, Sauvé and Stern2000; Sauvé Reference Sauvé2002). Much the same can be said of calls to negotiate subsidy disciplines for services trade as mandated by GATS Article XV, where a host of political economy reasons can be adduced to explain the tepid appetite of governments to curtail their ability to lend support to fledgling services industries. This is so even in countries devoid of the fiscal means of lending such support (Sauvé and Soprana Reference Sauvé and Soprana2018).
Disciplines on non-discriminatory domestic regulation: Rather than searching for novel ways to expand the frontiers of GATS rules, a first wave of PTAs concluded after the Uruguay Round relied on hoped-for GATS developments on unfinished rulemaking challenges, with many PTAs affirming the desire of parties to incorporate by reference any advances flowing from multilateral discussions. This was notably the case for disciplines on domestic regulation mandated by GATS Article VI:4. While the latest generation of PTAs shows no progress on the issues of emergency safeguards and subsidies, more notable advances can be observed on disciplining non-discriminatory domestic regulation where the notion that market access and national treatment commitments may not be sufficient to allow service suppliers to operate effectively in foreign services markets has gained increasing currency. This has led services negotiators to use trade agreements not only as a tool to remove quantitative restrictions and discriminatory measures relating to international services trade and investment, but also to address regulatory obstacles and promote improved good governance in services markets.
As a result, significant developments on domestic regulation matters can be observed in the drafting of ‘newer generation’ PTAs. Such a development offers evidence of a parallel and iterative dialogue between multilateral and preferential attempts at rulemaking. That dialogue accelerated when, faced with a negotiating stalemate in the GATS Working Group on Domestic Regulation, a subset of like-minded WTO Members launched a plurilateral Joint Statement Initiative (JSI) on services domestic regulation (SDR) at the WTO’s Eleventh Ministerial meeting held in Buenos Aires in December 2017 with a view to facilitating trade and reducing trade costs through streamlined procedures governing licencing, accreditation and standards-making in services. Mirroring their contentious multilateral and plurilateral nature, services PTAs have, with only a few exceptions, registered little headway in incorporating necessity or proportionality tests in services trade, an area on which the December 2021 JSI outcome on SDR also remained silent.
Assessing the prevalence of disciplines on domestic regulation in a sample of seventy-four services PTAs, a recent WTO study found that the trend towards including SDR provisions in PTAs held broadly across economies at all levels of development and regions, with only low-income economies participating to a more limited extent, as may be expected given the often-weaker state of LDC regulatory regimes. Almost 40 per cent of WTO Members in the study’s sample adopted, on average, at least half of the disciplines found in the SDR JSI. The study also shows that economies have already implemented SDR measures in their regulatory frameworks, with only low-income economies having done so to a lesser degree. Averaged across all service sectors and modes of supply, parties to more than half of the PTAs under review implemented at least two-thirds of the JSI-SDR measures (Baiker et al. Reference Baiker, Bertola and Jelitto2021).
Government procurement: Considerable negotiating headway has been achieved in opening government procurement markets (including in service sectors) within PTAs, though such advances are most notable in PTAs among developed countries already party to the WTO’s plurilateral Government Procurement Agreement that covers services. Such advances are also common within PTAs conducted along North–South lines and where procurement liberalisation was often one of the preconditions set by developed countries to launch preferential negotiations with developing country partners (Anderson et al. Reference Anderson, Muller, Osei-Lah, De Leon, Pelletier, Arrowsmith and Anderson2011). A prime example of the latter trend can be found in the NAFTA. It bears recalling, however, that the progress achieved on procurement matters at the PTA level has occurred in standalone dedicated chapters governing state purchases rather than in the services chapters of PTAs.
2.3.3.2 PTA Advances Relevant to Services Trade Governance
If PTAs register uneven progress on the GATS+ rulemaking front, significant progress can be reported in so-called GATS‐X terms, that is, in agreeing to services-relevant rules not explicitly included in the GATS. As in the case of government procurement, most such advances have been made outside of the services chapters of PTAs, within dedicated separate chapters, while others relate to generic (i.e. non-services-specific) provisions calling for stepped-up levels of regulatory cooperation and to government-to-government and public–private dialogue on services issues.
Advances on new rules relating to services have often been achieved in policy areas that feature a market access component. This is notably the case for government procurement, investment, the movement of business persons, as well as specific mandates to conclude mutual recognition agreements in selected professional service sectors. New sector-specific rules (recalling GATS annexes) have also been agreed as complements to the market opening commitments some PTAs have secured in novel sectors, such as express delivery, postal, and courier services.Footnote 14
Investment: Far-reaching advances have been made in incorporating comprehensive investment disciplines covering services that span investment promotion, protection, liberalisation, dispute settlement, and, most recently (mirroring WTO-JSI developments), investment facilitation into PTAs. Such a trend started with the NAFTA in 1994, anchoring the investment protection provisions found in bilateral investment treaties into PTAs to which a liberalisation (market access) component was added. Mounting controversy over investment litigation, particularly that flowing from investor–state arbitration, has either fuelled a major revision of many substantive and procedural features of investment rulemaking or induced significant policy aversion. Both trends have found their way into latest-generation PTAs. For instance, the USMCA that replaced the NAFTA in 2020 limited investor–state dispute settlement to Mexico and the United States, and with a reduced scope and subject to additional conditions triggering its use (Bernasconi-Osterwalder Reference Bernasconi-Osterwalder2018). Meanwhile, the Investment Protocol of the AfCFTA has altogether eschewed questions of investment protection, liberalisation, and investor–state dispute settlement,Footnote 15 focusing solely on questions of investment facilitation with market opening commitments limited to mode 3 transactions governed by the Agreement’s Protocol on Trade in Services.
Preferential trade agreements brokered along North–North and North–South lines are exhibiting notable adaptability in the investment sphere, with dedicated chapters reflecting a new equilibrium between the rights and obligations of investors and host states while also embracing a greater focus on quality investment aimed at satisfying sustainability aims (OECD 2022). The ability of some PTAs to produce a more balanced set of investment disciplines stands in marked and paradoxical contrast to the repeated inability of WTO Members to take up a more ambitious investment agenda (Echandi and Sauvé Reference Echandi and Sauvé2020).
Movement of business persons: A large number of PTAs also include separate chapters on the temporary entry of business persons featuring entry commitments. Such chapters, which apply to categories of business persons in different sectors, including services, generally go beyond the GATS by including obligations in relation to transparency and procedural matters, including on the processing of applications (Stephenson and Hufbauer Reference Stephenson, Hufbauer, Chauffour and Maur2011).
Digital trade: Unlike investment, where the adoption of multilateral norms rooted in preferential practice has not proven possible, the advent of increasingly prescriptive chapters on digital trade illustrates the important role that PTAs can potentially play as rulemaking laboratories (Monteiro and Teh Reference Monteiro and Teh2017). While close to two-thirds (64 per cent) of PTAs featuring digital trade rules involve developed countries endowed with greater digital capacities, more than a third (36 per cent) can be found in agreements among developing countries (Figure 2.4). This suggests the heightened interest that some developing countries have begun to show towards the development promise of the unfurling digital transformation, though many of them, particularly in Africa, currently abstain from rulemaking efforts in this area, whether in PTAs or in the ongoing plurilateral negotiations on e-commerce at the WTO.
Parties to PTAs with digital trade provisions, by levels of development.
The question arises of whether the growing corpus of digital rules adopted by countries, both within PTAs and stand alone digital trade agreements, can be replicated multilaterally. The distinctive feature of PTAs, involving as they do partnerships of choice, has allowed significant progress to be registered on matters of digital trade governance. This includes, among other key PTA developments, acceptance of a permanent moratorium on the application of customs duties on electronic transmissions, the prohibition of requirements on the localisation of computing facilities, provisions for the protection of personal data, as well as commitments aimed at freeing cross-border data flows. All are issues at play within ongoing JSI negotiations on e-commerce at the WTO.
Disciplines on SOEs: Disciplines on SOEs represent another new frontier in PTA rulemaking. Though horizontal (i.e. non-services-specific) disciplines on competition policy matters have long been embedded in PTAs, the CPTPP introduced detailed provisions specifically targeting the potential trade-distortive actions of SOEs, including those active in services.
The CPTPP’s SOEs and Designated Monopolies chapter (Chapter 17) aims at ensuring that commercial SOEs operate in a transparent manner and in a way that does not put privately owned companies at a competitive disadvantage. More specifically, the chapter aims at ensuring that SOEs act in accordance with commercial considerations in their purchase and sale of services, and that they do not discriminate against services and service suppliers of other parties. In addition, the chapter disciplines non-commercial assistance provided directly or indirectly to SOEs if it causes adverse effects on other parties’ interests. For example, parties are prevented from causing adverse effects by providing non-commercial assistance to an SOE that supplies a service into the territory of another party through mode 1 or through an investment in the territory of another party. Adverse effects arise when, for example, the services of SOEs having received non-commercial assistance displace or impede a like service from the market of another party.
Regulatory cooperation: Preferential trade agreements have been argued to offer greater scope for making speedier headway on matters relating to regulatory cooperation in services trade, notably in areas such as services‐related standards and the mutual recognition of licenses and professional or educational qualifications (Hoekman and Mattoo Reference Hoekman and Mattoo2007; Mattoo Reference Mattoo2015). The evidence in this area remains somewhat mixed. Regulatory harmonisation induced by PTAs beyond de minimis thresholds remains a rarity, and the conclusion of trade facilitating mutual recognition agreements has proven challenging even within the more comfortable confines of preferential partnerships of choice and within economies characterised by broadly convergent regulatory ecosystems (Sauvé Reference Sauvé1995).
The question thus arises of whether PTAs facilitate regulatory convergence. To the limited extent that negotiated convergence occurs, it tends to proceed far more under ‘closed’ Article V (Economic Integration) agreements than through the open regionalism called for under GATS Article VII (Recognition). Furthermore, with only a few exceptions, progress on regulatory issues tends to be less pronounced in trans‐regional PTAs. This suggests stronger returns to geographical proximity in matters of regulatory cooperation (Sauvé and Shingal, Reference Sauvé and Shingal2011). PTAs can, however, play a useful role in promoting dialogue between regulators, business groupings, and civil society organisations, the ‘regional public good’ benefits of which may be reaped outside of trade agreements but in a manner that nonetheless facilitates and promotes trade and investment, helps to promote improved policy outcomes, and strengthens investment climates.
2.3.4 Liberalisation Commitments
As for trade in goods, an obvious motivation for services PTAs is to allow parties to secure deeper and broader commitments on market access and national treatment than those benefiting all WTO Members under GATS schedules. Market opening commitments can liberalise services trade by reducing discriminatory measures or market access impediments to trade and investment in services. But just as importantly, such commitments can also enhance the predictability of policy regimes and the resulting conditions for business operations. By binding existing levels of openness, they will prevent trade-restrictive policy reversals. Various studies have documented the benefits deriving from higher levels of binding commitments in services, pointing to the value of PTA commitments that (merely) bind the status quo.Footnote 16
The levels of market opening secured by services PTAs and, in particular, the extent to which they go beyond GATS commitments, have been well documented in Roy et al. (Reference Roy, Marchetti and Lim2007), Marchetti and Roy (Reference Marchetti, Roy, Marchetti and Roy2008), Fink and Molinuevo (Reference Fink and Molinuevo2008), Mattoo and Sauvé (Reference Mattoo and Sauvé2010), Roy (Reference Roy, Sauvé and Shingal2014), and van der Marel and Miroudot (Reference Van der Marel and Miroudot2014). Overall, services PTAs tend to contain commitments that go well beyond those under the GATS (as well as beyond offers of improvements to GATS commitments made during the Doha Round), even if levels of GATS+ commitments naturally vary across agreements, parties, sectors, and modes of supply.
Van der Marel and Miroudot (Reference Van der Marel and Miroudot2014) found that such factors as the quality of governance, market size, skill endowments, and asymmetries between parties are relevant in accounting for GATS+ commitments in PTAs, while Shingal et al. (Reference Shingal, Roy and Sauvé2018) emphasised that the coherence and level of restrictiveness of parties’ regulatory frameworks, as well as the importance of parties’ bilateral merchandise trade had a positive impact. Roy et al. (Reference Roy, Marchetti and Lim2007) observed that countries that were parties to negative-list PTAs had undertaken greater commitments than they had under positive-list agreements.
In more recent years, the gap between market opening commitments in PTAs and the WTO has increased further, given that, as noted earlier, a host of new countries are now engaging in preferential services negotiations and in light of WTO Members’ continued inability (or absent appetite) to resume market access negotiations under the GATS.
Two other trends are at play in fuelling the gap between preferential and multilateral commitments in services trade. First, PTAs linking larger economies can be expected to yield a richer harvest of commitments compared with similar agreements reached with smaller trading partners. Indeed, as observed by Roy et al. (Reference Roy, Marchetti and Lim2007), countries that are parties to multiple PTAs tend to assume greater commitments in their agreements with the United States and, to a lesser degree, with the EU, than they do with other trading partners. Second, the greater number of negative-list agreements, including the adoption of such a negotiating modality by the EU in 2017 and by a growing number of other countries more recently, also shows a proclivity for yielding a qualitative and quantitative upgrade in services commitments over and above those found in GATS schedules.
Figure 2.5 presents a new analysis of PTA commitments in three sectors: environmental services, tourism and travel-related services, and computer services. Using the methodology developed by Roy et al. (Reference Roy, Marchetti and Lim2007), we assess the extent of GATS+ commitments in over 140 services PTAs, including some notified up to early 2022. The three sectors have different characteristics and have attracted different types of commitments under the GATS. While environmental services are subject to a relatively low level of commitments in the WTO, tourism is the most committed sector. For its part, the computer service sector, more than the other two, is subject to a relatively low level of market access and national treatment limitations and is highly reliant on cross-border supply (mode 1).
GATS+ commitments on environmental services, computer services, and tourism services in PTAs.
Note: On the basis of 142 of the 193 regional trade agreements notified under GATS Article V as of 1 March 2022. Counting the EU as one. ‘GATS/GATS offer – unimproved in PTA’ means the number of Members that have GATS commitments or that have made an offer in the WTO services negotiations in the relevant subsector, and that have not taken better commitments in PTAs. ‘PTA – Improved’ means the number of Members that have undertaken a commitment in PTAs that improve a GATS commitment or offer. ‘PTA – New’ means the number of Members that have undertaken a commitment in PTAs, where no commitment or offer had been made under the GATS. For computer services: 1.B.a = consultancy services related to the installation of computer hardware; 1.B.b = software implementation services; 1.B.c. = data processing services; 1.B.d. = data base services; 1.B.e. other computer services. For environmental services: 6.A. = sewage services (CPC 9401); 6.B. = refuse disposal services (CPC 9402); 6.C. = sanitation and similar services (CPC 9403); and 6.D. = other environmental services. For tourism and travel-related services: 9.A. = hotels and restaurants; 9.B. = travel agencies and tour operators services; 9.C. = tourist guides services.
Despite the differing characteristics noted above, PTA commitments in all three sectors largely exceed GATS commitments and Doha Round offers. In travel and tourism services, fifty-nine WTO Members (counting the EU as one) undertook greater commitments than in their GATS schedule or Doha Development Agenda (DDA) offer. Given that a large majority of WTO Members already have some commitments in the sector, PTA improvements have mostly taken the form of binding better levels of treatment in subsectors already committed in GATS schedules. This is particularly the case for the subsectors of hotels and restaurants and of travel agencies/tour operators, which had attracted a much higher level of commitment at the multilateral level than in the case of tourist guide services.
For computer and environmental services, GATS+ commitments have taken the form of better guarantees of access in sectors already committed, as well as commitments in subsectors that are uncommitted at the multilateral level. Overall, for the sample of agreements considered, fifty Members undertook GATS+ commitments in computer services, while sixty-two did so in environmental services. The rapid growth in digital trade and rising salience of concerns over environmental stewardship given evolving climate change dynamics would appear to explain the propensity for significantly improved commitments relative to when GATS commitments in both sectors were contemplated in the mid- to late 1990s.
Figure 2.6 offers a more detailed view by comparing GATS and PTA commitments in tourism and travel-related services across different modes of supply. It shows that GATS+ commitments were not concentrated in only certain modes of supply. Across the three sectors, a significant number of improvements took place under mode 1, which is of greatest significance for digital trade. For example, while 35% of GATS commitments in hotel and restaurant services were unrestricted (full commitments), the proportion jumped to 56% in the sample of PTAs under review. For travel agencies and tour operator services, full commitments were found in 73% of PTAs reviewed, compared to 62% when looking at GATS commitments/offers, or 56% when considering only existing GATS commitments. For tourist guide services, full commitments were found in 77% of PTAs reviewed, a 15% premium over the GATS.
GATS+ commitments on tourism services in PTAs, by mode of supply.
Note: See information on Figure 2.5.
As noted above, the significant and growing gap in services commitments between the WTO and PTAs does not carry, in principle, the same risks of trade diversion or preference erosion as is in goods trade. Indeed, unlike goods trade, where PTAs aim to eliminate tariffs over time for substantially all trade, services PTAs chiefly tend to bind the status quo, although examples of agreements bringing about de novo liberalisation in certain areas do exist (Roy et al. Reference Roy, Marchetti and Lim2007). Services PTAs hold lesser negative consequences for non-parties because, again in contrast to goods trade, services trade liberalisation is typically implemented on an MFN basis given the challenge and administrative costs of maintaining differentiated regulatory regimes. However, cases of applied discrimination do exist, including as a result of PTAs, such as in the case of differentiated screening thresholds for foreign investment. The movement of natural persons is another area where the potential for applied preferences exists, though comprehensive information documenting such practices is often lacking.
Given that most PTA commitments bind existing levels of openness and that such applied levels do not distinguish across different foreign services and suppliers to the same extent as for goods trade, WTO Members should, in principle, find it easier to formally extend their PTA commitments on an MFN basis under the GATS. The recent broadening and deepening of PTA commitments – new actors engaging in services PTAs, increasing links between larger economies – suggests that the basket of potential services commitments that could be multilateralised is significant and growing. This recalls the urgency of reviving the WTO’s market opening machinery in services, if only to consolidate what has been achieved preferentially if appetites for rolling back existing restrictive practices continue to prove deficient (World Bank and WTO 2023).
2.4 Concluding Remarks: What Future Directions for Services PTAs?
This chapter highlighted how advances in the services chapters of PTAs have chiefly taken the form of innovations with respect to architecture and negotiating modalities, as well as GATS+ advances in the level of bound commitments. Services PTAs, on the whole, display fewer notable advances regarding the rules governing services trade, though notable PTA progress on issues relevant to the functioning and external competitiveness of services markets has been secured in the latest-generation agreements.
Evidence of GATS+ rulemaking varies considerably across PTAs, with most innovations embedded in dedicated chapters (or annexes) that target specific service sectors (e.g. chapters on telecommunications services that build on the GATS Annex on Telecommunications and/or Reference Paper on Basic Telecommunications) or through dedicated chapters that usually do not distinguish between goods and services trade, such as those dealing with e-commerce/digital trade, the movement of business persons, investment, government procurement, regulatory convergence, or SOEs. The latter shows every sign of becoming more common in future PTAs and can be expected to add value to the obligations and commitments on services embedded in future PTAs.
Advances in services induced by PTAs are on the whole less marked than in other areas of trade and investment governance, a situation that in our view shows little prospect of significant change in the coming years. This is so because services trade negotiations confront much the same challenges and complexities regardless of the settings in which negotiations proceed. Indeed, the sheer diversity of the service economy poses significant challenges to governments as they enact policies aimed at harnessing the potential contribution of services trade to inclusive economic growth and development.
The service sector’s heterogeneity implies that, regardless of the setting, whether at the domestic level or in international negotiations, policy reforms in services need to pay close attention to – and be informed by – differences in the nature and roles that various services play, in the multiplicity of ways in which they are traded, in the broad range of public policy aims their supply pursues, and in the political economy forces they put in play. Services further differ in their skill- and capital-intensity, the degree to which they are connected to other sectors, their propensity to be supplied by micro-, small, and medium enterprises (MSMEs) or by large multinational firms, and the degree to which they can be remotely supplied. Such differentiation explains why service sector governance rarely, if ever, proceeds on a one-size-fits-all basis. It also recalls why domestic reforms anchored in trade agreements typically proceed in a progressive manner, including within preferential confines.
A further layer of complexity stems from the high degree of regulatory scrutiny services transactions tend to command, including across borders. Such scrutiny reflects the ubiquity of instances in which services markets can fail to produce socially optimal outcomes in the absence of regulatory measures pursuing legitimate public policy aims, such as consumer protection, the prevention of systemic risks in financial markets, environmental degradation, or undue market concentration in network industries. All are factors that define the realm of what can be negotiated, whether on a preferential or global basis.
Heterogeneity is centrally at play in the conduct of services negotiations, which typically involve a broad and complex set of policies, regulations, lead sectoral ministries, regulatory agencies, and diverse constituencies, domestic and foreign, public and private. For these reasons, and once again regardless of the negotiating setting, care is needed in assessing the nature, pace, and sequencing of regulatory reform and market opening undertakings in services markets if they are to maximise a country’s growth and development prospects. For all the above reasons, the process of conducting service sector reforms, whether autonomously or in the context of trade agreements, tends to prove more challenging than in other sectors (Sáez et al. Reference Sáez, Taglioni, van der Marel, Hollweg and Zavacka2014).
Still, despite the limitations noted above, innovations in recent years in relation to architecture and liberalisation modalities highlight some possible avenues in the design of future services PTAs. Building on recent trends, future PTAs can be expected to keep moving beyond the traditional dichotomy between positive and negative listing modalities. Taking account of the particular parties involved, elements of the two approaches can be used and combined so as to provide for maximum transparency of existing market access conditions. These innovations might also provide inspiration for revisiting and reviving certain positive-list PTAs that were concluded one or two decades ago.
Greater use can be made of transition periods to facilitate the shift towards negative listing approaches. Agreements should also incorporate commitments to lend support to developing countries through independent technical advice to facilitate the preparation of these lists, as transition periods and proper sequencing are relevant for those countries with limited or no prior experience with negative listing.
For those agreements and parties that cannot move to negative listing, even with transition periods, certain elements inspired by negative listing agreements could still be incorporated into a positive listing structure. One such element would be to use annotations in schedules to indicate which specific commitments contain limitations that reflect the existing, applied, situation. This feature has been used in a few PTAs but remains underutilised, even in the few positive-list agreements incorporating such features. PTAs could also usefully enshrine a standstill obligation under positive-list agreements with a view to reaping status quo commitments that have greater commercial meaning and stronger signalling effects.
Perhaps most importantly, another element inspired by negative listing approaches that could be used in positive-list agreements concerns the production of non-binding lists of existing non-conforming measures embedded in PTAs for transparency purposes. Such lists may help promote periodic domestic assessments of the trade and investment incidence of prevailing regulatory conditions. An important difference between the two traditional negotiating modalities is that positive-list agreements cannot be counted on to provide information on applied restrictions, whether for sectors committed or for sectors not committed to the extent that parties are allowed to commit less than the status quo or eschew commitments altogether (i.e. remain ‘unbound’) in specific sectors, subsectors, and/or modes of supply. In contrast, negative-list agreements are generally designed to bind the status quo and detail existing non-conforming measures and the measures maintained in sectors where future regulatory immunity is sought. When the transition to negative listing is not achievable, GATS-type approaches could be supplemented by an obligation to parties to prepare, make transparently available, and periodically update non-binding lists detailing their existing non-conforming measures in different modes and sectors, whether those are committed or not.
While such non-binding lists would not prevent the introduction of new restrictions, they would nevertheless reduce uncertainty by allowing for more informed investment and trade decisions, facilitating needed dialogue between government departments and with external stakeholders, promote a better understanding and assessment of any policy changes and their impacts, and help reduce trade costs. Nowadays, it appears increasingly odd to see PTAs aiming to boost trade and investment – most of which without the expectation of seeing liberalisation commitments reopened or supplemented after their entry into force – while providing little or no information on key existing trade measures across many service sectors and modes of supply.
Since such non-binding transparency lists carry no MFN or market access implications, they could also be considered at the multilateral level as a partial remedy to the stasis of WTO services negotiations (without substituting for them nor implying the initiation of new market opening talks) and to the low level of liberalisation but also of transparency provided by existing GATS commitments, where a majority of Members have no commitments in a plurality of sectors.
Another forward-looking element of services PTAs relates to the importance of integrating services-specific development assistance – Aid for Trade – features into services agreements. The growing prevalence of services in world trade and investment – to say nothing of their predominance in production and employment, their gendered gains, and lesser aggregate environmental footprint – points to their central integrating role in the global economy and key contribution to sustainable development strategies. Yet, as highlighted above, LDCs remain much less engaged in services PTAs than are middle- and higher-income developing countries, even if their participation has increased as a result of a few recent negotiations. And such participation is set to grow markedly in the wake of the AfCFTA’s implementation, thirty-three of whose fifty-four signatories are LDCs.
This deficit in participation is troubling given the benefits that could accrue to LDCs, in light of the potential for enhancing competition, strengthening regulatory practices, boosting inward direct investment in service sectors, facilitating access to quality and affordable services inputs for their companies in different sectors, and expanding their exports and diversification prospects, notably through digital means. Agreements to which LDCs are parties, especially those that also involve more advanced economies, should feature provisions on the supply of technical assistance focused not only on the implementation of negotiated outcomes, but also on durably strengthening supply capacities with a view to diversifying baskets of competitive service exports.
Open access