On 16 December 2011, Russia gained the approval of 153 other member states to join the World Trade Organization (WTO). During its arduous eighteen-year accession, Russian officials reformed a wide range of laws and policies that could affect trade.Footnote 1 Russia made these changes because senior Russian officials believed that improved governance would pay off in greater foreign investment, and that investment in turn would stimulate economic growth.Footnote 2 President Vladimir Putin acknowledged ‘countries compete in the attractiveness of their business climate … quality of state institutions and effectiveness of the court and legal system’. Ivan Tchakarov, chief economist at the Russian brokerage Renaissance Capital, asserted ‘By becoming a WTO member, Russia will have to import … rules and regulations that will address the very issues that foreign investors usually complain about, like corruption, the protection of minority shareholders, the independence of the judiciary.’Footnote 3
In this article, we define governance as ‘mechanisms, processes and institutions through which citizens and groups articulate their interests, exercise their legal rights … and mediate their differences’ (UNDP, 1997). Good governance follows the rule of law and is transparent, responsive, effective, and efficient.Footnote 4 Due process, evenhandedness, and transparency are important metrics for good governance and can act as anticorruption counterweights. They are also norms of the GATT/WTO (Nadakavukaren Schefer, Reference Nadakavukaren Schefer2008: 21; Ala'i, Reference Ala'I2009: 269–70; Halle and Wolfe, Reference Halle and Wolfe2007: 8).
The WTO and its predecessor organization, the GATT, comprise the international system of rules governing trade.Footnote 5 WTO rules do not actually discuss governance. But the WTO Secretariat and other observers have suggested that improved governance is a spillover of its efforts to promote open trade (Aaronson, Reference Aaronson2001; Trujillo, Reference Trujillo2007; World Trade Organization, 2005).Footnote 6
Members of the WTO also use the system to improve governance among nations, such as Russia that want to join this rules-based system.Footnote 7 Working parties comprised of other WTO member states closely monitor potential members, and make sure these nations meet their commitments. Once these acceding nations become members, they are monitored through trade policy reviews and can be challenged in a trade dispute. In short, the WTO has a feedback system to hold members to account for their trade practices and governance commitments.
Policymakers have termed this process of using membership as a lever to improve governance ‘policy anchoring’ (Francois, Reference Francois, Galal and Hoekman1997; Bachetta and Drabek, Reference Bacchetta and Drabek2002; Basu, Reference Basu2008). They have stressed that anchoring occurs during the year(s) of accession.Footnote 8 Our work differs from these scholars in two ways: strategy and findings. We used metrics that track WTO norms to explain how the process occurs and by so doing are able to nuance our understanding of when policy anchoring occurs. We argue that members improve governance by adhering to three norms that are core values of the WTO (as well as widely recognized attributes of good governance). The three norms of good governance promulgated by the WTO are even-handedness (what the WTO calls non-discrimination); access to information (what the WTO calls transparency); and administrative due process (the ability to review, comment upon, and challenge trade-related policies).Footnote 9 To ascertain when policy anchoring occurs, we divided countries into four groups. We call the first group non-members: it includes countries that never joined the WTO as well as countries that are attempting to join but have not yet been accepted. Some of the non-members completed their accession negotiations during our period of study; we label them the completed negotiating group. Countries that have gained approval to accede must then must gain domestic approval (which may take several years) to join the WTO. If they succeed, they become part of the new members group (countries that became members after 1995 – we note that our data begins in 1996). Finally, we call nations that joined the GATT between 1948 and 1995 long-standing members. Because our period of study extends from 1996 to 2011, a country such as Cambodia could be a member of three different groups, moving from non-member to completed negotiations (2003) to new member (2005).
We found that nations change their laws and policies to join the WTO during and after the accession process. Using our more exact metrics, we found that countries that successfully negotiate to join the WTO exhibit improved performance on metrics for evenhandedness and access to information during the negotiating process, but we did not find commensurate improvements in performance on metrics for due process. New members showed mixed effect; while this group improved their performance on metrics for access to information evenhandedness weakened in new WTO members and we found no significant effects with metrics for due process. Moreover, long-standing members improved their performance on access to information and due process, but worsened their performance on metrics for evenhandedness. In sum, the WTO's effect on governance was uneven – policymakers improved their performance on these metrics some of the time. Interestingly, because our data are not limited to governance in the trade regime but cover the polity as a whole, our empirical evidence provides partial support for our hypothesis that the norms of good governance promoted by the WTO transcend the trade sphere and affect the country's approach to governance in general.
Parsing the ‘how’ and ‘when’ of policy anchoring
Prominent scholars such as Basu, Tang, Wei, and Ferrantino have tested this policy anchoring process by correlating membership and quality of governance. They have relied on metrics for governance developed by widely respected sources such as the World Bank and Freedom House. These broad metrics are based on expert surveys of policymakers, human rights and governance activists, and business leaders. Analysts ask these experts a wide range of questions and then aggregate the answers into one numerical assessment of governance, which they then describe as broadly reflective of good governance. Thus, some scholars call these metrics ‘mash-ups’.Footnote 10
However, we believe these governance metrics may be too broad to examine how policymakers translate WTO norms into behavior. For example, scholars use Freedom House metrics to ascertain the extent to which government functions properly – whether policymakers act in a responsive, responsible, and accountable manner. Such proper functioning includes elements of ‘due process’. But the Freedom House metrics are much broader than due process per se. To ensure that our findings build on the work of these scholars, but effectively assess the impact of the WTO, we decided to test country performance on two sets of data: the metrics widely used by other scholars cited in the literature section below, and the narrower metrics which we believe more closely track the three WTO norms. We believe this strategy allows us to hone in on how the norms of the WTO are translated by national policymakers into governance practices. By relying on both broader and more narrow metrics, we are able to develop conclusions as to whether the WTO norms have spillover effects on the polity as a whole. However, our metrics are also inexact. For example, our metrics for even-handedness and due process clearly portray non-discrimination and the right to appeal, but other scholars might reasonably argue that these metrics do not exactly replicate how non-discrimination and due process work in the context of WTO membership. We acknowledge these concerns and hope that our work will encourage other scholars to develop more exact metrics that track WTO norms and/or reflect more narrow governance outcomes.
We hypothesized that if policy anchoring is occurring when nations attempt to join the WTO, these acceding states should show dramatic improvements in good governance during the years of accession. So we should see changes in both the completed negotiations group. However, if policy anchoring occurs during membership, we would see greater evidence of improvements in our metrics for governance for both new and long-standing members of the WTO.
We used several different models to test country performance on both the broad and narrow metrics. When we used the broad mash-up metrics for governance, we find no support for policy anchoring. The negotiating group showed declining performance on metrics for due process and access to information, while improvements in evenhandedness were insignificant. New members were weaker on all three metrics, while we found no significant effects for long-standing members. But, as noted above, we did find improvements when we used our more narrow metrics. As a result, we believe the mash-up metrics are too broad to accurately replicate the effects of WTO membership on governance.
We believe our varied findings can be explained by several factors. First, good governance takes expertise, funds, and will; hence we should not be surprised that our empirical analysis shows uneven improvements in governance over time. New member states may reform their laws and policies, but policymakers may at times be unwilling or unable to effectively change behavior, institutions, and culture. Policymakers must learn how to create inclusive, responsible, and responsive political and economic institutions. These institutions both stimulate and regulate an economy that can harness the talents, skills, and creativities of their citizens (Acemoglu, Reference Acemoglu, Aghion and Durlauf2005; Acemoglu and Robinson, Reference Acemoglu and Robinson2012). Members ‘learn’ how to improve governance and adopt the norms of improved governance as they participate in the day-to-day workings of the WTO (Sandholtz and Gray, Reference Sandholtz and Gray2003). However, this learning process is gradual and uneven. Sometimes countries make significant progress in some areas; the same countries may make less progress in others.
Our findings may also be explained by the inadequacy of WTO strategies to consistently hold countries to their obligations to adhere to the three WTO norms. Although members review each other's performance at trade policy reviews, trade disputes are the main means of holding countries accountable. But these disputes center on how non-adherence to WTO norms affects market access of other WTO member states, not on adherence with WTO norms per se. Moreover, these disputes are rare.Footnote 11 As a result, the threat of a dispute may not compel member states to change or improve their behavior to consistently adhere to their WTO obligations.Footnote 12 In sum, because efforts to enforce compliance are irregular, member states cannot always hold other member states to account unless they challenge another nation through the expensive process of a trade dispute. So while the dispute settlement mechanism is binding, policymakers are also well aware that they will not be consistently monitored for their governance practices, unless these practices directly distort trade.
Our analysis proceeds as follows. We begin by delineating the specific GATT/WTO norms of due process, evenhandedness, and transparency and show how they change the behavior of member states. We then discuss the theoretical literature that attempts to explain why nations change their behavior as they join the WTO and use that discussion as the basis for our theoretical argument and hypotheses. Next we examine qualitative evidence of accession and trade policy reviews to see if member states alter their approaches to governance to foster due process, even-handedness, and transparency. We then examine our hypotheses quantitatively to see if the accession process and membership in the GATT/WTO affect our metrics for due process, access to information, and evenhandedness. Finally, we conclude with a discussion of the theoretical and policy implications of this research.
Theoretical background: how membership in international organizations may improve governance
Many scholars believe that nations will improve their governance after policymakers join an international organization (Simmons and Hopkins, 2005; Von Stein, Reference Von Stein2005; Bauhr and Nasirtousi, Reference Bauhr and Nasiritousi2009; Sung and Chu, Reference Sung and Chu2003). But these same scholars posit different reasons for these spillovers. Robert Keohane argues that states join international organizations to reduce information asymmetries and ensure that all members benefit from the ‘public good’ of rules-oriented trade (Keohane, Reference Keohane2005: xi). Some scholars term this theory ‘neoliberalism institutionalism’ (World Trade Report, 2007). Other scholars see membership as a means of formalizing commitments (Goldstein and Martin, Reference Goldstein and Martin2000; Sung and Chu, Reference Sung and Chu2003) and spreading key norms of good governance and open markets (Simmons et al., Reference Simmons, Dobbin and Garrett2008; Sandholtz and Gray, Reference Sandholtz and Gray2003). By adhering to WTO rules, some scholars have found that governments may also advance some democratic rights (Aaronson and Abouharb, Reference Aaronson and Abouharb2011) or democracy (Keohane et al., Reference Keohane, Macedo and Moravcsik2009).
Other academics have sought to answer whether or not developing countries use the WTO to send a message to markets and policymakers. These scholars tend to view the WTO both as a signaling device and a policy anchor. Countries want to signal that like other mature democracies, officials will respect property rights and the rule of law (Elkins et al., Reference Elkins, Guzman and Simmons2006; Dobbin et al., Reference Dobbin, Simmons and Garrett2007; Buthe and Milner, 2008; Mansfield and Pevehouse, Reference Mansfield and Pevehouse2008: 273). Mansfield also believes emerging democracies use membership in the WTO to signal to their citizens that the government is accountable (Mansfield et al., Reference Mansfield, Milner and Rosendorff2002: 478–481). If Mansfield et al. are correct, these governments will use membership to learn how to make policies in a more transparent, accountable, and evenhanded manner.
It is not easy to join the WTO. First, as described above, new members must reduce trade distorting rules and regulations and make significant changes to their governance processes and strategies. Countries with better governance may find accession easier than countries that must make major institutional and policy changes (Bacchetta and Drabek, Reference Bacchetta and Drabek2002: 16; Evenett and Primo Braga, Reference Evenett and Primo Braga2005: 1). Federal officials may encounter resistance from once-privileged groups when they try to make these changes. Some citizens may view the reform process as coercive and alienating (Michalopoulos, Reference Michalopoulos1998; Langhammer and Lucke, Reference Langhammer and Lucke1999). In addition, some countries have trouble adopting WTO norms because their culture emphasizes the sovereignty of the bureaucracy over the right of individuals for access to information (Potter, Reference Potter2003; Biukovic, 2008). Hence, these countries may resist greater access to information and transparency. Thirdly, WTO membership also strengthens the power of the central government vs. regional elites, so these individuals may resist change (Langhammer and Lucke, Reference Langhammer and Lucke1999; Tang and Wei, Reference Tang and Wei2006). Finally, policymakers must accept increased scrutiny and foreign participation in the polity; leaders may find such changes mean they have less control over their nation's policy choices.
Despite these costs, almost every country outside of the organization wants to join the WTO. UNCTAD theorized that policymakers may see their commitments as ‘investments … they are payments today in the expectation that they will produce rewards in the future’ (Basu, Reference Basu2008: 5). Membership also signals foreign investors that the country will provide foreign and domestic actors with the information they need to assess market and political conditions (Barton et al., Reference Barton, Goldstein, Josling and Steinberg2006; Tang and Wei, Reference Tang and Wei2006; World Bank, 2006; Honda, Reference Honda2008).
Some scholars have examined this policy anchoring process in depth. Bachetta and Drabek used World Bank metrics of institutional quality and found that countries that joined the WTO between 1995 and 2000 had better institutional quality than non-members (Bacchetta and Drabek, Reference Bacchetta and Drabek2002: 17, 41). However, their study did not examine change over time and did not control for other factors that might affect institutional quality, such as foreign aid.
Tang and Wei hypothesize that WTO commitments made at accession have a stronger effect on countries with poorer governance. They checked their results on several different metrics for good governance and find that the policy changes required by accession benefit countries with inadequate governance (Tang and Wei, Reference Tang and Wei2006: 18–21). Ferrantino finds neither free trade agreements nor WTO accessions have a significant influence on governance. He argues that this finding reflects inadequacies in governance metrics rather than reality (Ferrantino, Reference Ferrantino2006). Basu delineates the specific changes required to join the WTO and compares newly acceded states to existing developing country members. He shows that obtaining WTO membership can lead to a positive improvement in domestic policies (Basu, Reference Basu2008). In short, Basu finds the investment in joining the WTO pays off not only in expanded trade, but also in better governance.
Like these scholars, we hypothesize that membership in the WTO can yield governance improvements. We agree that countries alter their laws and policies when they accede, but it takes time for such changes to improve governance. However, we focus our analysis on the transmittal of WTO norms.
Our theoretical and empirical contribution
We begin by delineating exactly how WTO norms change the behavior of new and long-standing member states. The section below delineates the specific WTO obligations that influence both the behavior of member states and market actors. We see these obligations as good governance norms. Specifically, the WTO requires that governments promote:
-
• Evenhandedness (also known as non-discrimination): Governments must not discriminate between foreign and domestic market actors (GATT's MFN and national treatment obligations). Member states must ensure ‘effective equality of competitive opportunities’ between foreign and domestic like (similar) products and services. Article III requires non-discrimination both in the letter of the law and in the manner in which laws are applied.Footnote 13 The WTO describes this as ‘treating other people equally’.Footnote 14
-
• Transparency and access to information: The WTO says that transparency is essential to the functioning of the global trading system, and a means of enhancing national and international legitimacy.Footnote 15 The WTO defines transparency as the ‘degree to which trade policies and practices, and the process by which they are established, are open and predictable’. Transparency and access to information help make the WTO's rules and processes accountable both to member states and their citizens. Governments must make trade-related policies in a transparent manner and ensure market actors can be made aware of such provisions.Footnote 16 These norms act as a check on arbitrary or discriminatory policies or practices. These rules also provide market actors with the information they need to challenge trade-related policies and decisions.
-
• Due process: Governments must accord due process rights to market actors. GATT Article X: 3(b) requires each party to maintain judicial, arbitral or administrative tribunals or procedures for the purpose … of the prompt review and correction of administrative action relating to customs matters. These tribunals must be independent of administering agencies and allow importers to lodge appeals. Moreover, individuals with interests in investigations have a right to receive notice, to present written evidence, to review the public docket, to challenge decisions and to seek judicial review (Charnovitz, Reference Charnovitz2001: 20).
As noted above, GATT/WTO rules obligate policymakers to act in certain ways when they develop and implement trade-related policies. Through WTO day-to-day activities and peer pressure, we hypothesize that the GATT/WTO should not only improve governance through adherence to their obligations but also over time through learning from other member states. We believe this process may begin during accession. If member states are anchoring to the WTO, policymakers will gradually begin to apply the norms of even-handedness, transparency, and due process to other aspects of economic and political governance after they join. But the effects of the WTO may not be limited to trade-related governance. We believe WTO norms will gradually spill over into the polity as a whole and be reflected in our empirical results. For example, under GATT/WTO rules, policymakers are obligated to act in an evenhanded and predictive manner on trade-related policies. They are obligated to create mechanisms to ensure that foreign and domestic actors are treated equivalently. We believe policymakers should gradually learn to avoid favoritism among foreign and domestic market actors, not just on trade issues, but on general governance. Secondly, trade policymakers are required to provide market actors with access to information about trade policymaking before decisions are made. In so doing, these officials should gradually learn to act in a transparent manner so as to provide clarity and certainty to market actors. Finally, policymakers are obligated to allow individuals to comment on and challenge trade-related regulations before they are adopted (a form of due process). As they accept questions and challenges to trade policies from market actors, they will learn to be responsive to their citizenry. Please note we are not saying WTO membership immediately transforms governance; the improvements are gradual, uneven, and differ among members. Table 1 above summarizes our argument and describes how governance learning spills over into the polity as a whole.
How GATT/WTO norms affect policymaker behavior and how this spills over into the polity
We developed two hypotheses to test our theoretical argument:
-
(H1) If states anchor during accession, they will improve their performance on the metrics for due process, access to information, and evenhandedness as they work to join the WTO.
-
(H2) If states anchor during membership, they will gradually learn to improve their governance and show improvements on the metrics for due process, access to information, and evenhandedness after accession.
If hypothesis one is correct, we will see significant improvements among countries in the successful negotiating group, compared to the non-member group. If hypothesis 2 is correct, we will see slow but significant improvements among new members compared to non-members and the negotiating group.
Qualitative evidence
Research design
In our qualitative research, we reviewed the channels with which member states may transmit these norms in the day-to-day workings of the WTO. These channels include accessions, daily discussions, trade disputes, trade negotiations, and trade policy reviews. Herein, we focused on accessions and trade policy reviews; the accession documents spell out country commitments, and trade policy reviews are a means to hold countries to account for their membership commitments. We examined every WTO accession from 1995 to 2011 (25 countries). We also examined the bulk of recent trade policy reviews for a diversity of WTO member states from 1995 to 2011 (24 countries). Our sample included a wide range of countries, including new and long-term members, large and smaller traders, as well as developing and industrialized countries. With these two avenues and a diversified sample, we could draw comprehensive conclusions as to whether member states were altering their policies and whether other states saw these nations as improving governance. We did not look at trade disputes because they are rare and these disputes center upon the trade consequences of non-compliance with WTO rules related to a particular good or service.
Accessions
Most of the world's 195 nations and customs entities are either members or countries that want to join the WTO.Footnote 17 As of May 2012, 155 nations belonged to the WTO, and 30 of the 39 countries that are not members of the WTO are trying to accede.Footnote 18 The accession process can be arduous and long. The candidate country produces a memorandum describing its trade policy. Next, a working party of WTO members delineates a protocol of accession which spells out what changes the country should make as well as a timetable for such reforms. The current members must agree on the terms of membership for a new member; these terms differ for each country.Footnote 19 Current members demand that potential new members make commitments on market access as well as commitments on rules. Ultimately, if two thirds of the members approve, the acceding country can ratify the agreement and then join the WTO.Footnote 20
Current members use the discussions over the protocol to prod the potential members to take steps that support capitalism, increase transparency, and advance the rule of law (Bacchetta and Drabek, Reference Bacchetta and Drabek2002: 11). Member states want to ensure that these new members do not make empty promises.
We reviewed working party reports and accession protocols for every recent accession from 1995 to 2011, or some 25 countries. We found countries promise significant changes and make many of these changes before accession is approved.Footnote 21 Moreover, the working parties make a checklist and monitor country steps to alter their laws as well as to regulate specific sectors or processes. At the behest of the working party, the WTO secretariat monitors specific actions such as a legislative action plan and a check list of changes to regulations on technical barriers to trade, sanitary regulations, and intellectual property rules, among other policies.Footnote 22 These documents illuminate the broad swath of legal and administrative reforms countries willingly make to join the WTO.
Many of the countries that have acceded since 1995 have significant problems with accountability, transparency, and evenhandedness. For example, Georgia admitted that its trade-related regulatory system was opaque, so it promised to change its legal system to conform to WTO rules. The representative of Georgia also confirmed that from the date of accession, Georgia's laws would provide for the right to appeal administrative rulings on matters subject to WTO provisions to an independent tribunal in conformity with WTO obligations.Footnote 23 Saudi Arabia and Nepal agreed to increase provisions for transparency and public comment.Footnote 24 Members spent hours asking questions about the rights of Saudis and non-Saudis to participate in the economy.Footnote 25 Saudi Arabia agreed to ‘provide a reasonable period … for members, individuals, associations and enterprises to provide comments to the appropriate authorities before such measures were adopted’.Footnote 26 Cambodia agreed to remake its judicial and administrative law systems. ‘Cambodia recognized the need to establish an appeals process, both administratively and to an independent tribunal to meet the requirements of Article X of the GATT 1994 and other such provisions in WTO Agreements.’ Working party member governments also reminded Cambodia that it was obliged to develop ‘mechanisms for publication and dissemination of draft legislation and standards for public comment; [and] the establishment of a TBT (technical barriers to trade) Inquiry point’, where foreign and domestic producers could learn how to meet Cambodian standards.Footnote 27 The representative of Jordan said that from the date of accession all laws, regulations, decrees, judicial decisions, and administrative rulings of general application related to trade would be published in a manner that fulfills the WTO requirements. Jordan was also questioned as to whether its court system provided rights for administrative appeal and the role of the king in ruling on such appeals.Footnote 28
China was the only country whose accession agreement required policymakers to improve the rule of law. The 2001 Protocol states that China must enforce ‘uniform administration of Chinese law’ throughout China (Aaronson, Reference Aaronson2007). The Protocol calls on China to ‘apply and administer in a uniform, impartial and reasonable manner all its laws, regulations, and other measures of the central government as well as local regulations, rules, and other measures pertaining to or affecting trade. China shall establish a mechanism under which individuals and enterprises can bring to the attention of the national authorities cases of non-uniform application.Footnote 29 The agreement requires China to ensure that ‘laws, regulations and other measures pertaining to and affecting trade shall be enforced’.Footnote 30
As the world's largest trading nation, trade policymakers, business leaders, and others closely monitor China's adherence to its WTO obligations.Footnote 31 While the US Government generally lauded China for meeting its obligations, it also concluded that parts of the Chinese government ‘had not yet fully embraced the key WTO principles of market access, non-discrimination and transparency, or the carefully negotiated conditions for China's WTO accession’.Footnote 32 In the 13 December 2011 testimony, the Assistant US Trade Representative for China noted, ‘China does not publish measures providing what China calls “internal guidance” to its agencies. These measures can bind agencies just as fully as officially public measures do, and the public should be able to see them.’Footnote 33 The US China Business Council (USCBC) regularly monitors China's adherence to WTO norms of transparency. In 2011, it found China posted less than one quarter of relevant documents for public comment and often didn't post for the required 30 days. USCBC also noted that China does not maintain records of items that have been released for comment, ‘making it difficult for anyone to track transparency in a comprehensive fashion’.Footnote 34 In 2012, the USCBC found China had improved its performance: some 62% of administrative regulations and departmental rules were posted for comment, but only 54% of these documents were posted for the required 30 days. Chinese observers also found mixed results. The Chinese Academy of Social Sciences evaluated government ministries on transparency. It ranked the highest scoring ministry 67 and the lowest scoring 51, on a scale of 0–100.Footnote 35 In sum, observers find China is making great strides in increasing its transparency, but it has not yet achieved open and responsive government.
Taken in sum, our qualitative evidence indicates that new members make major changes to both their laws and how they govern. Countries clearly adopt measures that enhance access to information, improve evenhandedness, and provide due process regarding trade-related policymaking. During the accession process, states engage in a debate with their citizens over the substance of trade policy and how trade policy is made.Footnote 36 Over time, these reforms may give citizens in weak democracies or authoritarian regimes opportunities to influence trade-related policymaking (Aaronson, Reference Aaronson, Chauffour and Maur2011; Aaronson and Abouharb, Reference Aaronson and Abouharb2011). In so doing, the WTO may have unanticipated democratic spillovers such as helping countries create a feedback loop between the government and the governed on trade and trade-related policies. In fact, according to the WTO, feedback loops create not only dialogue and expectations, but can also promote greater accountability.Footnote 37 However, as China illuminates, it takes a long time for government officials to improve governance and develop a culture of transparency and accountability. Traditional elites may resist change or thwart access to information, fearing that they could lose power. Hence, the accession process may lead to de jure changes but these changes may not affect market actors consistently or equitably.
The trade policy review mechanism
Since 1989 (under GATT), and continuing under the WTO, member states have formally and publicly reviewed each other's trade policies in trade policy reviews. The US, EU, China, and Japan are reviewed every two years, the next 16 nations (in terms of their share of world trade) are reviewed every four years, and the remaining countries are reviewed every six years.Footnote 38
Trade policy reviews (TPRs) address a wide range of governance issues. Trade diplomats from other member states review the subject country's approach to taxation, customs, agriculture, industrial policies, health, environmental and safety regulations, authority of sub-governments, as well as the effectiveness of the administrative and judicial systems. The review meetings are not open to the public but they are made public on the WTO's web site some six weeks after they occur. Therefore, citizens of WTO member states can use the review to gain broader insights into a country's policies, and trade officials can gain valuable feedback on their nation's WTO adherence.Footnote 39 Zahrnt argues (Reference Zahrnt2009: 6) that the TPR teaches the habits of good governance, because it ‘accustoms governments to tolerate reviews, stakeholders to contribute to the review process and the media to use the results’. These reviews also enable interested parties to compare trade policies across countries. Trade policymakers from the country under review may use these reviews to encourage legislators to make policy changes. Some scholars also argue that the review process makes reforms more credible (Francois, Reference Francois1999).
Trade officials try to use this process to hold other governments to account. Sometimes, officials laud other governments, at other times they criticize foreign states' policymakers for bad behavior. As we read the minutes of the TPRs, we found members often reject or ignore criticism – China is a good example. But the WTO Secretariat believes the process is improving governance. In 2010, WTO secretariat staff reviewed the impact of these TPRs in the Americas and concluded that these nations had become increasingly transparent and better governed (Valdés, Reference Valdés2010: 9, 32).
We examined TPRs for 24 countries under review in the period 2006–11. We created a sample group of geographically and politically diverse states; we included long-standing and relatively new members, poorer and wealthier states, as well as nations at different levels of industrialization and democratization. Our sample included long-standing members such as the US, Brazil, Sri Lanka, and Pakistan as well as developing and middle-income countries that were relatively new members (Georgia, Jordan, China, Slovenia, and the Kyrgyz Republic). We examined whether member states made the changes they promised at accession (for new members) or were held to account for their governance policies (for long-standing members such as the US).Footnote 40 Costa Rica was the only country in our sample where these issues never came up. For other countries, governance issues came up frequently.
Member states criticized countries such as the US as well as developing countries. Trade powers were not the only member states using the trade policy review process to praise or chide. Colombia, Turkey, Canada, and Chile also actively challenged the behavior of other member states.
In general, countries were asked how they made regulations transparent, how they encouraged foreign understanding of relevant policies, and how they responded to public comment.Footnote 41 Some countries such as Kyrgyzstan and Georgia were chided for unpredictable enforcement and unclear public policies.Footnote 42 Members often asked about policy transparency and evenhandedness; they less frequently mentioned the rule of law or judicial independence. Members sometimes disagreed on performance or the best strategy to discuss governance. Turkey, a member since 1951, was lauded by some countries for its openness and improved governance, although China and Japan said that Turkey was not transparent enough.Footnote 43 Not surprisingly, China received the most criticism and questioning. In its 2008 and 2010 TPR, members acknowledged that China had become more transparent and better governed, but most countries made it clear that they thought China remained complex, opaque, and inadequately governed.Footnote 44 Some countries challenged the US and the Philippines for supporting domestic industries in an opaque and uneven manner.Footnote 45 Member states directly mentioned problems of corruption, and the need to strengthen the rule of law during the trade policy reviews of Ghana, Bangladesh, Philippines, Pakistan, Thailand, the Kyrgyz Republic, Sri Lanka, Georgia, and the joint review of Tanzania, Kenya, and Uganda.Footnote 46
Members also used the trade policy review process to see if new member states were changing how they govern in non-trade-related aspects of governance (areas not explicitly covered by WTO rules). For example, the EU and the US also asked Georgia how it would provide for freedom of association, which is not addressed in the WTO. Both governments wanted to know if Georgia would ensure that workers and individuals in general would have the right to organize. The Georgian representative responded that its laws allowed individuals to organize in ‘any kind or form of association, organization, federation’. But he did not provide specific examples of individuals doing such organizing. The Georgian government was also asked by Chinese Taipei (Taiwan) how it was able to implement reform measures during a time of political upheaval. The government representative responded that ‘the main key to the success of Georgia in implementing reforms … was the strong will of Georgian population and Georgian government to overcome the old system and strengthen corruption-free market economy’.Footnote 47 The US also pressed Jordan on non-trade-related governance issues related to the role of civil society in the Jordanian polity. The US representative noted ‘Many NGOs have been highly critical of the Associations Law that was passed this summer’ possibly undermining NGOs ability to fundraise ‘and even their permission to exist’. The Jordanian representative responded that the law was amended to consider these concerns. The US also asked about the status of the Anticorruption Commission in Jordan. The Jordanian representative responded that the Anti-Corruption Commission operates with neutrality, objectivity, and independence.Footnote 48 The Canadian government asked how Jordan would ensure public input into its trade policy. The representative responded that the government has created advisory committees and ‘the Prime Ministry provides access to the public through its website’. Also, ‘two pilot projects have been launched to establish a consultation mechanism on trade and environment issues … The consultation mechanism aims at strengthening the dialogue between the government and the private sector and civil society in trade and economic policy formulation and related regulatory aspects’.Footnote 49 Clearly countries such as Jordan and Georgia wanted to show they were making headway on reforms. At the same time, countries such as the US and Canada were determined to use these reviews to challenge other member states on governance issues.
Taken in sum, the qualitative evidence indicates that members use the trade policy reviews to praise countries that have made governance progress and to name and shame countries that continue to have problems. They often discussed issues of transparency and evenhandedness and less frequently discussed due process. They chided some nations for corruption and inadequate governance; some countries even prodded other new members to encourage public participation and enforce other non-trade regulations such as labor rights. But these trade policy reviews revealed that members were not consistently, evenly, or effectively adopting WTO norms. Members don't work with a check list to monitor accession commitments or have a coordinated approach to further encourage diffusion of WTO norms with incentives such as capacity building. Hence while the trade policy review process is useful as a means of ‘outing’ bad or inadequate behavior, it cannot stop such behavior. Moreover, members may talk about de jure policy changes, but such change may not yet result in facts on the ground – where citizens can consistently obtain information, challenge trade-related policies, and monitor their government commitments. Nonetheless, our review of WTO documents showed some member states are pushing new member states to transmit WTO norms in areas of non-WTO competence – including advancing human rights or reducing corruption. These countries see the trade policy review process as a means of transmitting a broader conception of good governance.
Quantitative evidence
Research design
Quantitative model
This section delineates how we test our hypotheses and assess the influence of the three WTO norms upon country behavior. The quantitative model includes all independent countries using the Correlates of War framework (Correlates of War, 2011). Our unit of analysis is the country year. The years covered by our analysis vary depending on the availability of data. We used a variety of models to test our hypotheses about the role of WTO membership in improving country performance on evenhandedness, access to information, and due process. As noted above, we also relied on two different data sets: the broader data sets used by several other scholars and those we think are more exact.
To many scholars, these broad data sets have significant advantages of duration and scope. They cover most countries for many years. However, both the World Bank and Freedom House data sets include information and questions that go beyond the more narrow concepts of evenhandedness, access to information, or due process. Thus, we relied on other datasets from Global Integrity and the World Bank that better track these norms. However, the Global Integrity data set does not cover every country for each year. Moreover, although our metric for access to information fully describes the WTO norm, our metrics for due process and evenhandedness measure only aspects of the norm (e.g. for due process – the right to appeal elections, and for evenhandedness – contract enforcement.)
Our empirical analysis proceeds as follows. We began with a single stage approach and then, to control for the effects of endogoneity, we use an instrumental variables approach. As noted earlier, we divided countries into four groups (1) non-members, (2) the completed negotiations group (countries that completed negotiations and are eligible to join the WTO), (3) new WTO members (those joining from 1996 to 2011)Footnote 50 and (4) long-standing GATT/WTO members (countries that joined the GATT before 1995). We list the countries and the years they were members of each group in Web Appendices A–B.
We then use these models to see if members at different states of WTO membership exhibit different performance on our broad and narrow metrics of evenhandedness, non-discrimination, and access to information in comparison to other members and non-members. We are specifically interested in learning if the completed negotiations group and new members exhibit stronger performance on our metrics compared to existing WTO members and non-members, because this would show us that the accession process is leading to measureable governance improvements. We then also examine if long-standing members of the GATT/WTO (those that joined before 1995) improved performance on these metrics in comparison to all other states (including non-members and new members).
In our single stage model, we use an ordinary least squares model. We limit the effects of potential heteroskedasticity across panels often found in cross-national data through the use of robust standard errors. In order to limit the effects of any broader trends towards or away from better governance outcomes, we include a year measure in all the models and also lag our controls in all of our models by one year to limit the possibility of reverse causality driving any of our results. In our robustness tests, we utilize instrumental variables analysis to account for the possibility of an endogenous relationship between those countries that are members of the GATT/WTO and better governance outcomes.
As we noted above, several factors that affect GATT/WTO membership may also affect quality of governance in member states. For example, wealthier and more democratic countries are also likely to have been long-standing members of the GATT/WTO and tend to have better governance outcomes. We control for endogeneity to avoid any resulting biases. Thus, we adopted a treatment–effects model (StataCorp, 2011: 2257),Footnote 51 which allows one to estimate the effect of an endogenously chosen binary treatment (GATT/WTO membership) on a continuous fully observed variable (our various governance measures) which are a function of the covariates in our governance equations as well as our key endogenous variable GATT/WTO membership (described as the treatment). The first stage of the model takes an instrumental variables approach that first predicts GATT/WTO membership. We treat our measure of the membership in the GATT/WTO as our endogenous regressor. In the model specification, it is important to have one or more additional variables (excluding exogenous variables) that are correlated with our measure of membership in the GATT/WTO, but not with the governance measures (the dependent variable in the second equation) except via membership. Our instrument is good at explaining membership in the GATT/WTO, but is not directly related to, or good at explaining, governance outcomes (Simmons, Reference Simmons2009; Greenhill, Reference Greenhill2010). For our new WTO and long-standing GATT/WTO groups, total trade is an appropriate instrumental variable, because it predicts the likelihood of membership, but does not directly predict governance outcomes. The included exogenous variables form the basis of our second stage regression equation and are displayed in our results.Footnote 52 Although the treatment regression approach is the correct estimator, given the distribution of our dependent variables, we cannot directly test the strength of our instruments.
Dependent variables
As noted above, we rely on six measures of governance to examine if membership in the GATT/WTO affects country behavior. To ensure consistency in our results, we re-scaled all our good governance measures (where not already) onto a 0–100 scale to permit comparison of the relative impact of GATT/WTO accession/membership on our dependent variables of interest.
Due process
Broad metric
Many scholars have relied on metrics from Freedom House which describe expert perceptions about the extent to which government functions properly. The index ranges from 0 to 100; higher values indicate stronger perceptions (original scale 0–12). These metrics cover all countries over the 2005–2011 period. To develop these metrics, Freedom House asks expert observers several questions including: Does the freely elected head of government and national legislative representatives determine the policies of the government? Is the government free from pervasive corruption? Is the government accountable to the electorate between elections, and does it operate with openness and transparency? Hence, these metrics go beyond due process and provides users with a broad overview of governance effectiveness.
Narrow metric
Scholars and policymakers generally define due process as regulations, laws, and procedures designed to ensure that the state respects the legal rights that accrue under domestic and international law to individuals. With due process, individuals have the right to challenge official administrative and judicial decisions as well as elections.Footnote 53 Although we found it difficult to find a single data set encompassing this concept, we rely on survey data from Global Integrity, which focuses on the extent to which individuals can appeal election results. The variables range between 0 and 100 where a ‘0’ indicates that election results cannot be effectively appealed and a ‘100’ indicates that elections can be effectively appealed. The statistics come from the Global Integrity Report (2004, 2006, 2007, 2008, 2010, and 2011). Global Integrity is an NGO funded by the World Bank and international foundations, which helps individuals and NGOs measure and improve governance and anticorruption counterweights. The Global Integrity Reports use questionnaire responses from country specialists to generate their data. Unfortunately, the Global Integrity dataset covered only 25 countries in 2004, 42 countries in 2006, 50 countries in 2007, 47 countries in 2008, 30 countries in both 2009 and 2010, and 31 countries in 2011. However, the dataset does show geographic and developmental variance. The sample also provides a good deal of variation in the extent to which results can be effectively appealed across the entire 0–100 range of the measure. Thus, we think it provides a good overview of whether or not a country respects due process.
Access to information
Broad metric
This metric describes perceptions about the extent to which to which a country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media. Thus, it does not describe transparency or access to information per se. The variable ranges from 0 to 100, higher values indicate better perceptions (original scale −2.5 – +2.5). This variable covers all countries of the world over the 1996–2010 periods and comes from Kaufman et al. (Reference Kaufman, Kraay and Mastruzzi2010) ‘Governance matters’ data set.
Narrow metric
Our narrow metric allows us to hone in on the state willingness to provide information to market actors. We relied on a variable from Global Integrity which describes how much information a government provides to its citizenry. The Global Integrity Reports (2004, 2006, 2007, 2008, 2010, and 2011) use questionnaire responses to generate a 0–100 scale measure. A value of ‘0’ indicates that the public has great difficulty accessing government information. A value of ‘100’ indicates that the public can easily access all aspects of government information. However, the data set does not cover every country for every year. Global Integrity covered 26 countries in 2004, 43 countries in 2006, 52 countries in 2007, 46 in 2008, 30 countries in 2010, and 31 countries in 2011. Although the sample size was limited, the data sample revealed a wide range of variation in actual levels of access to government information. The values range from 0 to 98.33.
Evenhandedness
Broad metric
This metric describes perceptions about the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. Thus, although this metric includes aspects of evenhandedness, it is much broader than evenhandedness. The variable, which covers all countries of the world from 1996–2010, comes from the Kaufman et al. (2011) ‘Governance Matters’ data set. For this variable, higher values indicate better perceptions (original scale −2.5 – +2.5).
Narrow metric
We relied on the World Bank Doing Business Report, which describes the degree to which individual states fairly and evenly enforce contracts within their borders. For ease of interpretation, we reversed the original measure so now lower values indicate worsening government enforcement of contracts, while higher values indicates that governments have better records enforcing contracts. The variable ranges from 0 to 100, higher values indicate better enforcement of contract (original scale 1–180). The data set covered 171 countries in 2007, 174 countries in 2008, 178 countries in 2009, 180 countries in 2010, and 178 countries in 2011. However, because the data were only available for five years, we may miss reforms that occurred earlier.
Key independent variables
Whether a country is negotiating to join the WTO, whether a country is a new member of the WTO, and whether a country is a long-standing member of the GATT/WTO
All three measures are dichotomous. The first measure describes whether a country was negotiating to join the WTO. The measure takes on a value of ‘1’ if the state is negotiating to join and a ‘0’ for those countries not involved in negotiations.Footnote 54 The second measure describes whether or not a country is a new member of the WTO and takes a value of ‘1’ in every country year of a state's membership for those nations that joined the WTO in 1995 or after. It takes a value of ‘0’ otherwise. The third measure indicates long-standing GATT/WTO membership and it takes a value of ‘1’ for every country year that a state was a member of GATT and ‘0’ otherwise. We used membership data from the GATT documents library at Stanford University (GATT governed trade from 1948 to 1995) and the WTO website.Footnote 55 For a more detailed discussion of our key independent variables and the country year groupings for each category of our independent variables, we direct the reader to Web Appendices A–B. This GATT/WTO membership measure is also one of the key dependent variables in the first stage of our equations.
Other correlates of joining international organizations and good governance
We note that factors that make countries more likely to become members of the GATT/WTO may also be factors describing or related to good governance. For example, countries which have better records on evenhandedness, access to information, and due process may be more likely to join the GATT/WTO. By controlling for these endogeneity issues, we are better able to assess the effect of GATT/WTO membership on good governance (Neumayer, Reference Neumayer2005; Landman, Reference Landman2005).
Our model explaining governance outcomes builds on literature from economics, political science, and law. Scholars have shown that wealthier countries tend to be better governed (La Porta et al., Reference La Porta, Lopez-de-Silanes, Shleifer and Vishny1999) and better governed economies have less corruption (Wei, Reference Wei1999; Lambsdorff, Reference Lambsdorff2007). Some academics have demonstrated that democratic countries which trade more are also likely to respect the rights of their citizens and to act in a more even-handed manner (Poe et al., Reference Poe, Tate and Camp Keith1999; Poe, Reference Poe, Carey and Poe2004; Blanton and Blanton, Reference Blanton and Blanton2007). In contrast, countries with relatively large populations, and that experience civil wars tend to have worsened governance outcomes (Poe et al., Reference Poe, Tate and Camp Keith1999; Poe, Reference Poe, Carey and Poe2004). Olson theorized that rapid economic growth has a disruptive impact on social stability, which in turn can reduce government respect for citizens' rights (Olson, Reference Olson1965). We also control for some contextual factors such as ethno-linguistic fractionalization (especially for issues of evenhandedness) and location (LaPorta et al., Reference La Porta, Lopez-de-Silanes, Shleifer and Vishny1999). Table 2 summarizes the independent, dependent, and control variables used in the first GATT/WTO membership equation and good governance stage equations.
Operationalization of GATT/WTO negotiations and GAT WTO membership good governance equations variables
Web Appendix C reports the descriptive statistics for the variables used in our analyses. Web Appendix D also includes descriptive statistics for our governance indicators across the four groups we studied. Web Appendix E reports the pairwise correlations amongst our independent and control variables. Pairwise correlations indicate no problems of multicollinearity, with the highest correlation of 0.63 between our measure of the level of democracy and the level of GDP per capita. VIF tests reported no variables with a value greater than 2.61, well below the threshold of 10 indicating that multicollinearity is not a problem in our data. We discuss the control variables at the end of each following section. As noted above, for ease of discussion, we transformed all our good governance dependent variables onto a common 0–100 scale with higher values indicating better governance outcomes for ease of discussion.
Results
Negotiation and GATT/WTO membership on broad metrics for good governance
Our tables illuminate the relationship between membership status and governance improvements. Table 3 displays the results from our first set of OLS analyses that utilize our broad metrics of good governance. The first three rows of results indicate the association between each stage of association or membership in the WTO relative to non-members, the excluded group.Footnote 56 Our first analysis focuses on the completed negotiating group (those nations which we hypothesize made major changes to governance during the accession process, and compare these findings to those of non-members). Models 1–3 indicate that the negotiating group had poorer outcomes across two of the three metrics of governance: due process and access to information at the 0.01 level of confidence in comparison to existing WTO members and non-members. The period of negotiation to join the WTO was negatively correlated with these outcomes. Also, we found that the size of these negative effects varied, depending on the norm in question: there was a sizeable negative −6.0 unit reduction in our broad metric of access to information, but a much larger and negative −17.25 unit difference in due process. We found no significant effects when we attempted to measure changes in evenhandedness. Next we examine how new members performed on our broad metrics of governance. In these analyses, we compare states that have joined the WTO since 1995 (new members) with non-members and long-standing members. Models 1–3 indicate that new members of the WTO (those who joined since 1995) had a worse performance on all three broad metrics of governance significant at a 0.05 level of confidence or higher. On average they showed weaker performance on due process (by –10.4 units), on access to information (by −5.15 units) and on evenhandedness (by −8.58). Finally, we examined long-standing GATT/WTO members relative to non-members and new WTO members. The results from Models 1 to 3 indicate that countries which were members of the GATT/WTO did not have significantly better governance scores on any of our broad metrics of good governance.
OLS model, the negotiation and membership effects of WTO and GATT/WTO on good governance, all countries, broad indices
Notes: Z–Score in parentheses, * p<0.1; ** p<0.05; **** p<0.01. Two-tailed tests of significance.
Web Appendices G and H display the results from our instrumental variables regression, which we used to examine the robustness of our results presented in Table 3. Web Appendix G examines countries that completed negotiations and their performance on our broad metrics of governance. We found that countries which have completed negotiations are associated with weaker outcomes in the areas of due process and access to information. However, countries that completed negotiations performed better on the broad metric of evenhandedness. Web Appendix H displays the IV regression results examining the link between new WTO and long-standing GATT/WTO member performance using our broad metrics of governance. The results from Web Appendix H indicate that new WTO members perform better on our broad metric of access to information but that new WTO members do not perform significantly differently from non-members and long-standing members in the areas of due process and evenhandedness. Finally, Web Appendix H examines the association between long-standing GATT/WTO members in comparison to new WTO members and non-members. The results indicate that long-standing members performed significantly worse in the areas of due process and evenhandedness on our broad metrics of governance in comparison to non-members and new WTO members, but did not perform significantly differently from these groups on our metric of access to information.
In Web Appendices G and H, we present first stage results findings for all nine instrumental variables analysis. In Web Appendix G, Membership in IGOs was our instrument in the completed negotiations equation. It significantly increased the likelihood of successfully completing negotiations to join the WTO across all three models at the 0.05 level of confidence or higher. In Web Appendix H, openness was our instrument in the new WTO member and long-standing GATT/WTO membership equations. Greater openness was positively associated with being a new WTO member across all three of our governance models at the 0.01 level of confidence. In contrast, greater openness was negatively associated with being a long-standing GATT/WTO member across all three of our governance models at the 0.01 level of confidence.
In our instrumental variables analyses, we report not ρ, the correlation of the error terms between the two stages of the model, but the inverse hyperbolic tangent of ρ, labeled athro (STATA, 2009: 2263).Footnote 57 This measure indicates whether there is a significant correlation in the error terms of the two equations and the direction of that error. A significant value of athro means that endogenous processes affect membership in the WTO and our governance indicators and was present in all nine models we present in Web Appendices G and H, but because we used an instrumental variables approach, we were able to limit the effects of this endogoneity.
Negotiation and GATT/WTO membership on narrow metrics of good governance
Table 4 displays the results from our first set of OLS analyses that utilize our narrow metrics of good governance. The first three rows of results indicate how each group performed on our metrics of good governance directly describing WTO norms.Footnote 58 Because we had a limited number of cases available for each of our dependent variables, we included only the most important controls (based on the extant literature) in these models. Our first analysis focuses on the completed negotiating group. In Table 4, Model 1 indicates that our completed negotiating group of countries exhibited poorer performance on our metrics of due process using our narrow metric of governance, significant at the 0.1 level of confidence. Models 2 and 3 of Table 4 indicate that our completed negotiating group of countries exhibited better performance on our metrics of access to information and evenhandedness using our narrow metric of governance, significant at the 0.1 level of confidence or greater. We found that the size of positive effects varied, depending on the norm in question: there was a sizeable positive +15.82 unit increase in our narrow metrics of access to information, but a slightly smaller positive +12.19 unit increase in evenhandedness.
OLS model: the negotiation and membership effects of WTO and GATT/WTO on good governance, all countries, narrower metrics
Notes: Z-scores in parentheses * p<0.1; * p<0.05; *** p<0.01. Two-tailed tests of significance.
Next we examine the effect of new WTO member performance using our narrow metrics of governance. In comparison, Model 3 indicates that new members of the WTO performed worse on our narrow metric of evenhandedness, significant at the 0.01 level of confidence. The results also indicate that new WTO members did not perform significantly differently on our narrow metric of due process in comparison with non-members and long-standing members. Finally, the results from Models 1 and 2 indicate that countries which were long-standing members of the GATT/WTO performed significantly better on our narrow metrics of due process and access to information than new members or non-members. While in Model 3 we found that long-standing membership was associated with poorer performance on our narrow metrics of evenhandedness significant at the 0.01 level of confidence. On average they showed stronger performance on due process (by 15.76 units), access to information (by 30.26 units), and weaker performance on evenhandedness (by −19.95 units).
Web Appendices I and J displays the results from our instrumental variables regression, which we used to examine the robustness of our results presented in Table 4. The results from Web Appendix I indicate that completed negotiations are associated with better outcomes in the areas of access to information and evenhandedness significant at the 0.01 level of confidence. In Web Appendix J, we present the effects of WTO membership and GATT/WTO membership on our narrower metrics of governance using instrumental variables analysis. We were unable to achieve convergence for either of our due process models. Models 1 and 2 in Web Appendix J indicate that new WTO members did not perform better on our narrow metrics of good governance in comparison to non-members and long-standing members. Web Appendix J indicates the mixed consequences of long-standing GATT/WTO membership on their governance in comparison to non-members and new WTO members. Long-standing members perform better on our metrics of access to information than non-members or new members. Meanwhile long-standing GATT/WTO members performed worse on our narrow metric of evenhandedness in comparison to new WTO members and non-members. We find that long-standing GATT/WTO members improve their performance on our metric of access to information by 51.61 units and worsened their evenhandedness by 44.74 units. Web Appendices I and J present first stage results findings for all seven instrumental variables analyses. In Web Appendix I, membership in IGOs was our instrument in the completed negotiations equation. It significantly increased the likelihood of successfully completing negotiations to join the WTO in one of the models at the 0.05 level of confidence. In Web Appendix J, openness was our instrument in the new WTO member and long-standing GATT/WTO membership equations. Greater openness was positively associated with being a new WTO member in both of our governance models at the 0.01 level of confidence. In contrast, greater openness was negatively associated with being a long-standing GATT/WTO member again across both of our governance models at the 0.01 level of confidence. The athro was significant in seven out of our eight models. Thus, we are confident that endogenous processes affected how member states performed on our narrow governance indicators, as shown in Web Appendices I and J.
When we use our limited but important control variables, our findings replicate the work of some governance scholars who have found wealthier, more democratic countries further from the equator are more likely to have better governance indicators (see Tables 3 and 4). We found wealthier countries had better governance outcomes across all six models presented, significant at the 0.05 level of confidence or higher. We also found that democratic states showed better performance on our metrics in five of the six models presented at the 0.10 level of significance or higher. Countries further away from the equator had better governance outcomes across five of the six models presented significant at the 0.01 level of confidence or higher. Finally, one of our two models showed a significant correlation between greater levels of ethno-linguistic fractionalization and better performance on our metric of evenhandedness at the 0.01 level of confidence.
However, we did not always find governance improvements. All three models indicated a negative correlation between larger populations and governance outcomes using our broad metrics, significant at the 0.10 level of confidence or higher. We also found countries in civil wars were associated with lower performance on two of our three broad metrics of governance at the 0.01 level of confidence. Finally, when we controlled for time using a year variable, we found a secular trend towards weaker governance.
Conclusion
The WTO governs trade and doesn't explicitly address governance. Yet as noted above, the WTO Secretariat, policymakers, and some scholars argue that improved governance is a spillover of membership in the WTO. We hypothesized that if countries anchor during accession, we would see significant improvements in governance in the completed negotiations group. We also expected we would see similar improvements in the new member group, although other member states have lost the leverage embedded in the accession process. If countries anchored during membership, we should see gradual improvements in the performance on our metrics for new and long-standing members. When we examined WTO documents, we found considerable supporting evidence for our hypotheses. Before they accede, countries make major changes to their laws, regulations, and behavior related to a wide range of trade and trade-related policies, from tax and competition policies to health and safety standards. They are coached and closely monitored by a working party and the WTO Secretariat. Once countries join the WTO, these nations are monitored by other member states at trade policy reviews. During these reviews, trade officials ask questions about compliance with WTO norms in trade policymaking. These trade diplomats also ask new members if they have adopted strategies to inform and involve the public, provide administrative due process, and ensure evenhandedness. In particular, the US, Canada, and the EU questioned other member states about areas of governance that are not regulated by the WTO, including advancing human rights or reducing corruption. Policymakers from these countries see the WTO as an instrument that can help them help member states learn to create inclusive, responsible, and accountable institutions of governance.
However, our quantitative analysis revealed a more complex story. We used both broad and specific metrics to examine country performance among groups at different stages of their engagement with WTO. Using the broad metrics, we found no support that policy anchoring occurs during the accession process or through membership.
When we used the more narrow metrics, we found mixed results. The negotiating group showed stronger performance on metrics of access to information and evenhandedness, but weaker performance on due process. Although new members showed better results on access to information, they had lower levels of performance on our metric of evenhandedness. We found no significant effect of new WTO membership on our metric of due process. Finally, long-standing members had stronger performance on metrics of due process and access to information, but they had weaker performance on metrics of evenhandedness. Thus, our data analysis provides some support for policy anchoring during accession and greater support for policy anchoring as a function of membership over time. We note that our findings are preliminary because we included few controls and had a limited number of cases.
Both our broad and narrow metrics describe each country's governance in general rather than specifically the rules related to trade. Thus, we believe our findings support our argument that the norms of good governance promoted by the WTO gradually filter into the polity as a whole.
Readers understandably will ask how we explain the dissonance between our strong qualitative and mixed quantitative results. We note our data may have measurement errors. Our broad metrics may be too broad and our narrow metrics too narrow. But we also note that there is no direct path or road map to improved governance. Nations take time to anchor to the WTO and occasionally nations may drift. Policymakers may lack capacity, funds, or will to effectively implement reforms. Business and governmental elites may resist change. (Russia provides a good example of this phenomenon.) Scholars such as Rodrik, as well as Acemoglu and Robinson, have shown that social, cultural, and historical factors may impede the development of effective institutions.
We also emphasize that the WTO cannot directly compel improved governance in the trade or domestic spheres. But member states could use the trade policy review mechanism more effectively and consistently. WTO members don't seem to compare accession commitments and membership achievements – they have not developed strategies to measure change over time. Nor have they developed incentives to encourage states to meet their reform objectives. For example, the WTO Secretariat wrote of Belize's review: ‘WTO commitments could anchor future reform efforts needed to sustain growth.’Footnote 59 The WTO report did not mention any countries providing capacity building assistance to help Belize achieve such reforms. We suggest that policymakers do a better job linking accession commitments to the TPRs and provide such assistance.
Our findings suggest some new areas of research. Although our metrics closely track WTO norms, we could not find statistics that covered all WTO members for our period of review. We hope other scholars will work on more exact metrics of governance that more closely track WTO norms. In addition, we found that membership in the WTO did not yield consistent improvements for all three norms. Although it is beyond the scope of this paper to tease out why, we want to encourage scholars to examine these distinctions.
In recent years, some scholars have attempted to measure the impact of the WTO on trade (Rose, Reference Rose2002, Reference Rose2005; Goldstein et al., Reference Goldstein, Rivers and Tomz2007). They have generally concluded that the WTO is a paper tiger – fiercer on paper than in reality. Economist Andrew Rose asserts that this should not surprise us – members deliberately designed the WTO to be a weak international institution (Rose, Reference Rose2005).
But the WTO is, without direct intent, having some effects on governance. Hence, scholars, citizens, and policymakers should consider the governance spillovers of the WTO when they judge the record of the WTO over time.
