1. Introduction
These are not cheery days for trade proponents. The Doha Round has collapsed. The Trans-Pacific Partnership (TPP) is dead. The Trans-Atlantic Trade and Investment Partnership (TTIP) is badly wounded. There is however a bright spot amidst the storm clouds. The expansion of the Information Technology Agreement (ITA) in December 2015 marked the most significant liberalization under the WTO since its inception in 1995. The ITA expansion also took on a somewhat special structure in that it was a critical mass agreement.
In this article, I have two principal aims. First, I hope to give other scholars a basis for further study of the ITA expansion. Despite the need to understand how this agreement came about and what it means for trade politics and International Political Economy scholarship, to date there has been no unified account of the ITA expansion. This article provides that account and in so doing advances the scholarly literature on trade politics. To provide this account, I rely on newspapers, and the specialist press, most especially Inside US Trade, to process trace the ITA negotiations.
My second aim is to use the ITA expansion to extrapolate lessons on which dynamics make trade liberalization more likely in the contemporary context. This agreement is one of the most significant success stories in trade negotiations in the last two decades, and given that the sectoral plurilateral critical mass approach appears to be one of the primary templates for trade officials moving forward, this agreement merits scholarly attention. Hoekman and Mavroidis (Reference Hoekman and Mavroidis2015) have already explained the legal dimensions of how plurilaterals can promote trade. This article by contrast focuses on the more explicitly political dimensions of a particular kind of plurilateral, critical mass agreement, and how those political dimensions affect the prospects for moving trade liberalization forward.
In order to achieve these aims, I proceed as follows. First, I briefly explain what critical mass sectoral plurilaterals are and discuss their special features. Second, I conduct an in-depth case study of the negotiations that surrounded the ITA expansion in order to provide a factual basis for the analysis that follows and for other scholars interested in conducting further research on the ITA expansion. Third, I analyze the case to identify the crucial factors that allowed it to be successfully concluded and suggest implications this analysis has for the International Political Economy scholarly literature. Finally, I discuss the lessons that this has for ongoing trade negotiations, and for trade politics generally speaking, in an era of rising opposition to trade.
2. Critical mass agreements and how they're different
Critical mass plurilateral agreements have three special characteristics. First, they have a narrower scope and do not use a single undertaking format, which was traditionally used in the GATT and WTO rounds and involved negotiating across many sectors of economic activity. The idea was that by having these rounds be a single undertaking in which all the issues were on the table and nothing was settled until everything was agreed, concessions in one area could be exchanged for concessions in a different area and thus facilitate compromise. If country A and country B could not find a compromise on issue X, then perhaps country B could be convinced to relent on issue X if it received a concession on issue Y. Sectoral critical mass agreements by contrast generally deal with a single industry or product category.Footnote 1
The second feature of critical mass agreements is that they involve many but not all WTO members. GATT and WTO rounds have involved all WTO members in a fully multilateral way. At the other end of the spectrum are bilateral free trade agreements between two countries. Regional trade agreements sit somewhere between these two. Like regional deals, plurilateral negotiations occur between a number of countries but not all WTO members, and unlike regional deals, when plurilaterals like the ITA extend benefits to all members on a most-favored nation basis, they do not discriminate between members. Furthermore, because states are not required to accept all aspects of this kind of open plurilateral, this kind of package is seen as a smaller derogation of sovereignty than a single undertaking (Van Grasstek, Reference Van Grasstek2013: 203).
Third, in the case of the ITA expansion as well as the ongoing Environmental Goods Agreement (EGA), these plurilaterals focus just on tariff reductions. Many trade agreements deal with a vast array of non-tariff barriers such as subsidies, investment rules, and regulatory differences. Plurilaterals can be constructed to avoid most, if not all, of these issues.
Finally, it is worth noting that the ITA was conceived of as open plurilaterals. The way an open plurilateral works is that a ‘critical mass’ of major producers and importers of a product class agree among themselves to eliminate tariffs and then extend those tariff reductions to all WTO members on a most favored nation basis.Footnote 2 This means that those other members can get access to those major importers’ markets without having to lower tariffs themselves and thus effectively free-ride. This is why it is important to get the critical mass as high as possible. Doing so is how the participants keep the free riding to a minimum. The importance of this aspect of these negotiations is why they are referred to as critical mass agreements.
Additionally, because critical mass agreements extend benefits to non-participating developing countries on a most favored nation basis without them having to make concessions, they also effectively preserve the principle of special and differential treatment for developing countries.
Given the lack of progress in the Doha Round, a number of analyses of how to advance trade liberalization have highlighted the potential of agreements with sectoral, plurilateral, and/or critical mass features as a way forward. The 2007 Warwick Commission Report advocated for a more robust consideration of critical mass approaches in the future.Footnote 3 It argued that there is more precedent for this approach; a critical mass kind of logic was used in telecommunications and in zero-for-zero tariff initiatives considered in the NAMA (non-agricultural market access) part of the Doha Round as well as the Tokyo Round Codes, though in the case of the Codes, it led to some important deficiencies, in particular, the perception that the Codes had balkanized the trading system.Footnote 4
Former trade negotiator Andrew Stoler (Reference Stoler2008) has also advocated for critical mass agreements, highlighting the problems created by attempting to force negotiations to have a single undertaking format. Gallagher and Stoler (Reference Gallagher and Stoler2009) expound on this argument, pointing out that critical mass agreements helpfully quarantine the potentially deleterious effects of non-reciprocity in multilateral trade negotiations and assert that a critical mass approach could be particularly helpful in agriculture. Manfred Elsig (Reference Elsig and Steger2009: 67–80) has discussed the ways in which critical mass frameworks, in conjunction with a stronger Secretariat, could promote more fruitful negotiations.
It is also important to note that the Warwick Commission, Stoler, and Elsig all add caveats to their support for critical mass agreements. The Warwick Commission provides several criteria that critical mass agreements need to meet, such as being binding, justiciable, and with consideration given to distributional consequences.Footnote 5 Gallagher and Stoler (Reference Gallagher and Stoler2009: 388–389) note that a critical mass approach is of course not a panacea to all difficulties in trade negotiations in agriculture; it would still run into the problem of there being a wide variety of politically sensitive support programs. Elsig (Reference Elsig and Steger2009: 79–80) also makes the important point that the contracting parties to a critical mass agreement need to ensure that there are effective incentives for non-participants in those agreements to want to join rather than continually free-ride. Of course, that is much easier said than done.
In any event, because the ITA expansion was negotiated on these critical mass grounds, those negotiations offer a good opportunity to further test-drive critical mass agreements and the arguments made about them. The ITA expansion negotiations were important in their own right, but were additionally important as a demonstration of the viability and efficacy of the critical mass approach.
3. The expansion of the Information Technology Agreement
Background
The original ITA was completed in 1996 and abolished tariffs among 72 countries on a range of IT goods such as personal computers, modems, and certain kinds of semiconductors. One of the key difficulties surrounding the negotiation of tariff reductions on IT products had to do with the different ways that some products can be classified under different tariff codes. To try to alleviate this problem, the original ITA contained a section called Attachment B in which products were described in plain language rather than tariff codes so as to make it clear that those items were definitively covered by the ITA regardless of how they were classified for tariff purposes. The US position was that Attachment B was intended to make the ITA a living agreement. An additional shortcoming of the original ITA was that, though it cut tariffs on many products, there were still a number of IT related goods that the ITA did not cover. These included video game consoles, tablet devices, and smartphones.
An additional difficulty involved technological change in IT products. If a product advanced technologically, could a state reclassify that product and thus increase tariffs on it? For example, the original ITA eliminated tariffs on mobile phones but once mobiles phones had other functions such as a camera, the EU wanted to consider them multifunctional devices that would be subject to a 14% tariff. This question became the focal point of one of the most significant WTO cases involving IT products. In that case, the EU wanted to reclassify flat-panel displays, set top boxes, and multi-function fax machines based on technological changes in those product categories. Alistair MacDonald, the EU Ambassador to the Philippines (one of the counterparties in the WTO dispute), summed up the EU position – having a product covered no matter how much it morphed would be a slippery slope, arguing that if the WTO disagreed with the EU, then ‘a concession on a horse carriage will soon be understood to cover a car’ (Hermosa, Reference Hermosa2009). The US and the Philippines thought that an EU win would open the way to backdoor invalidation of the ITA because it would allow states to simply reclassify products whenever they wanted to raise tariffs on them, and this would be easy to do because technology in IT products was changing so quickly. The US and the EU had by this point already begun strategizing over how to further liberalize trade in IT goods, but neither was willing to move on the issue while this case before the WTO was ongoing.
The WTO ultimately ruled against the EU, and the EU decided to adjust its policies rather than appeal the WTO decision; the EU also called for a renegotiation of the ITA (Beattie, Reference Beattie2010). The resolution of this case was important not only because it removed a political barrier to the EU and the US seriously discussing liberalization in IT goods, but also because it galvanized the global IT industry to press for a new ITA. In March 2011, 40 industry associations joined together in urging governments to expand the ITA.Footnote 6
Not long thereafter, in April 2011, the US Trade Representative started to put together a proposal to expand the ITA in product coverage and geographic scope. The EU agreed with these aims, and also wanted to address non-tariff barriers (NTBs). NTBs were, and still are, a relatively bigger problem in Japan, and so EU producers felt they would be receiving little in return for tariff decreases if NTBs were not also dealt with.Footnote 7 In particular, EU producers, and in turn the EU, wanted to promote regulatory cooperation by having members allow electronics manufacturers to independently certify that their products met a foreign country's regulatory standards.Footnote 8 Other states however were quite strongly opposed to this, and to the inclusion of NTBs generally, because they believed it would make the negotiations much more difficult. Once it became clear that they were not going to get their negotiating partners to agree to regulatory cooperation or other rules designed to mitigate NTBs, the EU accepted the idea that NTBs would not be included in the negotiations.Footnote 9
Product coverage – early progress
Once negotiations got under way in May 2012, the central negotiating task was to decide on which products tariffs would be eliminated. This aspect of the negotiations began with states talking about exactly how countries would put together their individual list; they quickly decided on the six-digit level for tariff codes as this would be the most generally feasible level.Footnote 10
Several countries, including the US, had already developed informal lists to give other participants a sense of the products they prioritized.Footnote 11 Nine participants, including the US, the EU, Japan, South Korea, and Canada, had all tabled lists by July 2012.Footnote 12 The US had the most ambitious wish list.Footnote 13 The EU was reluctant to drop tariffs on televisions and monitors, two commercially important items, since it wanted to use that as a bargaining chip in its ongoing trade negotiations with Japan.Footnote 14 The ambition of the EU's stance was also checked by its internal division. Eastern European countries, such as Poland, were less in favor of ITA expansion than the Nordic countries.Footnote 15 China did not table a list at all.Footnote 16 Japan and South Korea generally agreed with the desires of the US and the EU and, despite other states’ skepticism, also wanted to include some items such as kitchen appliances that are not normally thought of as IT goods.Footnote 17 Even when Japan clarified its position on refrigerators, saying it only wanted to cut tariffs on high-tech refrigerators with LCD screens and other features, a number of other states, and especially China, did not seem to find that to be a meaningful distinction.Footnote 18
By November 2012, the collective wish list had swelled to 439 items, meaning that there were 439 items for which at least one state wanted to see tariffs eliminated.Footnote 19 This list notably contained the three items from the earlier WTO dispute: multi-function office machines, set-top boxes, and flat-panel displays.Footnote 20 In December 2012, the negotiating partners started to whittle down the list to those items on which agreement could be found.
Amongst the first items to be cut were the sophisticated home appliances that Japan and Korea wanted, as many states did not accept that there was a genuine IT connection.Footnote 21 These were not the only goods that had their IT relevance questioned. IT relevance significantly played into states’ considerations at this point. Of the items on the list, the US believed that 60 did not have any IT application, while China questioned the IT relevance of one-third of the listed products.Footnote 22
By January 2013, the wish list had been cut to 300 items with negotiating countries agreeing to try to cut another 30–40 items in February and an additional 30–40 in March with the aim of concluding the talks by the end of 2013.Footnote 23
Product coverage – Chinese objections
The Chinese government had begun to discuss IT trade liberalization with its domestic businesses in January 2012.Footnote 24 In the autumn of 2012, China joined the core negotiating group and, even though it still had not produced an actual wish list, immediately called for a much narrower list of goods to be liberalized.Footnote 25 China's hardline position was rooted in opposition from the Ministry of Finance (MOF), which was generally interested in maintaining revenues from tariffs, and the National Development and Reform Commission (NDRC), which was the leading agency charged with rolling out industrial policies meant to bolster nascent industries, including in the tech sector.Footnote 26
When China finally produced its official position in June 2013, it said it had sensitivities on 148 of the 256 items on the list.Footnote 27 Of those, it said it absolutely could not compromise on 106 of them, though in reality there was no real sense among negotiators of how many of these items China actually could compromise on.Footnote 28 Negotiators were surprised by the extent of this reluctance and also confused by China's decision to list some items that were already covered by the original ITA.Footnote 29 Chinese officials defended this position, as well as their later refusals to compromise, by arguing that China was a developing country and so should not be asked to agree to a deal that did not give it special treatment (Enheng, Reference Enheng2013; Jiabao Reference Jiabao2013).Footnote 30 Developed countries could accept LDCs free-riding on an ITA expansion as they were not major producers of IT goods. China however was the largest single producer of IT goods and so there was no way that developed countries could accept lowering their own tariffs without China making meaningful concessions.
China's position threw a major roadblock into the negotiations. As one official put it, ‘the mood, once optimistic, has darkened considerably’ (Politi and Donnan, Reference Politi and Donnan2013). In response, the US, the EU, and Japan suspended negotiations to pressure China to compromise on its list of non-negotiable items.Footnote 31 They strongly wanted an ITA expansion of about 200 items, which would require China to seriously cut down its list of sensitivities.Footnote 32 The IT industry also got involved, with a wide range of trade associations sending a global industry letter to Chinese Vice Premier Wang Yang urging the Chinese government to pursue an expanded ITA and to seriously reduce its sensitivities list (Perez, Reference Perez2013a).
In October 2013, the Chinese government softened its stance somewhat by saying that there were only about 60 items that it wanted excluded altogether and it might accept long phase-outs on the rest, which led to a restart of negotiations, but China still frustrated other negotiating partners by demanding so many exclusions and long phase-outs on items such as semiconductors and flat panel displays that were high-priority items for those other states.Footnote 33
As a point of comparison, whereas China was arguing that it could not accept expeditious liberalization on roughly 140 items, the US took that position on just one item (fiber optic cables) while Japan took that position on zero items under consideration (Perez, Reference Perez2013b). As one trade official put it, ‘we have a situation today where there is just a broad frustration that China is not ready to negotiate this … they are the world's biggest beneficiary of IT exports. Other people look at that and say there's got to be a little bit of reciprocity there’ (Donnan, Reference Donnan2013a).
When China refused to budge from this position, the talks once again collapsed in November 2013 (Donnan, Reference Donnan2013b). What particularly frustrated US negotiators was that China was clearly making spurious claims of import sensitivity on economically meaningful items. For example, even though it did not produce MCOs (an advanced type of semiconductor) it was claiming that it could not accept liberalization of those items because it wanted to eventually become a significant producer of them given their commercial importance.Footnote 34 Another factor that made other states even more reluctant to concede to Chinese demands was that there were broader WTO negotiations going on in Bali at the time and so those officials were loathe to do anything that would set a bad precedent.Footnote 35
US–China informal agreement
In March 2014, after traveling to China, US Congressman Charles Boustany (R-LA) said that Beijing appeared willing to make further concessions in the plurilateral talks to expand the Information Technology Agreement (ITA) but first wanted to ensure that the US would be more flexible in its position through bilateral negotiations and said that the Chinese did not ‘like the pressure being exerted collectively by the EU and Japan and us. They feel that a US–China agreement should drive the process.’Footnote 36 When he was asked if he thought China would be willing to adopt a less hardline position, he answered, ‘Yeah, I get the sense that's what they want. They want to directly negotiate with us on this.’Footnote 37
In April 2014, US IT industry CEOs met with Chinese officials and pointedly reminded them that many investments were made in China as a result of the original ITA and that if China did not support an expanded ITA, future investment would go to other Asian locations.Footnote 38 They also said that while they could not accept Chinese demands on product scope, they would be more positively inclined toward the idea of phase-outs.Footnote 39 Less than a month later, the US and China had bilateral discussions on the ITA at an APEC meeting in May 2014.Footnote 40 In July, they held more talks while discussing other bilateral issues ranging from natural gas exports to financial flows; there was enough informal compromise on these matters that it laid the groundwork for a meeting between the two presidents to be scheduled for November.Footnote 41 The two states continued to discuss their differences on IT goods throughout the autumn. By the end of October, they had found compromise on all but about ten tariff lines related to contentious items such as MCOs.Footnote 42
In November, during President Obama's visit to China as part of APEC meetings, the two states announced that they had reached an agreement to expand the ITA. Their list of products that would see tariffs eliminated included a number of US priorities such as MCOs and medical devices (Donnan, Reference Donnan2014a). It also included an array of items from GPS devices to video cameras and in sum reduced tariffs on roughly 1 trillion dollars in traded goods (Donnan, Reference Donnan2014b).
The US and China left the details on the timeline of tariff elimination unresolved for the time being. They had agreed that there would be four levels of phase-out length: immediate, three year, five year, and seven year. However, they had not agreed on which items would be which length.
One major concern was that flat-panel displays, which were a high priority for Japan, South Korea, and Taiwan had not been included in the US–China unofficial deal. Still, the expectation was that given how beneficial the expansion would be, no participant would block the deal over this.Footnote 43 Furthermore, the agreement did not need to be ratified by the US Congress because Under Section 111(b) of the Uruguay Agreements Act in 1994, the president can lower tariffs when the agreement is part of the Uruguay Round and this ITA expansion would function as an amendment to the original ITA and thus fit under that legal framework. At this point, negotiators believed that an officially completed expansion of the ITA was achievable by the end of December 2014.Footnote 44
Flat-panels, Korea, and Taiwan
South Korea and Taiwan were more upset about the exclusion of flat-panel displays than the US and China had estimated they would be. South Korea is home to the largest manufacturers of flat panel displays. Flat-panel displays were also really important to Taiwanese domestic stakeholders. Domestic hard liners and industry groups were very angry that flat-panel displays had not been included. Of the 12 tariff lines on machine tools and the one tariff line on flat panel displays on which Taiwan had wanted tariffs cut from the beginning, none made the final deal and so Taiwan – with some justification – felt that it had been marginalized and ignored.Footnote 45
South Korea and Taiwan both thus refused to go along with an ITA expansion that did not include flat-panels. At the very least, they were hoping for other concessions from China if not on flat-panels. China however refused to make further concessions, even small ones. Chinese officials had been adamant about the exclusion of flat-panels from the beginning and believed the deal with the US should be final (Donnan, Reference Donnan2014c). The flat-panel dispute threatened to totally derail the agreement.
The US, the EU, and others pressured China to make concessions to South Korea and Taiwan to compensate for the lack of flat-panel displays. Korea wanted organic light emitting diodes, rechargeable batteries, and some types of monitors.Footnote 46 China's counter-demands were car stereos, watt-hour meters, and switching equipment, but Costa Rica, El Salvador, Guatemala, and Australia did not want to accept those.Footnote 47 Once no agreement could be found on these issues and with a solution to the larger flat-panel displays contention seemingly off the table, the talks came to a halt once again, despite how close the negotiating countries were to reaching a deal. South Korea and China then began bilateral discussions because nothing else could go forward until they had settled their differences.Footnote 48
Negotiators badly wanted to finish a deal in principle by the middle of the summer so that a finalized agreement could be announced at the WTO ministerial in Nairobi in December 2015. The US, the EU, and Japan were all pressuring South Korea to relax its stance so that a deal, which would still be beneficial to it overall, could pass.Footnote 49 What made South Korea's intransigence seem particularly obstinate was that a 2006 bilateral deal between China and South Korea stipulated that Chinese tariffs on South Korean flat-panels would be eliminated by the end of 2016, so it seemed that the Korean government was holding out as a matter of principle rather than out of commercial benefit. In any event, it was proving difficult to identify goods that were commercially significant enough to satisfy South Korea and Taiwan and also acceptable to China.Footnote 50
Product coverage finalized, on to phase-outs
In April 2015, China called for a new round of talks dealing with both product coverage and phase-out periods.Footnote 51 Basically, China was signaling that it might be willing to accept coverage of items it had previously said no to if it could secure a long phase-out period. Finally, in July 2015, the negotiating parties reached a breakthrough when the US made small concessions to China to help South Korea and the EU make their own deals with China (Donnan, Reference Donnan2015).Footnote 52 The final deal covered 201 items worth 1.3 trillion dollars annually, representing roughly 7% of all traded goods.Footnote 53
This was still not the end of the negotiations though. The parties had settled on which products would see their tariffs eliminated, but they still had to come to a settlement on the timeline of that elimination. The parties agreed on four categories based on the US–China bilateral deal: immediate, three years, five years, and in rare cases seven years. China's hardline position on phase-outs mirrored its earlier position on product coverage; it wanted much longer phase-outs than many other countries were willing to allow.Footnote 54 China wanted five or seven year phase-outs for 90 of the 201 products, including many of the most significant products such as MCOs.Footnote 55 In contrast, the US’ offer was to eliminate tariffs on all items within three years and Japan was eliminating tariffs on all items immediately.Footnote 56 Negotiations over these phase–outs ran all the way up to the December ministerial in Nairobi. China did not even submit a final tariff schedule until the ministerial. Sources said, both during and after the negotiations, that they believed China's delay in submitting its offer was a negotiating tactic aimed at running out the clock so that other ITA participants would be forced to accept its final schedule without demanding further changes on its phase-outs.Footnote 57 In the end, China's final tariff schedule included five or seven year phase-outs on over half of the 201 items, including MCOs and medical devices.Footnote 58 China did at least agree not to backload the tariff reductions. So, for example, if it levied a 35% tariff on a good for which it had agreed a five-year phase-out, it would reduce that tariff to 28% after one year, to 21% after two years, etc. The US would be eliminating 80% of tariff lines immediately and the rest after three years.Footnote 59
The parties also agreed to revisit the terms of the agreement, if the critical mass of producers dropped below 90%.Footnote 60 India was the country left unstated but widely understood to be the one that, if it became a major IT goods producer, could reduce the critical mass below 90% since it was not one of the participants in the ITA expansion.
The final agreement was announced at the Nairobi ministerial in December 2015. Despite the long and difficult negotiations, it nevertheless amounted to the most significant victory for trade liberalization under the WTO since the organization was created in 1995. The tariffs slated to be eliminated immediately came down on 1 July 2016. The three-, five-, and seven-year phase-outs will happen in 2019, 2021, and 2023 respectively.
4. The four factors that led to the ITA expansion and what they suggest about the international political economy literature on trade
Earlier, I discussed the three characteristics of plurilaterals (i.e. no single undertaking, many but not all WTO members, and a focus on tariffs only). All three of these played a major role in facilitating the ITA's passage. An additional factor (an absence of nationalistic opposition) was critical as well.
No single undertaking/narrower scope
Some scholars have argued that the Single Undertaking is a significant impediment to further liberalization.Footnote 61 Others have defended it and argued that it was not responsible for the failure of the Doha Round (Wolfe, Reference Wolfe2015). That the ITA expansion, unlike the Doha Round, did not employ a single undertaking and succeeded lends considerable evidence to the former of those two arguments.
First, the eschewing of a single undertaking approach reduced opportunities for concession extortion and in general kept the negotiations simpler. In a single undertaking, nothing is agreed to until everything is agreed. This means that even if all parties have agreed to a package of trade liberalization measures, in say textiles, that package is not a done deal until everything else is also agreed to, from agriculture to services. What this means is that a party that has no substantive issue with the textile package may uphold agreement on that package until it gets a concession it wants in a totally unrelated area. In 2014, India did this exact kind of concession extortion over an unrelated issue when it refused to accept a Trade Facilitation Agreement under the WTO until it received concessions on support for agricultural production. Because these kinds of critical mass agreements generally deal with just one industry or product, once agreement on that has been found, it cannot be spoiled by demands for concessions on some unrelated matter.
This is exactly what happened in the ITA. While states were negotiating over the product list, no state refused to accept the inclusion of a priority item for another country in order to extract a concession on some unrelated product because there were no unrelated products to deal with. This was precisely why states made a point to only include items that had clear IT relevance. Keeping the negotiation to those products reduced the room for spoilers.
Sticking to only IT related products also prevented each participant's stickiest trade issues from gumming up the works. Practically all states have some sort of trade policy position that abrades almost everyone else. Had those been allowed to creep in, the ITA would have become a hostage to movement on those more viscid issues. The US might have tried to tie IT tariff reductions to more stringent enforcement rules on intellectual property, to the chagrin of everyone else. The EU might have tried to insert labor provisions no one else wanted to include. Developing countries might have demanded new commitments from the EU and Japan regarding agriculture, which they would not have agreed to. That none of these was allowed to happen was a key permissive cause in the advancement of the ITA expansion.
Relatedly, because critical mass agreements like the ITA focus on only one industry or product class, they reduce the number of societal interests that become politically engaged in the negotiations. Single undertaking negotiations, because they deal with a myriad of products and industries, motivate a vast array of domestic political groups to take an interest in those negotiations. By bringing a greater number and variety of societal actors, a single undertaking increases the chance that one or more of these actors will try to act as a veto player and block the deal because it is not to that actor's liking. Critical mass agreements, by avoiding this, have a better chance of being completed and ratified.
This too helped the ITA. The only major industry involved in the ITA was not surprisingly the IT industry and it wanted an ITA expansion. Other business interests in industries that might have been less enthusiastic about trade liberalization never became involved. This provides some evidence that the practitioners’ story, in which trade agreements are the product of engaging with exporters and reducing protectionist demands, is a better explanation of what leads to those successful agreements than the more apolitical economic model-based explanations, such as Bagwell and Staiger and Grossman and Helpman (Regan, Reference Regan2015). Trade deals do not happen if and when they maximize the utility on an economist's blackboard. They happen if and when those who want to see trade liberalized politically overpower those who do not.
Many, But Not All, WTO Members
Second, the plurilateral nature of this kind of agreement, i.e. including many but not all WTO members, means that recalcitrant states that are not major producers in the industry in question can be excluded, thus further marginalizing potential spoilers. Developing country opposition to moving forward on the four Singapore Issues was one of the major reasons behind the eventual collapse of the Doha Round (Gallagher, Reference Gallagher2007: 62–85). In the case of the ITA by contrast, the sectoral nature of the agreement, combined with the fact that India's production in IT goods was not large enough to jeopardize the 90% critical mass, meant that developing countries that did not want to facilitate IT trade liberalization, and most especially India, could be prevented from derailing the negotiations.
Relatedly, informal agreements between key players were the building blocks of the ITA expansion. Unlike in the Doha Round when the US and the EU could not find agreement on agricultural proposals, they were able to reach agreement on what the scope of the ITA expansion should be (tariffs only rather than addressing services and NTBs). Furthermore, whereas the trajectory of the ITA expansion suggests that Wolfe (Reference Wolfe2015) underappreciated the importance of the single undertaking as an impediment, it lends evidence in favor of his argument that China's newfound centrality is crucial to the successful promotion of multilateral liberalization.Footnote 62
The informal deal between China and the US was critical to getting the ITA negotiations to move forward. The results of that deal left some other parties such as South Korea and Taiwan disappointed, but it did overcome the greater hurdles preventing China from getting to a position that others found acceptable. This suggests that in future negotiations, parties may be better off if key participants such as the US, the EU, and China find informal agreement among themselves first and only then consult more formally with a wider set of negotiating partners.
Tariffs only focus
Third, because they generally only seek to reduce tariffs rather than deal with non-tariff barriers, such as subsidies and regulatory differences, critical mass agreements defuse NGO opposition. This is important. On some occasions, such as in the Transatlantic Trade and Investment Partnership, NGOs have been significant drivers of opposition to trade agreements (Young, Reference Young2016). NGOs provided the main opposition to the Anti-Counterfeiting Trade Agreement that caused its downfall (Dur and Mateo, Reference Dur and Mateo2014). Their primary set of concerns about trade agreements is that those agreements will undermine the domestic regulations they value. By not dealing with regulatory trade barriers at all, plurilaterals do not attract the opposition of NGOs. Relatedly, by not affecting regulations and by being generally lower in salience, these plurilaterals provided NGOs with fewer opportunities to generate attention and resources for themselves via opposition to a trade agreement.
This avoidance was not a foregone conclusion. Had the ITA expansion delved into IT-relevant regulations on a sensitive issue such as data privacy, it is easy to imagine NGOs and activists decrying this as a giveaway to big business and a sacrificing of consumers’ rights. This is not the only way that including non-tariff barriers in the negotiations could have proved fatal. Even without NGO opposition, states simply may have been unable to reach agreement on these more complicated issues. The lack of NGO engagement on the ITA expansion also meant that there was no counterweight to business interests’ lobbying. As the case showed, at key moments, the business community pushed governments to work harder to find a compromise solution. In these moments, there was never an NGO voice on government officials’ other shoulder giving contrary advice.
Counterfactuals are obviously impossible to prove, but let us assume for a moment that the ITA expansion negotiations had involved the kinds of regulatory cooperation that the EU had envisioned in the early stages. Recall that the EU wanted IT goods suppliers to be allowed to self-certify that their goods met different countries’ regulatory standards. Had this been a drawn-out topic of negotiation, it is possible, and in fact quite probable, that some states, most likely Japan, would have been opposed to firms being allowed to do this. One can also easily imagine developed states being skeptical that firms from countries with a history of lax regulatory stringency (read developing countries) would actually meet developed country's standards in the way they said they would.
Critical mass agreements’ relatively narrow focus on tariffs also means that their benefits are quite tangible. It is difficult to explain to the average voters how reducing regulatory trade barriers related to data exclusivity in pharmaceuticals or to localization requirements regarding e-commerce affects them. By contrast, it is relatively straight forward to make the case that reducing tariffs on a given product from 25% to 0% will make that product more affordable for consumers.
This approach has shown promise before, as it did in the zero-for-zero tariff reductions in the Tokyo and Uruguay Rounds. In the case of the ITA, had some kind of societal opposition to the ITA emerged, it would have been relatively simple to explain to voters that the ITA expansion would make their smartphones more affordable, and everyone wants their next smartphone to cost them less money.
Admittedly, avoiding regulatory issues where possible does amount to a scaling back of the ambition of trade liberalization. Especially for developed countries, the major impediments to trade are not tariffs but non-tariff barriers, and particularly regulatory issues. Regulatory differences are some of the most significant impediments to trade and are only growing in importance (Winslett, Reference Winslett2016).
Still, limited gains are better than nothing. The importance of dealing with regulatory differences as trade barriers does not mean that more easily ratifiable treaties should be held back because of the difficulties surrounding negotiations over those regulatory differences. Furthermore, there is a lesson here even for those trade negotiations that must deal with non-tariff barriers: avoid contentious issues if at all possible. If investor–state dispute settlement provisions or some other disciplines in a proposed trade agreement are going to generate significant opposition to a trade deal, then perhaps it is better to have a trade deal without those provisions than no trade deal at all.
Finally, it bears noting that though this critical mass approach seems especially promising for tariff negotiations, it may be helpful in other contexts too that deal with trade impediments other than tariffs. Indeed, a critical mass approach was used in both the Financial Services Agreement and the Basic Telecommunication Services Agreement negotiated in the 1990s.
Lack of nationalistic opposition
Other major trade deals have faced withering criticism from right-wing populists that they amounted to concessions that benefitted foreigners while hurting the nation. Donald Trump and his supporters vehemently denounced NAFTA and the Trans-Pacific Partnership. Nigel Farage and the Leave campaign railed against the EU. Marine Le Pen criticized TTIP. All of them launched nationalistic broadsides against trade deals, arguing that trade openness created job losses and weakened their country while helping foreigners. The rise of the populist right coupled with the left's longstanding skepticism of trade means that there are now more numerous and more varied political actors willing to veto trade agreements and as Mansfield and Milner (Reference Mansfield and Milner2012) have shown, the presence of these veto players is a key predictor of democratic leaders’ ability to sign on to trade deals.
The ITA never generated a similar broad-based nationalistic opposition and thus mitigated the extent to which it created these potential vetoes. At no point was there a groundswell of critics of the deal claiming that it would lead to job losses or a relinquishment of sovereignty. I would argue that this was because the production of IT was already organized internationally with complex supply chains. What this means is that many producers of IT goods are themselves consumers of imported basic parts and raw materials. Additionally, it meant that there were fewer domestic manufacturers that both depended on trade protectionism for their survival and would be able to generate salience and visibility for potential job losses.
The agreement got wrapped up as Donald Trump was taking over the Republican Party and loudly challenging GOP orthodoxy on trade and yet there was hardly a peep from him on the ITA expansion, TTIP opposition in Germany had grown substantially by then and yet there was no similar opposition to an ITA expansion in Germany. Had there been a nationalist backlash to an ITA expansion from any of the major negotiation participants, that might have forced that state to withdraw, which likely would have scuttled the whole deal.
This kind of a nationalist backlash did not occur for three main reasons. First, the salience of the ITA negotiations remained very low. The only people who seemed to know about it and care about it were people in IT-intensive industries and trade policy works. This is evidenced by the fact that a LexisNexis search shows that in 2014 and 2015, in The New York Times, The Guardian, The Japan Times, The Sydney Morning Herald, Der Spiegel, and The Washington Post, there were just five articles across all six sources on the ITA expansion negotiations. Over that same period, these six publications ran a combined 481 articles on the Trans-Pacific Partnership and 356 articles on the Transatlantic Trade and Investment Partnership.Footnote 63
Second, the decision to avoid dealing with non-tariff barriers, and especially with domestic regulations, helped keep the ITA from becoming controversial. Citizens generally believe that their country's regulations are superior to other countries’ and are often opposed to the idea of those regulatory differences being diminished in order to further a trade agreement (Pew Research Center, 2016: 19–20).
Third, ITA negotiators refused to get stuck on relative gains concerns. That is, negotiators’ main focus was on constructing a deal in which all parties benefitted enough to support the deal rather than each of them focusing on benefitting more than other parties. Relative gains concerns did appear at times. For example, when the negotiations did bog down, it was over the concern that China was making too few concessions in return for the benefits it was receiving. Still, despite these temporary setbacks, negotiators refused to allow those kinds of concerns to kill the negotiations. The focus on everyone benefitting kept the negotiations moving forward and ultimately produced an agreement that all of the major parties were satisfied with.
Taken collectively, these four factors suggest that, at the very least, the Uruguay Round was a product of a very specific set of circumstances, and that using the Uruguay Round as the model for moving forward is a mistake. The Doha Round collapse has, if nothing else, made that abundantly clear. Some scholars have talked about how a new round might be launched and what negotiators can try to do to ensure that a round would ultimately succeed rather than fall apart, as Doha did (Odell, Reference Odell2015). The lessons from the ITA suggest that launching a new Doha-like round is probably not advisable; liberalization can best be advanced by other plurilateral critical mass means.
5. Conclusion: prospects for other ongoing negotiations
The Environmental Goods Agreement (EGA), an effort to promote trade in environmentally beneficial goods, is currently still being negotiated. EGA negotiations began in 2014 and largely resemble the ITA expansion talks. Like the ITA expansion, the EGA is sectoral in nature, i.e. focused on reducing tariffs only on environmental goods such as solar panels and wind turbines. It is also an open plurilateral with a critical mass target of 90% and focused solely on tariffs rather than non-tariff barriers, just like the ITA expansion. Also, just as APEC played an important role in facilitating progress on ITA expansion negotiations, APEC is also spurring progress on an EGA; at Vladivostok, APEC members agreed to lower tariffs on environmental goods, which could make it easier for those states to accept further liberalization in these goods at a future date. Another similarity between the two is that some of the key issues involved whether certain goods were truly relevant to the sector being discussed, what goods would make it on the list, and how long phase-outs would be. Given its similarity to the ITA, the EGA provides an additional test of the critical mass model's relevancy.
The EGA negotiating parties included the US, the EU, China, and 14 other mostly developed states. They seemed to be making progress throughout 2014 and 2015. By November 2015, most of the participants had found broad agreement on a list of approximately 350–375 items.Footnote 64 As in the ITA negotiations, China preferred a much narrower list of goods on which to have tariffs eliminated, but negotiators believed that they could reach a deal by the end of 2016.Footnote 65 At the G20 meeting in China in September 2016, the negotiating partners said they had arrived at a ‘landing zone’, suggesting that they had reached an agreement in principle and implying that they would be able to announce a completed deal at a ministerial meeting scheduled for 3–4 December 2016.
Then Donald Trump won the US presidential election. After the election, the Chinese negotiators became noticeably less engaged and less ambitious in their offer, and many of the other parties were unsure as to what this signaled about their intentions.Footnote 66 They soon found out. At 11:00am on the final day of the December ministerial, just hours before the press conference announcing a completed deal was scheduled to be held, the Chinese delegation proposed a list that was nowhere close to what other countries wanted in a final deal and was actually less ambitious than the APEC list that it had agreed to years earlier (Keaten, Reference Keaten2016).Footnote 67 Between the uncompromising nature of the list and its timing, it seemed clear to many that the Chinese negotiators believed that the Obama Administration would be so eager to complete a deal before their time in office ended that they would accept China's new more hardline demands.Footnote 68 They miscalculated. The other parties walked away from the negotiations instead, with the US, the EU, and global industry blaming China for the failed negotiations (Miles, Reference Miles2016).Footnote 69
Where the EGA negotiations go from here is anyone's guess. China may return to the table with a product list that is more acceptable to the other parties. It may use the Trump Administration's rhetoric on Taiwan and other issues as a reason to not re-engage in talks at all. Environmental NGOs may or may not fully approve of the final product list.Footnote 70 China and the EU have had a long running disagreement over bicycles while the US and China have disagreed strongly on high-tech batteries; even with the best of intentions the parties may not be able to settle their differences over these items.Footnote 71 In terms of the US position, Trump has railed against trade but never criticized the EGA specifically, and at least one notable Republican, House Ways and Means Chairman Kevin Brady (R-Texas), has urged the Trump Administration to complete the EGA (Needham, Reference Needham2016). Trump could use the completion of an EGA to trumpet how much better the deals that he makes are and how his toughness led to the Chinese making concessions to their December offer. On the other hand, it is also possible, and I mean this in all seriousness, that after months of careful negotiating between the various parties, President Trump scuttles the whole thing with a dyspeptic 4:00am tweet.
Regardless of the broader international political context, the ITA expansion suggests that keeping the EGA away from non-tariff barriers and especially regulatory differences, having it be a critical mass agreement, employing phase-outs, and minimizing the extent to which nationalistic concerns about relative gains become a major sticking point, will all be crucial to liberalizing trade in environmental goods.
The politics surrounding trade agreements like the TPP and TTIP have grown even more contentious over the last few years. Right-wing populists have railed against trade as an elite-driven project that hurts their nation's citizens. Leftists have maintained their opposition. Even agreements such as the Comprehensive Economic and Trade Agreement (CETA) between Canada and the EU – that might have once been considered unobjectionable – now face stiff resistance.
That does not mean however that all trade liberalization efforts must be forced into hibernation through this winter of trade discontent. On the contrary, Doha may be dead but tariffs still exist and, where they do, they present a somewhat less challenging target than other trade barriers. A transatlantic trade agreement on wine for example may not be as ambitious as TTIP was, but it is still a step in the direction of liberalization.Footnote 72 Painstaking, sector-by-sector negotiations may lack the rhetorical flashiness or obvious ambition of a new WTO round, but if completed, those negotiations would nevertheless be important steps toward liberalizing trade. The ITA expansion offers a blueprint for how trade liberalization in these and other issues can still be achieved. Eschewing a single undertaking, limiting the set of negotiating countries, bypassing contentious issues where possible, employing differential phase-outs, and avoiding a nationalistic backlash, can all be hugely helpful.
