Information gaps and the demand for information

In order to understand variations in the designs of information systems, we need to understand the demand for information by outlining the information gaps that countries seek to bridge. What does imperfect information really mean? Information is a necessary input in all phases of regime design and sustenance: agenda-setting, negotiation, implementation, monitoring, and enforcement.Footnote ¹⁴ Information gaps hamper attempts to enhance cooperation among states during all of these phases. Not all information gaps, however, are of the same variety. Institutional IR theory recognizes that information asymmetries imply that one actor has more information than another. But it does not explain how these asymmetries affect different stages of a regime or different categories of states. Moreover, information gaps afflict states not just in a relational sense (as asymmetries) but in an absolute sense as well (absence of salient and credible information for all parties). Viewed this way, we can conceptualize information gaps, which would incorporate both the relational as well as the absolute aspects of information barriers to international cooperation.

Resource-constrained information

Information gaps also arise because of a lack of resources. States have to invest considerable time and money to gather information during the negotiation phase as well as to formulate implementation strategies and monitor other states. In regimes with large membership, effective monitoring can become a prohibitive exercise. Resource-constrained information gaps are particularly problematic for developing countries, which suffer from a lack of financial and technical resources to collect and analyse the necessary information.

Of course, not all non-compliance need be deliberate. There are several reasons why states might find themselves in non-compliance of a treaty without wilfully wanting to be so: ambiguity in the actual agreement, incapacity of states to fulfil a commitment, and the temporal dimension to account for the period between treaty conclusion and implementation.Footnote ²⁰ Axelrod and Keohane argue that international regimes reduce this uncertainty and help to avoid mistaken disputes.Footnote ²¹

Table 1 summarizes the information gaps and their implications for regimes and developing countries. Issue uncertainty during negotiations is doubly problematic for developing countries since they are uncertain about their own preferences as well. This could be partly resolved if countries invest in additional research, but the lack of financial and technical resources can hinder their efforts. Hidden/inaccurate information is problematic if developing countries remain unaware of non-compliance by other states. In the trade regime, protectionist policies abroad affect a country's export interests. But, while in rich countries, export-oriented groups can be expected to monitor actions elsewhere, such capacities are limited for developing countries. Even governments are constrained in their ability to monitor policies due to a lack of resources. For developing countries, it is ultimately the resource-constrained information gap that is the most critical. It adversely affects the potential for developing countries to participate in all phases of a regime.

Table 1.

Information gaps and their impact on regimes and states

Alternative designs of information systems

The design of information systems depends on a range of factors. The framework presented here extends existing literature by accounting for more types of information systems and for variations in them over time. Information systems are designed based on the nature of information gaps affecting regimes and their member states, the power imbalances between states and their incentives towards the regime, and the capacities of state and non-state actors to monitor developments. The interaction of these variables – information gaps, actors' interests, actors' capacities – yields five design types.

• • No reporting – Although all regimes require information, it is conceivable for some to have no explicit reporting or information gathering mechanism. We should expect this outcome when regime membership is small and behaviour is clearly observable. We should also expect no reporting where members are similar in size and capacity, resulting in a degree of trust among the states.

• • Self-reporting – States report on their own activities, policy direction, and legislative changes to show compliance or justify non-compliance with regime obligations. We should expect this outcome when the number of actors increases, the threat of hidden/inaccurate information increases, but where states are unwilling to cede sovereignty to the regime secretariat. Self-reporting would also be more acceptable where inaccurate information is the predominant threat, but behaviour is observable with relative ease.

• • Other-reporting – States gather and report information about the actions of others, whether states or non-state actors. We should expect this outcome when states face hidden/inaccurate information hurdles, but there is reasonable equivalence in the capacities of members to monitor others.

• • Institutional reporting – The regime secretariat collects information on members' actions and the functioning of the regime as a whole. We should expect this outcome when regime membership is large, the wide scope of issues adds to more uncertainties, and a high proportion of states are resource-constrained.

• • Non-state actors (NSA) reporting – Non-state actors collect and analyse information, to report on member states' behaviour and on the impact of the regime on an NSA's interests in a particular issue area. We should expect this outcome when the incentives of states and NSAs do not coincide, when states are resource-constrained, or when states suffer from issue uncertainty in new areas of negotiations.

In practice, information systems in international regimes are often a combination of these designs. The next section explains how information gaps and members' interests and capacities influenced the trade regime's information system, in particular the TPRM.

1950s: ad hoc reviews

During the GATT's initial years (1947–1955) its membership was small (only 33) and developing countries accounted for less than half the membership (48%). Information gaps would have been minimal in the mostly developed country-dominated trade regime. As a result, there was no institutionalized monitoring mechanism in these early years (a case of ‘no reporting’).

The earliest example of formal surveillance emerged in the mid-1950s. The GATT's 9th Session (28 October 1954–18 March 1955) was unique in its history in that it served as a Review Session as well. The previous session had recognized that ‘international trade continues to be restricted by high tariff barriers…’ There was also a perception that trade was being hampered ‘by widespread application of other restrictions [including quantitative restrictions]’.Footnote ²² Contracting parties recognized the need to garner information about whether these measures were being legitimately used or not.

The Review Session looked into quantitative restrictions, schedules and customs administration, other barriers, and organizational questions. Following the submission of working party reports, regular consultations were established to review quantitative restrictions maintained for balance-of-payments purposes.Footnote ²³ Thus, limited surveillance was established to counter a potential threat of hidden information on non-tariff barriers. But contracting parties decided not to extend additional authority to the GATT Secretariat to conduct reviews on other trade issues.

Another example of ad hoc institutional reporting occurred in 1958. The small number of developing countries in the GATT continued to assert that their interests were not being taken into account. Nor did they have the capacity or resources to undertake a detailed review of rich countries' trade policies on their own. Largely in response to such complaints, contracting parties established a Committee under the chairmanship of Gottfried Haberler, a Harvard-based trade economist. The Haberler Report, noting the decline in the terms of trade for primary commodity producers, argued that the trade policies of developed countries were hindering less developed ones from participating in the trade regime. In response, a Committee dealing with agricultural trade conducted consultations with 40 contracting parties during 1959–61 on agricultural policies.Footnote ²⁴ With the Common Agricultural Policy emerging in 1962, the Committee also held consultations with the European Economic Community.Footnote ²⁵ These were, however, rare examples of systematic reviews of whether the GATT system was achieving its objectives for a growing proportion of its membership. Thus, although a case of institutional reporting, the incentives of contracting parties remained issue-specific and the regime was not yet large enough for members to mandate a more permanent information system.

1960s–1970s: mixed information systems

The situation began to change as the GATT grew in size, the proportion of developing countries increased, and issues that are more complex started dominating the trade agenda. By the end of the 1960s, GATT membership had more than doubled to 75. The issues under negotiation also broadened in scope, such as those concerning balance-of-payments restrictions, agriculture, various multilateral trade negotiations (MTN) codes, and textiles. The risk of hidden information increased in a heavily laden trade agenda, and contracting parties developed an incentive to monitor issues of specific interest to them.

Several ad hoc and sector-specific surveillance mechanisms were established in the 1960s and 1970s. The Committee on Balance-of-Payments Restrictions consulted annually with relevant developed countries, and biennially with developing countries. The Textile Surveillance Body, set up under the Multi Fibre Agreement (MFA), had a ‘small but geographically representative membership, meeting as often as necessary’.Footnote ²⁶ It received annual notifications on textile restrictions from all MFA signatories, and reviewed bilateral and unilateral actions by the signatories. The Textiles Committee also established a Sub-Committee on Adjustment in 1981 to review measures adopted to facilitate adjustment. The various MTN codes were also subject to notification requirements for contracting parties that had signed the codes. Some countries independently conducted surveillance over the implementation of the codes as well. However, all documentation was restricted only to the code signatories and observers.

The 1960s and 1970s were a mix of self-reporting (textiles, MTN codes, BoP restrictions), other reporting (the European Community and the United States conducted their own surveillance targeted at major trading partners), but minimal institutional reporting. Contracting parties recognized the need for surveillance on specific issues. But, given that signatories to specific agreements were fewer than the larger GATT membership, reporting behaviour was more easily observable. Thus, countries were not yet willing to give a strong mandate to the GATT Secretariat.

Early 1980s: issue-specific institutional reporting

It was not until the 1980s that a series of proposals for strengthened institutional surveillance in the GATT came forward. In 1979, the Tokyo Round had concluded with an Understanding Regarding Notification, Consultation, Dispute Settlement, and Surveillance, which provided for special GATT Council meetings to review developments in the trading system.Footnote ²⁷ Furthermore, although tariffs had been cut significantly, there was a growing need to monitor the use of non-tariff barriers. From 1980 onwards, Special Council meetings were held biannually with the Secretariat submitting a note detailing developments in the preceding six months. In 1983, the scope of the meetings was broadened to ensure that trade policies were consistent with GATT principles and to ‘avoid measures which would limit or distort international trade’.Footnote ²⁸ This was largely due to issue uncertainties in the negotiations, growing concern about hidden information on non-tariff barriers, and a poor record on compliance with notifications.

Surveillance on issues of concern for developing countries also increased to some extent via the Committee on Trade and Development (CTD). The CTD had emerged as a result of the inclusion of Part IV in the GATT in 1964, which focused on the special circumstances and needs of less developed countries.Footnote ²⁹ In 1979, when poor countries comprised 70% of the GATT membership, the Sub-Committee on Protective Measures was established to monitor actions by rich countries that could block imports from developing ones. Also, the Sub-Committee on Trade and Least Developed Countries was made responsible for reviewing trade policy developments of interest to this category of contracting parties. After a GATT Ministerial Declaration in 1982, the CTD adopted a programme of consultations ‘to examine how individual contracting parties have responded to the requirements of Part IV’.Footnote ³⁰ But no conclusions were reached until the start of the Uruguay Round.

Developing country concerns did not translate into wider institutional review. Although developing countries were increasing in number, their engagement with the trading system was very limited in the 1960s and 1970s, focusing on securing preferential access to markets. Thus, the degree to which issue uncertainty would have hindered their negotiating capacities was limited. Instead, they were largely affected by hidden information about rich countries' policies and were constrained by a lack of resources to undertake reviews themselves; hence, the call for institutional reporting within the CTD. At the same time, developing countries had been traditionally very protective of their sovereignty; they were reluctant to let the GATT Secretariat monitor trade policies more regularly.Footnote ³¹ So, a combination of limited issue-specific interests and resource-constrained information gaps resulted in limited surveillance oriented towards developing country interests.

Late 1980s: debating and designing the TPRM

In 1985, an Eminent Persons Group (EPG, chaired by Dr Fritz Leutwiler), arguing that ‘governments should be required regularly to explain and defend their overall trade policies’, so as to ‘prevent departures from the [GATT] rules’, proposed that ‘countries should be subject to regular oversight or surveillance of their policies and actions …’Footnote ³² The GATT Secretariat was to collect and publish this information and the Secretariat would initiate studies of trade policy changes. But the group made it clear that the Secretariat would act ‘as watchdog (though not judge)’ on behalf of the trading system.Footnote ³³ A watchdog, by definition, acts as the guardian of others' rights. So, an appropriate surveillance mechanism for the EPG would not only review members' trade policies but would also raise the alarm when non-compliance affected other members' rights in the system. For developing countries, the value of a multilateral watchdog would become evident only if they perceived it to be defending their trade interests.

Until the mid-1980s, there was no mechanism that involved the periodic review of all contracting parties as well as of the trade system as a whole. There was a growing perception among developing countries that an ‘unwarranted asymmetry’ existed in surveillance obligations: some countries had to get their trade policies reviewed under balance-of-payments consultations and as conditions of accession to the GATT, but there was no surveillance of the policies of major trading powers.Footnote ³⁴ Moreover, there was the question of neutrality and credibility of information. A chief concern among contracting parties was that the only available review of global trade policies was one published by the United States Trade Representative (expectedly, these reports steered clear of US trade barriers).Footnote ³⁵

The Punta del Este Ministerial Declaration that launched the Uruguay Round committed ‘to enhance … surveillance in the GATT’ with regular monitoring of trade policies and their impact on the trade system. On 28 January 1987, a Surveillance Body was established to monitor the standstill and rollback provisions in the Declaration. By relying on notifications (self-reporting), the Surveillance Body would submit reports to the Trade Negotiations Committee. Thus, surveillance in this case was a way of increasing confidence in the GATT system while a major trade round was being negotiated.Footnote ³⁶

During the Uruguay Round, a negotiating group on the Functioning of the GATT System (FOGS) was established. In the first proposal submitted to the group, Australia criticized the existing system of surveillance through biannual reviews, pointing to ‘arbitrary data collection, and lack of quantification of support measures and their trade effects’.Footnote ³⁷ The United States wanted the Secretariat to prepare country reviews, so that the GATT would engage in ‘enhanced surveillance’ of countries' trade regimes. It also wanted greater economic analysis and research and quantification of the effects of trade measures.Footnote ³⁸ The European Community wanted to rationalize existing GATT notification and surveillance mechanisms. It demanded that the country reports be policy-oriented, highlighting the objectives, challenges, and direction of a country's trade policies.Footnote ³⁹

Among the similarities in developed country positions was recognition that surveillance activities should serve not only as an ‘early warning system’ but should also be ‘used to ensure compliance with contractual obligations …’. Multilateral surveillance would provide a ‘counter-weight to pressures from domestic special interest lobbies’.Footnote ⁴⁰

In principle, developing countries also favoured a multilateral surveillance system. Strengthening the existing surveillance procedures meant regular monitoring of the trade policies of major trading powers, ‘so as to ensure symmetry in surveillance in the GATT’.Footnote ⁴¹ At the same time, developing countries did not want any new obligations. Instead, they wanted a new surveillance mechanism to target richer countries.

Jamaica (a member of the more moderate coalition, the Group of Twenty) was keen to ensure that the TPRM ‘does not explicitly require any new obligations …’.Footnote ⁴² Arguing that the credibility of the GATT system was undermined by the actions of ‘those having major shares and stakes in world production and trade’, Jamaica contended that any effort to improve the GATT system had to start with ‘trade policy … at the national level’.Footnote ⁴³ It insisted on rationalizing the several GATT bodies that carried out surveillance activities; improving the monitoring of MTN codes, the functioning of the Textiles Surveillance Body, the surveillance of a strengthened Article XIX; and reviewing the liberalization of trade in agriculture.Footnote ⁴⁴

Several developing countries echoed similar sentiments during negotiations. They argued that the GATT system needed new mechanisms less, and improved adherence to existing principles more.Footnote ⁴⁵ The Leutweiler Group had noted that GATT surveillance was meant to ‘strengthen the ability of all countries – and especially the smaller and developing countries – to defend their trade interests’.Footnote ⁴⁶ India, submitting a note on behalf of developing countries, pointed out that the reviews were not intended to lead to any binding conclusions or recommendations, or to establish a legal process.Footnote ⁴⁷ But, it also hoped that the process would lead to improved adherence with GATT rules. Developing countries warned that GATT surveillance should not become another means to pressure them to change their trade policies.Footnote ⁴⁸ Negotiations on the GATT system had to address the issue of special and differential treatment.Footnote ⁴⁹

At the Mid-term Review of the Uruguay Round in Montreal in December 1988, contracting parties agreed to a proposal to set up a Trade Policy Review Mechanism; the decision was confirmed on 12 April 1989. At a time when developing countries were opposed to several changes in the trade regime, why did they agree to the TPRM? The debates reveal that parties did not disagree on the need for surveillance, but had differing interests in establishing such a mechanism. If rich countries wanted to inject a degree of formality to institutional monitoring, poor ones had their own expectations.

Among the incentives, as already pointed out, was a desire to fix the asymmetry in surveillance obligations in the regime. Here was a chance for poorer countries to influence the functioning of the GATT system (the Uruguay Round provide ‘a useful opportunity for a review of the institutional framework of the GATT’).Footnote ⁵⁰ As one former GATT/WTO official put it, ‘Developing countries wanted to use the resulting TPRs to put developed countries on the block.’Footnote ⁵¹ According to this view, given that developing countries had little recourse to the dispute mechanisms in the GATT, they decided that moral suasion would be a better option.

Secondly, periodic surveillance could serve as a ‘multilateral stamp of approval’ for their policies.Footnote ⁵² At a time when many developing countries were under pressure from international financial institutions to reform their economies, a TPR would have been a legitimate forum to defend one's trade policies. Countries also expected the reviews to be an opportunity to attract investment. In other words, a positive review would be a ‘signal’ to the markets that the country had a conducive environment for investment.Footnote ⁵³ Joseph Francois argues that TPRs increase the credibility of domestic policy reform and strengthen trade-promoting policies.Footnote ⁵⁴ By reducing uncertainty about conditions in export markets, it should also increase investments in tradable sectors. The reduction in risk, in turn, would have a salutary effect in reducing the cost of capital for poor countries.

Thirdly, there was a potential benefit of improved policymaking within countries. A senior WTO official who had been involved in the TPRM in its early days argued that the reviews ensured coordination between government agencies that would not otherwise have communicated with each other.Footnote ⁵⁵ The TPRM was expected to have a demonstration effect by highlighting the beneficial impact of trade liberalization. Donald Keesing writes that the TPRM ‘imposes a useful discipline’ by forcing members to re-evaluate their own trade policies.Footnote ⁵⁶ The information generated in the reviews, by being available to NGOs, businesses and academics, would add to the pressure for change in policies. Several officials claimed that one of the chief attractions of the TPRM was its value in stimulating policy reform.Footnote ⁵⁷ The reviews were an opportunity ‘to ask GATT/WTO staff to do what domestic reformers [could not] do politically’.Footnote ⁵⁸

Developing countries were also aware of the limits to their capacity and resources to gather and process information. As Keesing notes, by making information about foreign trade practices a public good, the TPRM has the greatest marginal value to small countries that do not have the resources to collect information by themselves.Footnote ⁵⁹ For the first time in the trade regime, the TPRM promised to become a source of ‘neutral and unbiased information’ and reduce hidden information by making barriers in export markets more transparent.Footnote ⁶⁰ The TPRM could also reduce issue uncertainties that poor countries encountered during the Uruguay Round. Developing countries were unsure about new trade issues, including services and intellectual property. They hoped that a neutral information system would assist them in understanding the issues and in demanding changes to rules that adversely affected their welfare.

Table 2 illustrates how surveillance has evolved in the trade regime over the past half a century. Unlike existing theoretical literature, the analytical account presented here can explain variations over time within a single regime. Developing countries agreed to the TPRM for reasons that had to do with their incentives in the trade regime, monitoring capacities, and the information gaps they faced. At the Marrakesh Ministerial in 1994, which heralded the close of the Uruguay Round, the Jamaican Trade Minister said that his country strongly supported a monitoring mechanism in the WTO because ‘there [was] a difference of opinion as to the benefits to be derived …’.Footnote ⁶¹ Monitoring would evaluate the negative effects from multilateral trade liberalization so that the losing parties could seek appropriate redress. Thus, for developing countries institutionalized monitoring could reveal hidden information and reduce issue uncertainties for future negotiations.Footnote ⁶²

Table 2.

Evolution of surveillance mechanisms in the multilateral trade regime

Functions of an ideal-type information system

Relevant and timely information yields many benefits. It creates a level playing field for rich and poor members of a regime, while giving an opportunity for members to justify their policies. Further, information generated through periodic reviews could influence policymaking and promote compliance within countries.Footnote ⁶³ Moreover, monitoring not only facilitates compliance, it can also provide in-depth analysis, alert regime members to potential risks, and evaluate the effectiveness of the regime itself, thus helping to develop new norms.Footnote ⁶⁴ Thus, three functions for information systems can be identified:Footnote ⁶⁵

• • Information dissemination – International regimes review national policies by comparing them with international standards and the commitments that states have made. The task of collecting and disseminating information is in response to hidden information problems as well as resource constraints. Information collection and dissemination by neutral organizations is necessary to reduce the risk of systematic information biases.

• • Compliance promotion – Where deviations from commitments are found, surveillance systems can recommend corrective action. This is the task of promoting compliance, whether by means of peer pressure among member states, the threat of use of explicit sanctions following litigation, or by pressure from NSAs whose interests are directly affected by non-compliance. The use of independently collected information for the purposes of improving compliance is also a response to the risk of deliberate deception by member states.

• • Regime evaluation – Regime evaluation occurs in response to issue uncertainty as well as resource constraints. When countries are unsure about the preferences of other parties they would be less committed to implementing the rules themselves. By evaluating the impact of the regime (or its underlying rules) on different states, monitoring informs negotiators more clearly about others' preferences. The same exercise can reduce general uncertainty about the impact of the regime on global welfare and thereby shape preferences for the development of new norms and rules.Footnote ⁶⁶

High expectations but restricted mandate for the TPRM

So, what did GATT contracting parties expect the TPRM to do? The TPRM's aim was to promote ‘improved adherence’ to rules and commitments through ‘regular collective appreciation and evaluation’ of members' policies and practices.Footnote ⁶⁷ But it was not clear how exactly contracting parties expected to use the information generated in the TPRs.

Most countries agreed to make TPRs available to the public. Although Japan had initially suggested that only press releases would be issued,Footnote ⁶⁸ the function of ‘information dissemination’ was largely uncontested. Delegations expected the reviews to be used to exert pressure to encourage internal debates on policy questions.Footnote ⁶⁹ However, members also had the autonomy to decide how widely the reviews would be circulated.

There was also the hope that the information collected during the reviews would be useful in future negotiations. Delegations suggested that the reports had to be relevant and useful, so that parties could ‘keep track of how political pressures on trade policies were developing’.Footnote ⁷⁰ After the WTO started functioning, member states suggested that they should be free to use the TPRs as reference material in WTO committees.Footnote ⁷¹ This was the first time that member states had explicitly recognized the use of TPRs in negotiations, rather than think of them as merely information-gathering exercises. The TPRM was potentially a response to the resource-constrained information gaps that developing countries encountered in negotiations and post-negotiation phases.

Some developed countries highlighted the importance of using reviews to encourage changes in non-compliant policies. For instance, Canada argued that while surveillance was the ‘fundamental question’ for the Negotiating Group, there was also the need to consider what the follow-up to the surveillance would be and its relationship with the dispute settlement system.Footnote ⁷² Canada had pursued improved transparency and trade policy coordination at the domestic policy level since the late 1970s, so its position was consistent in GATT negotiations as well.

Yet, no proposal clarified how the expected policy changes would occur in practice. Information derived from the TPRs could not be used in dispute settlement proceedings or to enforce specific GATT obligations.Footnote ⁷³ If adherence to GATT rules had to improve, then countries could only rely on ‘moral suasion’ or ‘peer pressure’. But how would peer pressure work when the power asymmetries between contracting parties were so vast? The ‘power of large numbers’, drawn from developing country coalitions, would not work because ultimately all decisions are based on consensus.Footnote ⁷⁴ Without majority voting, there was little scope for poorer countries to exert collective pressure on richer ones to change their policies. Consequently, developing countries resisted any explicit provision to change policies, fearing that it would be asymmetrically applied against them. Thus, ‘compliance promotion’ remained an implicit aim for the TPRM.

As regards ‘regime evaluation’, the Nordic countries and the European Community wanted the Secretariat to make periodic assessments of trends in trade policies based on TPRs.Footnote ⁷⁵ India stated that since Special Sessions of the GATT Council had served as ‘early warning mechanisms’ in the past, enhanced surveillance should encompass an overview of general developments affecting the trading system.Footnote ⁷⁶ In fact, according to the Indian delegation, the concept of ‘enhanced surveillance’ had a stronger connotation of moral suasion for developing countries than did ‘trade policy review’.Footnote ⁷⁷

But there was no consensus on the yardsticks that would measure the welfare impact of the regime. The limited demands for regime evaluation translated into an institutionalized ‘overview of developments in the trading environment which are having an impact on the multilateral trade system’.Footnote ⁷⁸ The procedure involved annual meetings of the GATT Council, assisted by a report from the Director-General. This form of enhanced surveillance, it was believed, would ‘strengthen the existing “early warning” aspect’ of Council meetings.Footnote ⁷⁹ Annual overviews would be the closest the trade regime would come to having an explicit regime evaluative mandate.

The mandate of the TPRM limits the ability of the mechanism to perform the three surveillance functions. Reporting requirements could increase transparency and force countries to explain their policies to others. But the impact of reviews depends on how widely information is disseminated, a decision that was left to individual countries. For domestic policy, compliance promotion depends on the extent of consultations between state and non-state actors. There are no structured follow-up procedures to promote compliance abroad; it depends on whether members choose to apply pressure in negotiations or via disputes.

Finally, it was possible for the TPR to have what Asif Qureshi called a ‘prescriptive trait’ or an evaluative role.Footnote ⁸⁰ As WTO Director General Pascal Lamy noted in a recent speech, the surveillance function requires the WTO to offer a forum to discuss ‘how best to operationalize … market opening’ as well as ‘the need for measures that translate the results of negotiations into real benefits for all Members’.Footnote ⁸¹ But there was no decision on which issues/rules to evaluate and for what purposes.

It was evident from the beginning that, without structural changes in the trade regime's governance, any new information system would only have limited impact. Julius Katz, the Chairman of the negotiating group and a prime mover of the TPRM, wanted countries to exert ‘moral suasion’ over each others' trading practices.Footnote ⁸² But developing countries feared that too broad an interpretation of surveillance spelled more chances for abuse of the process. Economic analysis of trade policies could easily slide into the realm of economic prescription of ‘good’ policies, which would then be imposed on them. They did not expect the reviews to give them greater flexibility or influence within the trade system.

As the first mechanism by which developed countries would offer their policies for scrutiny and criticism by developing countries, the TPRM represented the first attempt to ‘institutionalize peer pressure’ in the multilateral trade regime.Footnote ⁸³ But the TPRM was not a standalone mechanism; its effectiveness depended not only on its own design but also on how developing countries could leverage their power in a new multilateral trade institution.

Content-related challenges

Trade delegates, WTO officials, and business groups all note the importance of the material contained in TPR reports. At the same time, many officials call for more analysis and better presentation of data as necessary changes to make the reports salient and usable. Set against these contradictory views, we need a scientific basis for evaluating the content of the WTO Secretariat's reports. The Secretariat has the task of preparing reports that cover the macroeconomic environment, the domestic institutional structures for trade policymaking, and full examination of trade policies by measures and by sectors. It cannot make legal judgments about members' policies, but can provide economic analyses. Although it relies on official government data, it also uses other documents and sources for the reports.

Transparency via the TPR process is important for members in two ways: when a country is itself under review, and when its major trading partners are under review. Much existing scholarship focuses on the former, the belief that domestic transparency could strengthen the hands of trade policy reformers.Footnote ⁸⁴ But there is no analysis of the latter aspect. If the reports have to be relevant, they should contain information and analysis on issues that are of greatest concern to member states.Footnote ⁸⁵ I use this standard of ‘issue salience’ to evaluate the content of a large sample of TPR reports. The TPRM has the potential to increase compliance with trade rules at home and abroad.

What are the critical or salient issues? I conduct an analysis that links the content of government and Secretariat reports to the Dispute Settlement Mechanism (DSM). The disputes that a member initiates are treated as proxy indicators of the most important trade concerns for that country. The analysis asks whether information about these issues and concerns were revealed in TPR-related documents.

Note that this analytical link between TPRs and the DSM does not claim any causal link between multilateral surveillance and legal dispute settlement. It merely evaluates the ‘information dissemination’ and ‘compliance promotion’ functions of the WTO's information system, namely that of revealing information that would be useful in promoting better adherence with the regime's rules. That, indeed, is the mandated purpose of the TPRM.

For the analysis, I construct a database of dispute-dyads of complainants and respondents in the DSM. A total of 172 dispute-dyads were analysed. The complainants in the dispute-pairs cover all developing, least developed economies and transition economies that had initiated a dispute at the WTO (until end-2007).Footnote ⁸⁶

Each dispute might have more than one complainant, but every complainant is treated separately, in order to consider its interests in the issue under contention. The focus, however, is on the TPRs of the respondent in each dispute. The objective of the analysis is to examine whether the respondent's TPR refers to, analyses, or criticizes the policy that is disputed. For each dispute, the database gives evidence from the TPR reports that both preceded and followed the request for consultations on a dispute.

The rationale for examining the preceding report is to check if the report highlighted the contentious policy. It also helps to examine whether WTO members other than the complainant also use the TPR as an opportunity to raise questions about the disputed issue.

After analysing 172 dispute-dyads, the results reveal interesting insights. In some cases, the Secretariat reports did mention policies that would eventually be taken up by complainants as formal disputes. This occurred only 53% of the time for TPRs of respondents preceding the initiation of disputes, by no means a signal of extensive coverage. Further, only 15% of the reports had any detailed policy analysis or criticism of the contentious policies. There is no significant difference between the reports of developed and developing countries. For the former, the average is 54% and for the latter it is 50%. But analysing the reports by year we find that after 2000 the coverage of critical issues for developing countries increased a great extent from negligible levels; for developed countries' reports, the share remains more or less stable.

The rationale for examining the succeeding report (after disputes are initiated) is slightly different. It relates to the dynamic of peer pressure by which the TPRM is expected to promote compliance. If succeeding reports highlight and analyse the issue under contention, they serve to apply pressure on the respondent. Moreover, the Dispute Settlement Body almost invariably defers the establishment of a panel, which gives parties time to resolve the dispute via consultations. A TPR during this period provides a further opportunity to analyse the merits of a disputed trade measure.

After disputes had been initiated, the Secretariat reports mentioned or analysed the contested policies 80% of the time but in many cases, it was only a footnote mention. Policy analysis or criticism could be found in only 16% of the reports. In fact, since 2003 there has been a downward trend in the average coverage of disputed issues. The proportion of reports covering such issues fell from a high of 90% in 1997 down to 43% in 2004 and 50% in 2006. This is accounted for largely in the reports of developed countries: in 2005 and 2006, only a third of the reports of developed country respondents mentioned or analysed the contentious issues under dispute. Finally, just a quarter of the government reports even mentioned the disputed policies; even fewer offered any clarification on them.

The above trends reveal a real content-related challenge for the TPRs. The Secretariat has a vast amount of information to collect and collate in the report with limited resources at its disposal. In the process of aiming to be comprehensive, it is perhaps falling short in highlighting the more critical aspects of members' policies, aspects that would be relevant for other members of the WTO.

Participation-related challenges

Participation in review meetings takes many forms. Members are allowed to pose questions in advance of the meetings. Members also raise issues of concern during the meetings. Both avenues of peer pressure require that delegates actively engage with the process. A third way is participation by mere presence, since TPRs are also learning exercises from which even passive participants could gain. No empirical evidence currently exists on any aspect of participation.

When it comes to applying peer pressure on the critical/disputed issues, the record is poorer. Prior to disputes, in only 24% of the cases members pose questions on issues that they consider critical enough to request consultations. After disputes had been initiated, whether reports mention or analyse the issue should be irrelevant to complainants. Yet, the most surprising result is that the proportion of complainants submitting advance questions after disputes had been initiated fell to 19%. Thus, contrary to what the TPRM had been envisaged for, in general member states do not consider the TPR process to be an effective forum for applying pressure. Moreover, declining trends in recent years suggest a growing disillusionment with the purpose or use of the mechanism.

Peer pressure can also be applied by countries other than dispute complainants. In fact, this is the advantage of a multilateral review process. Thus, for each dispute-dyad we also investigated how other members behaved. The results are more encouraging: on average, prior to a dispute, other states sent in advance questions on the contested issue 61% of the time. After the onset of disputes, other parties continued to submit written questions 55% of the time. Although complainants have not submitted advance questions, the reviews have given opportunities for the WTO's membership as a whole to collectively apply pressure on each other. However, comparing the number of questions members submit when their biggest trading partners are under review with how many questions they get in return for their own reviews, we find that the direction of peer pressure is largely skewed against developing countries.

A related challenge pertains to participation in review meetings. Many supporters of the TPRM stress that simply participating is a good exercise in transparency and learning. The meetings hold an important status in the WTO. The Trade Policy Review Body is equivalent to the General Council, comprising all member states and giving developing countries equal opportunities to review the policies of richer members. Among other international economic institutions, this is a very different setup, compared to say the weighted voting system in the IMF, which limits the voice of poor countries, or OECD reviews in which developing countries cannot participate.

However, developing country participation in the TPR meetings remains poor. I measure which members raise questions or make statements during the meetings. I use documents from all TPR meetings held during 1995–2007 (190 in total, barring nine apparently random cases where some documents were missing) to measure participation by different countries. The analysis records the participation of 70 countries: the top 20 trading powers, additional member states that have initiated disputes, leaders of negotiating coalitions, and LDCs that have undergone a TPR. This survey presents the first comprehensive account of participation in TPR meetings (arguably, the first in any monitoring mechanism in the GATT/WTO).

Contrary to expectations for larger trading powers, the results show very low levels of participation even among major developing countries. The top four traders almost always participate in review meetings. But even among the next 16, there are only a few active participants: only a minority has participated in more than half the meetings (Figure 1). Worse still, an average LDC has participated in just three meetings. The likelihood that an LDC will ask questions or raise a point is just 2%.

Figure 1.

A majority of WTO members has not participated in even half the TPR meetings

How can we explain the low participation of even the larger developing countries? The strength of delegations in Geneva, the scheduling of meetings, and the challenges of analytical capacity are only partial explanations. A bigger problem is the lack of political interest in the reviews. Although there is no recorded evidence, anecdotal accounts suggest that countries often send middle-ranking officials, whether they are under review themselves or they are participating in other members' reviews.Footnote ⁸⁷ Another reason for low participation is the unwillingness of members, especially developing countries, to ask each other embarrassing questions. My own observations of the review meetings found that much of the commentary was a repetition of what was already in the Secretariat's reports. And where delegates did speak up, it was usually to praise the country under review or make long statements about the historical and cultural links between countries, issues that had little to do with trade policy.

TPR meetings, instead of being opportunities to debate members' policies, have increasingly become forums for ‘negotiated reviews’ at best, according to one Latin American delegate.Footnote ⁸⁸ At worst, because of the lack of interest and the low levels of participation, the meetings are so lacklustre that another ambassador suggested they were ‘boring to death’.Footnote ⁸⁹

Despite some exceptions, the general trends reject our initial conjectures that at least larger traders, with offensive interests in the regime, would participate more actively in its monitoring mechanism. And the fact that countries do not raise questions even on issues directly affecting them suggests that the peer pressure envisaged for the TPRM does not work. For developing countries, the TPRM proves to be weak on both content and pressure: not only are TPR reports often silent on controversial issues, TPR meetings are toothless as well.

Relevant content for developing countries

WTO members need to identify minimum standards for surveillance. Much discussion about the scope of issues to be covered occurs at a general level. But member states need advance warning of policy changes to lower the risk of hidden/inaccurate information. Policymakers have to decide which policies monitoring mechanisms (whether the TPRM or issue-specific ones) would highlight, and to establish templates for factual and timely reporting.

The Secretariat could also use more sector- and country-specific expertise for drafting TPR reports. Collaboration and coordination between its divisions could be made more formal for the purposes of monitoring. Moreover, developing countries might prefer other forums like the Committee on Trade and Development to have an input in the review process. They have secured this provision for RTA monitoring.Footnote ¹⁰² Overcoming issue uncertainties by drawing on multiple sources of expertise could greatly increase confidence in the quality of reporting and provide information on flexibilities for implementing rules.

Reviews should also analyse the systemic impact of policies in major trading countries. This is not restricted to developed countries, but large developing countries as well. Resource-constrained members would support analyses that explain how regulations in a major country affect their trading interests. Assigning desk officers who would engage in regular research (not just at the time of TPRs) would be a step towards building such capacity. Similarly, analysing the systemic impact of the regime's rules warrants a closer working relationship between individual committees, the TPRD, and the Economic Research and Statistics Division. It also means that coordination between the WTO and the IMF/World Bank would have to increase: building tariff and related databases, evaluation of macroeconomic conditions, impact analyses of trade negotiations, and linkages between technical assistance needs and development assistance.

Alternative sources of information

Although many agencies and actors research trade policies, their efforts are not always coordinated in a way that countries could benefit from timely information.

Therefore, domestic transparency and monitoring has to improve, with particular emphasis on the processes of consultation and coordination. Among developing countries, Brazil, Chile, China, India, Hong Kong, and Korea are among the countries that already have wide-ranging consultation processes, albeit to varying degrees of success. National trade policy forums could draw in legislators, business groups, and NGOs in periodic dialogues on trade policy and get inputs from multiple sources about regulatory changes abroad. Long gaps between multilateral reviews often result in forgotten policy lessons. Therefore, interim national reviews could take stock of changes and improvements in policies.

The role of non-state actors in monitoring should not be underestimated. The more states remain reluctant to strengthen WTO-led monitoring, the greater will be the potential of NSAs to close the information gaps.Footnote ¹⁰³ Existing literature often assumes that in the trade regime, non-state actors would be equally disposed towards trade monitoring. In reality, their capacity varies greatly across countries. Capacity building efforts that focus on explaining WTO rules to domestic actors would need to be supplemented with measures to improve the policy review and monitoring capabilities of non-state actors.

Resource-constrained countries could also gain from better regional surveillance. The aid-for-trade review process has already incorporated this dimension and the TPRM also undertakes a few regional reviews. Regional peer review could increase the availability of timely data, discussion on the appropriate models and frameworks for analysis, and generate more options for policy implementation. These dynamics could also translate into more effective peer pressure.

More effective forums for peer pressure

The record of WTO surveillance shows that members are reluctant to incorporate strong follow-up procedures for monitoring exercises, fearing that they could lead to disputes. So, it would be unreasonable to expect new monitoring mechanisms to buck the trend. Nevertheless, there are other means to increase peer pressure and promote compliance.

Members could target each others' reporting behaviour. Thus, stronger provisions for notifications would have to be combined with pressure against non-reporting countries. Failure to report on changes in policies should attract more explicit criticism in various WTO committees as well as during TPRs. As opposed to sanctioning non-compliance, which has been the main worry for members, an emphasis on better notifications would sanction non-reporting and increase the available information.

Another route is continuing to push for procedural changes in review meetings. Although this has had limited impact in the past in the absence of other reforms, improvements could continue incrementally, by using external discussants, by ensuring that reviews do not clash with other meetings, and by making high-level representation in meetings mandatory.

Further, members and the WTO have to think about greatly increasing the dissemination and visibility of reports and their use. This implies generating more media coverage of the reviews and permitting non-state actors to participate as observers. Moreover, within the WTO, coordinators of negotiating coalitions could outline the implications of their trade rivals' policies, so that reviews feed into negotiations and a structured process of demanding policy change. It also means that, at least on less controversial issues, specific procedures are in place to follow-up after reviews. More structured links with the Integrated Framework and with wider aid-for-trade efforts could go a long way in giving poor countries a stake in more effective monitoring.

Effective monitoring is a necessary condition for international cooperation. By undertaking an in-depth empirical investigation of the TPRM, this paper has offered ideas to experiment with new innovations in the trade system. With a growing demand for information, trust in multilateral monitoring would increase when it shines the light on issues that matter to developing countries and is seen to have an impact.