The determinants of product-specific rules of origin: An econometric analysis in the regional comprehensive economic partnership

Crivelli, Pramila; Inama, Stefano; Marand, Jeremy Working Paper The determinants of product-specific rules of origin: An econometric analysis in the regional comprehensive economic partnership ADB Economics Working Paper Series, No. 713 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Crivelli, Pramila; Inama, Stefano; Marand, Jeremy (2024) : The determinants of product-specific rules of origin: An econometric analysis in the regional comprehensive economic partnership, ADB Economics Working Paper Series, No. 713, Asian Development Bank (ADB), Manila, https://doi.org/10.22617/WPS240022-2 This Version is available at: https://hdl.handle.net/10419/298159 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/3.0/igo/ ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org ADB ECONOMICS WORKING PAPER SERIES NO. 713 January 2024 The Determinants of Product-Specific Rules of Origin An Econometric Analysis in the Regional Comprehensive Economic Partnership Rules of origin differ among overlapping free trade agreements, raising firm compliance costs, discouraging utilization of trade preferences, and hindering regional value chains. This study exploits a unique dataset comparing the restrictiveness of product-specific rules of origin (PSRO) between the Regional Comprehensive Economic Partnership (RCEP) and other free trade agreements in Asia based on manufacturing requirements. Using maximum-likelihood models, the econometric analysis shows that economic sectors, political economy determinants, and negotiating capacities significantly influence PSRO stringency under the RCEP. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 68 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. THE DETERMINANTS OF PRODUCT-SPECIFIC RULES OF ORIGIN AN ECONOMETRIC ANALYSIS IN THE REGIONAL COMPREHENSIVE ECONOMIC PARTNERSHIP Pramila Crivelli, Stefano Inama, and Jeremy Marand ASIAN DEVELOPMENT BANK The ADB Economics Working Paper Series presents research in progress to elicit comments and encourage debate on development issues in Asia and the Pacific. The views expressed are those of the authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent. ADB Economics Working Paper Series Pramila Crivelli, Stefano Inama, and Jeremy Marand No. 713 | January 2024 Pramila Crivelli ([email protected]) is an economist and Jeremy Marand is a consulant (mjeremy.consultant@ adb.org) at the Economic Research and Development Impact Department, Asian Development Bank. Stefano Inama (stef[email protected]g) is the chief of the Technical Assistance, Trade and Customs in the Division on African and Least Developed Countries, United Nations Conference on Trade and Development. The Determinants of Product-Specific Rules of Origin: An Econometric Analysis in the Regional Comprehensive Economic Partnership Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) © 2024 Asian Development Bank 6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, Philippines Tel +63 2 8632 4444; Fax +63 2 8636 2444 www.adb.org Some rights reserved. Published in 2024. ISSN 2313-6537 (print), 2313-6545 (electronic) Publication Stock No. WPS240022-2 DOI: http://dx.doi.org/10.22617/WPS240022-2 The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies ofthe Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned. By making any designation of or reference to a particular territory or geographic area, or by using the term “country” inthis publication, ADB does not intend to make any judgments as to the legal or other status of any territory or area. This publication is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound bytheterms of this license. For attribution, translations, adaptations, and permissions, please read the provisions andterms of use at https://www.adb.org/terms-use#openaccess. This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed toanother source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held liable for any claims that arise as a result of your use of the material. Please contact [email protected] if you have questions or comments with respect to content, or if you wish toobtain copyright permission for your intended use that does not fall within these terms, or for permission to use theADB logo. Corrigenda to ADB publications may be found at http://www.adb.org/publications/corrigenda. ABSTRACT Rules of origin differ among overlapping free trade agreements, raising firm compliance costs, discouraging utilization of trade preferences, and hindering regional value chains. Using a unique dataset comparing the restrictiveness of product-specific rules of origin (PSRO) between the Regional Comprehensive Economic Partnership (RCEP) and other free trade agreements in Asia based on manufacturing requirements, this study aims to identify the factors explaining the stringency of RCEP PSROs. The econometric analysis based on maximum-likelihood models shows that economic sectors and political economy determinants and negotiating capacities significantly influence PSRO stringency under the RCEP. In particular, restrictiveness scores we develop show that (i) products for which developing RCEP members exhibit a strong revealed comparative advantage face stricter PSROs, and (ii) the degree of sophistication of the production process is positively related to PSRO leniency. This casts doubt on the potential for RCEP to provide a viable solution to the existing “noodle bowl” of rules of origin in Asia and reduce firms’ costs of compliance in developing economies. It also demonstrates the need to strengthen aid for trade in policy and negotiations in developing Asia. Keywords: international trade agreements, product-specific rules of origin, market access, RCEP, regional integration, political economy JEL codes: F13, F15 I. Introduction In a globalized world dominated by international value chains, preferential rules of origin have become one of most impactful trade policy tools to foster regional integration. 1 Defined by the World Trade Organization (WTO) as “the criteria needed to determine the national source of a product”,2 in free trade agreements (FTAs), they formally differentiate products originating in a member country from those originating in nonmember economies, thus determining the eligibility of a product for preferential tariff treatment. While not a recent phenomenon, global value chains have rapidly emerged in recent decades because of technological innovations in transportation and communication, reducing costs and enabling countries to specialize in the production of components.3 The fragmentation of production processes across countries has therefore substantially grown and inputs are sourced from suppliers around the world. In 2022, intermediate goods accounted for approximately half of total trade.4 With widespread internationalization of production, determining the origin of a good is increasingly challenging. The vast majority of non-primary products traded in international markets rarely originate from a single country and the involvement of multinational firms further complicates determination of origin (Gourevitch, Bohn, and McKendrick 2000 in Harilal and Beena 2005). In this context, rules of origin determination is critical, especially when combined with the flourishing of free trade areas. Rules of origin are an essential instrument to limit the cost of potential trade deflection while ensuring a sufficient degree of trade creation and ensure a positive welfare impact.5 They ensure that goods traded between member countries and benefitting from the preferential treatment are not simply assembled in a member country by using components exclusively originating from third countries. In the absence of rules of origin, products from third countries could be imported by the country exhibiting the lowest tariff to enter the FTA and benefit from preferential treatment when reexported within the trade bloc (Felbermayr, Teti, and Yalcin). At the same time, it is essential to take into account the costs such rules impose on firms ensuring that these do not act as a deterrent to their utilization of FTA. Understanding the industrial context, needs of companies, availability of inputs, and available technologies is essential in designing rules of origin. The tradeoff between the benefits of lenient rules of origin allowing firms—particularly medium and small enterprises—to reap the benefits of an FTA and the potential cost of trade deflection 1 Preferential rules of origin are used in reciprocal or non-reciprocal agreements. Non-preferential rules of origin determine the origin of the good independently of any preferential treatment at the border and are mostly used to apply WTO agreements (ex. anti-dumping or countervailing duties). 2 WTO. Technical Information on Rules of Origin. https://www.wto.org/english/tratop_e/roi_e/roi_info_e.htm. 3 WTO. 2014. World Trade Report 2014. Chapter C. The Rise of Global Value Chains. https://www.wto.org/english/res_e/booksp_e/wtr14-2c_e.pdf. 4 WTO. 2023. Exports of Intermediate Goods Post Sustained Growth in Second Quarter of 2022. https://www.wto.org/english/news_e/news23_e/stat_01feb23_e.htm#:~:text=The%20share%20of%20IGs%20in,activity%20in%20g lobal%20supply%20chains. 5 Trade deflection refers to cases where products from nonmember exporters enter the free trade bloc via the member with the lowest tariff, and subsequently move duty-free to higher-tariff members. 2 has long been discussed in the trade policy sphere. While such economic objectives could be assumed to be at the heart of trade negotiations, this paper shows that political considerations with disproportionate influence from large economies may be the major factor at play. Understanding these dynamics is essential to assess whether an FTA will deliver its expected benefits and to orient future (re-)negotiations towards more effective trade deals, such as in the case the Regional Comprehensive Economic Partnership (RCEP). The creation of a common rules of origin platform among 15 economies in Asia and the Pacific is often perceived as one of the most significant trade-liberalizing benefits offered by the RCEP. The wide geographical coverage of the agreement and its cumulation provision is expected to foster regional value chains through stronger incentives to source inputs within the region, and lower costs of compliance and heterogeneity for exporting companies. 6 However, RCEP rules of origin only provide an alternative to the existing “noodle bowl” of overlapping rules of origin in the region, but do not supersede rules applicable in pre-existing agreements. RCEP rules of origin thus may well turn into an additional layer of complexity depending on how they have been designed and negotiated. While the RCEP does provide an opportunity to simplify the complex mosaic of overlapping rules of origin, the market access benefits and resulting wider economic gains of the agreement can only be assessed in the industrial context of the region based on the stringency of the productspecific rules of origin (PSRO) and tariff concessions. Recent research (Crivelli, Inama, and Pearson 2022, 2023) has shown that the PSROs contained in the RCEP are neither more liberal than those of the ASEAN Trade in Goods Agreement (ATIGA), nor than those of the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) and selected ASEAN+1 FTAs. This already suggests the possible use of rules of origin as a protectionist instrument by some RCEP members with significant bargaining power, compensating or offsetting the benefits of tariff reduction. Negotiations to determine RCEP PSROs for the 5,205 tariff lines were notably slow and difficult, according to Rillo, Robeniol, and Buban (2022). Prior to the 6th RCEP Intersessional Trade Negotiating Committee Meeting and Related Meetings of 24 August 2019, 86.72% of total tariff lines (or 4,514 subheadings) were agreed on by the TF-PSR (PSRO Task Force)7, the last body under the committee to complete its task (Rillo, Robeniol, and Buban 2022). These difficult negotiations, mirroring those of tariff concessions that tediously lasted for 8 years,8 can be seen as the result of diverging interests among member countries. Studies on PSRO stringency such as Portugal-Perez (2011) in the case of North American Free Trade Agreement (NAFTA) show that developed countries have significant bargaining power during negotiations and can craft rules of origin according to their interests and at the expense of developing partners. Did similar dynamics apply to RCEP? 6 RCEP Article 3.4 provides for diagonal cumulation of originating materials with the possibility to design a full cumulation scheme in the future. 7 RCEP TNC Chair’s Guidance to Working Groups and Sub-Working Groups, 24 August 2019, the ASEAN Secretariat, Jakarta. 8 Rillo, Robeniol, and Buban (2022) note: “One key challenge was the lack of readiness of [countries] to exercise flexibility. Many, if not all [countries], took hard-line positions, making it difficult to reach consensus.” 3 Using a unique dataset codifying the PSRO stringency based on their industrial process requirements at the 6-digit level of the Harmonized Commodity Description and Coding System (HS 6), this study aims to identify the factors explaining the relative restrictiveness of RCEP PSROs when compared with ATIGA, CPTPP, and ASEAN+1 trade agreements, focusing on political economy determinants. The innovative codification of the PSRO restrictiveness and methodology developed in this paper represents a major contribution to the literature. For the first time, PSRO restrictiveness has been tailored to the subtance of the PSRO in a given sector rather than being based on an abstract predetermined coding related essentially to the form of the rules.9 To identify potential differences in bargaining power between RCEP countries depending on their level of economic advancement,10 we separate RCEP members into developing countries (Cambodia, Indonesia, the Lao People’s Democratic Republic, Malaysia, Myanmar,11 the Philippines, Thailand, and Viet Nam) and developed countries (Australia, Brunei Darussalam, Japan, the Republic of Korea, New Zealand, and Singapore). Due to its economic size, the People’s Republic of China is analyzed separately. First, we compare RCEP PSROs with those of six other agreements one by one using a probit model. In a second stage, we build restrictiveness scores at the product level to investigate the factors influencing the relative PSRO stringency employing ordered logit models. The empirical specifications take into account sectors and variables reflecting potential trade deflection as well as revealed comparative advantages. We find that political determinants and negotiating capacities significantly influence the stringency of PSROs under RCEP. The restrictiveness scores we develop show that products for which developing countries have a strong revealed advantage face stricter PSROs. Comparing the RCEP with other agreements individually shows that the People’s Republic of China’s (PRC) revealed comparative advantages favorably influence PSRO stringency. Both analyses reveal that the degree of knowledge sophistication needed to produce a product is negatively associated with PSRO stringency. Such a finding is favorable to the few RCEP developed members where these highly sophisticated goods are produced. Overall, the higher the value of exports from developing to developed RCEP partners for a particular product, the stricter its associated PSRO. In contrast, PSROs tend to be lenient for highly traded products from developed to developing members.12 Section II provides a brief overview of the major contribution to the literature on the the cost of compliance of rules of origin and their determinants. Section III reviews the indexes commonly 9 The substance of a PSRO refers to the requirement of working or processing for a product to achieve substantial transformation, and therefore acquire originating status and eligibility for preferential treatment at the time of exportation. The form of the PSRO refers to its drafting technique, independently of its content or stringency. 10 Based on World Bank income groups. Developed countries correspond to high-income countries while developing countries are all other countries. See https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bankcountry-and-lending-groups for more information. 11 Effective 1 February 2021, ADB placed a temporary hold on sovereign project disbursements and new contracts in Myanmar. 12 RCEP developed members are Australia, Japan, the Republic of Korea, Brunei Darussalam, New Zealand, and Singapore. Developing members are Cambodia, Indonesia, the Philippines, Malaysia, Myanmar, Thailand, and Viet Nam. 4 used in the literature to quantify PSRO restrictiveness and highlights their shortcomings. Section IV introduces our new method to build measures based on relative stringency. Section V presents the variables and econometric frameworks employed to examine the determinants of RCEP PSRO restrictiveness using our new measurements. Estimation results obtained when comparing RCEP with the six other agreements using a probit and fractional logit models are presented in section V. Section VI concludes. II. Literature Rules of origin have been characterized in the economic literature as carrying a series of disadvantages. In addition to compliance with the PSROs in industrial process, rules of origin also imply substantial costs in red tape (Cadot et al. 2006 and Carrère and de Melo 2006), such as documentation and certification procedures. For example, in the case of NAFTA, Anson et al. (2005) find that administrative costs account for almost half of the preference margin. Unsurprisingly rules of origin are used as protectionist tools. Several authors in the 1990s and 2000s, such as Harilal and Beena (2005), note a growing tendency to use them as nontariff barriers to trade, especially for producers of intermediates (Hoekman 1993; James 1997, in Harilal and Beena 2005), as traditional protectionist policies are increasingly constrained (Destler 2006). More stringent rules of origin on local content to protect regional producers of intermediate goods affect companies producing final goods by preventing them from choosing the most efficient supplier around the world (Conconi et al. 2018). As a consequence, regional producers have to rely on less efficient internal input sources, raising production costs. This induces these producers to also request protection, leading to “cascading protection along the production chain” (Harilal and Beena 2005, Hoekman 1993). The stringency of PSROs is influenced by trading interests, and lobbies defending the interests of specific product groups in each member country may try to influence negotiation positions. In addition, the stringency of rules of origin can result from differing bargaining power between member countries (Portugal-Perez 2011). When trade agreements involve member countries with varying levels of development, more economically developed countries may have considerably more influence on the stringency of rules of origin. Authors such as Anson et al. (2005) argue that “southern partners are left on their participation constraint” when they are involved in vertical trade agreements, as they are eventually provided with little market access. Moreover, rules of origin can be used as an instrument to create captive markets, according to the trade suppression mechanism described by Rodriguez (2001). Strict rules of origin are imposed by intermediate goods producers to force downstream companies located in other parties to buy their products instead of inputs produced by more efficient and cheaper competitors located in third countries. Since developed member countries are more likely to produce capitalintensive intermediate goods and firms located in developing member countries are more likely to only assemble final goods, northern countries tend to benefit from vertical trade agreements. 11 AANZA, AFCTA, AJCEP, and AKFTA using the HS 2012 codes at the six-digit level following a line-by-line comparison of around 25,000 observations. Based on this data, we compute four relative measures: two stringency scores (score 1 and score 2) based on the RCEP and the 6 comparative agreements (ATIGA, CPTPP, ASEAN+1) combining the two datasets described above, and two leniency ranking measures (rank 1 and rank 2) using the two individual datasets separately (ATIGA, CPTPP; ASEAN+1): Pooled analysis – Stringency scores • Average stringency (score 1): This score is calculated as follows. For each HS6 line and each RCEP comparative agreement, we first assign the value of 0 if the PSRO of the comparative agreement is stricter than RCEP, 0.5 if they have equal stringency, and 1 if the RCEP’s PSROs are more stringent. For each HS6 line, we therefore have six values—one for each agreement the RCEP is compared with. The average stringency score corresponds to the simple average of the stringency value across the six agreements. The score ranges between zero and one. A higher value indicates more stringency. • Count stringency (score 2): This score between 0 and 6 counts the number of agreements having more lenient PSROs than RCEP at the HS6 digit level. For example, if for a particular tariff line, a value of 4 indicates that 4 agreements have more lenient PSROs than RCEP, and two are equally or more stringent. Higher values indicate more stringency. Individual analysis—Leniency ranks • Rank leniency (rank 1 and rank 2): The two scores apply the same ranking system as in Crivelli, Inama, and Pearson 2023, and 2022, respectively, based on RCEP PSRO leniency. If the RCEP has the most lenient PSRO for a particular tariff line among all agreements used in the study, the rank takes the value 1. If the RCEP has the second most lenient PSRO, the rank takes the value 2 and so on. The first stringency rank (rank 1) compares the RCEP with the ASEAN+1 agreements. The second (rank 2) compares the RCEP with the CPTPP and ATIGA agreements. Higher values indicate more stringency. Values for rank 1 and 2 range between 1 and 5 and between 1 and 3, respectively. 12 V. Data and Econometric Model To identify the determinants of RCEP PSRO stringency, the following specification is adopted: 𝑌𝑌𝑖𝑖=β0+�β𝑗𝑗 𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝑗𝑗 16 𝑗𝑗=1 +𝛿𝛿0𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐷𝐷𝐷𝐷𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝑖𝑖+𝛿𝛿1𝐼𝐼𝐶𝐶𝐶𝐶𝐼𝐼𝐼𝐼𝐷𝐷𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖+𝛿𝛿2𝑃𝑃𝐶𝐶𝐼𝐼𝑖𝑖+𝛿𝛿3𝑀𝑀𝐼𝐼𝐼𝐼𝐶𝐶𝑀𝑀𝐼𝐼𝐶𝐶𝐼𝐼𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖𝑎𝑎𝑎𝑎 +𝛿𝛿4𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑎𝑎_𝑎𝑎𝑑𝑑𝑑𝑑 +𝛿𝛿5𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑_𝑎𝑎𝑎𝑎 + 𝛿𝛿6𝑅𝑅𝐶𝐶𝑅𝑅𝑖𝑖 𝑎𝑎𝑎𝑎 +𝛿𝛿7𝑅𝑅𝐶𝐶𝑅𝑅𝑖𝑖 𝑎𝑎𝑑𝑑𝑑𝑑 + 𝛿𝛿8𝑋𝑋_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑 + 𝛿𝛿9𝑋𝑋_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖𝑎𝑎𝑎𝑎 +𝛿𝛿10𝑋𝑋_𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑎𝑎𝑎𝑎 +𝛿𝛿11𝑋𝑋_𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑 +𝛿𝛿12𝑇𝑇𝐼𝐼𝐼𝐼𝐶𝐶𝑇𝑇𝑇𝑇_𝐼𝐼𝐼𝐼𝐼𝐼𝐷𝐷𝑟𝑟𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝑖𝑖 + 𝛿𝛿13𝑃𝑃𝑅𝑅𝐶𝐶_𝑀𝑀𝐼𝐼𝐶𝐶𝐼𝐼𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖+ 𝛿𝛿14𝑋𝑋_𝑃𝑃𝑅𝑅𝐶𝐶_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖+𝛿𝛿15𝑀𝑀_𝑃𝑃𝑅𝑅𝐶𝐶_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖+ 𝛿𝛿16𝑅𝑅𝐶𝐶𝑅𝑅_𝑃𝑃𝑅𝑅𝐶𝐶𝑖𝑖+𝜀𝜀𝑖𝑖 Where Yi reflects the relative stringency of the RCEP rules of origin for product i, defined at the HS-6 digit level, captured by 10 different measures divided into two categories: i. six pairwise indicators comparing RCEP with each agreement k separately (individual regressions for each agreement), and; ii. four stringency measures (two scores, two ranks) as defined in section III. Under (i) Yi takes the value 1 if the PSRO for product i is more (less) stringent under RCEP than under agreement k (ATIGA, CPTPP or any of the ASEAN+1 agreements), and 0 if the PSRO is equally or less (more) stringent under RCEP than under the other agreement. This specification is estimated using a probit model. Under (ii), the model is estimated using an ordered logit approach and the dependent variable (Yi) takes the values of the different stringency scores/ranks detailed in section III. In the case of score 1, we use a fractional response approach as the variable is continuous and ranges between 0 to 1. Our independent variables include a wide range of potential PSRO determinants focusing on both, standard rules of origin economic objectives of maximizing trade creation while minimizing trade deflection, and on political economy factors. The latter aim to capture bargaining power differences between RCEP developed and developing members 21 and the potential use of PSROs as a protectionist tool. In line with the literature,22 the PSRO determinants related to political economy considerations and potential trade deflection are based on the following variables. 21 To repeat, the RCEP developed members are Australia, Japan, the Republic of Korea, Brunei Darussalam, New Zealand, and Singapore. Developing members are Cambodia, Indonesia, the Philippines, Malaysia, Myanmar, Thailand, and Viet Nam. The People’s Republic of China is singled out from the developing group due to the size of its economy and incorporated separately in the regressions. The results are robust to the exclusion of Singapore and Brunei Darussalam, the two developed ASEAN economies. 22 For example, see the empirical framework in Portugal-Perez (2011). 13 The mean base rate 23 under RCEP for economically advanced countries is denoted by 𝑀𝑀𝐼𝐼𝐼𝐼𝐶𝐶𝑀𝑀𝐼𝐼𝐶𝐶𝐼𝐼𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖𝑎𝑎𝑎𝑎. We expect the coefficient of this variable (𝛿𝛿3) to be positive and statistically significant if developed countries are using PSROs as an instrument to protect industries that received higher protection through high tariff rates before RCEP came into force. This would corroborate the results obtained by Cadot et al. (2006) in the case of NAFTA and the EU’s PANEURO system. The propensity of trade deflection through developing (or developed) RCEP parties is proxied by two variables, 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑎𝑎_𝑎𝑎𝑑𝑑𝑑𝑑 (or 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑎𝑎_𝑎𝑎𝑑𝑑𝑑𝑑), capturing the difference in base rates between the two country groups. More specifically, the trade deflection variables are computed as follows: 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑎𝑎_𝑎𝑎𝑑𝑑𝑑𝑑 =𝑀𝑀𝐼𝐼𝑀𝑀�0; 𝐵𝐵𝐼𝐼𝐶𝐶𝐼𝐼 𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖𝑎𝑎𝑎𝑎𝑑𝑑𝑎𝑎𝑎𝑎𝑎𝑎𝑑𝑑𝑎𝑎 −𝐵𝐵𝐼𝐼𝐶𝐶𝐼𝐼 𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑖𝑖𝑎𝑎𝑑𝑑� 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑_𝑎𝑎𝑎𝑎 =𝑀𝑀𝐼𝐼𝑀𝑀�0; 𝐵𝐵𝐼𝐼𝐶𝐶𝐼𝐼 𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑖𝑖𝑎𝑎𝑑𝑑 −𝐵𝐵𝐼𝐼𝐶𝐶𝐼𝐼 𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖𝑎𝑎𝑎𝑎𝑑𝑑𝑎𝑎𝑎𝑎𝑎𝑎𝑑𝑑𝑎𝑎� 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑎𝑎_𝑎𝑎𝑑𝑑𝑑𝑑 takes the value of the base rate of developed (or advanced) economies minus the value of the base rate of developing countries whenever the difference between the two is positive, and 0 otherwise. 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑_𝑎𝑎𝑎𝑎 takes the value of the base rate of developing countries minus the value of the base rate of developed countries if positive, and 0 otherwise. As described in the introduction, the main objective of rules of origin is to prevent trade deflection. If negotiations followed an economic rationale, we expect the coefficients (𝛿𝛿4 and 𝛿𝛿5) to be positive and significant. A positive coefficient on 𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑎𝑎_𝑎𝑎𝑑𝑑𝑑𝑑(𝑇𝑇𝐷𝐷𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑_𝑎𝑎𝑎𝑎) indicates that stricter PSROs are implemented in response to greater incentives of trade deflection through RCEP developing (advanced) members. (𝑇𝑇𝐼𝐼𝐼𝐼𝐶𝐶𝑇𝑇𝑇𝑇_𝐼𝐼𝐼𝐼𝐼𝐼𝐷𝐷𝑟𝑟𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝑖𝑖) denotes the difference between the baserate and the new tariff in the first year RCEP came into force. Larger tariffs reduction in year 1 could potentially induce stricter PSROs as member countries may try to deter any trade deflection dynamic that could be induced by RCEP lower tariffs. In this case, we would expect 𝛿𝛿12 to be statistically significant and positive. In a way similar to Portugal-Perez (2011), we introduce two variables to proxy the presence of a revealed comparative advantage 24 (RCA). 𝑅𝑅𝐶𝐶𝑅𝑅𝑖𝑖 𝑎𝑎𝑑𝑑𝑑𝑑 and 𝑅𝑅𝐶𝐶𝑅𝑅𝑖𝑖 𝑎𝑎𝑎𝑎 are the revealed comparative indexes for RCEP developing and developed countries, respectively, calculated over 2016‒2019. RCA variables are computed as the ratio between the share of exports of a specific good in the total exports of an RCEP developing (developed) economy, and the world’s exports share of this good in global exports.25 If RCEP developing (developed) countries have a revealed comparative advantage, the RCA index is greater than 1. A negative coefficient on the RCA index variable indicates that PSROs are more lenient for sectors in which a comparative advantage is observed. A significant difference in RCA coefficients between RCEP developed and developing economies 23 All tariff rates used in this paper (base rate and tariff phasing down) are extracted from the schedule of tariff commitments submitted by RCEP members. 24 Revealed comparative advantages are computed as the ratio between the share of exports of a specific good in the total exports of an economy, and the world’s exports share of this good in global exports. The average RCA for 2016‒ 2019 is used in the estimations. 25 All the export and import data used in this paper, directly as independent variable or serving to compute other indicators such as the RCA, are yearly data obtained from UN Comtrade, averaged over 2016‒2019, the period over which the vast majority of RCEP PSROs were negotiated. 14 suggests an asymmetric bargaining power between parties. For example, a negative coefficient on the RCA variable of developed countries associated with a positive or insignificant coefficient in developing members indicate that the PSROs favor the interests of exporters located in RCEP developed members over those of their counterparts located in developing RCEP economies. Similarly, an insignificant coefficient on the RCA variable of developed countries associated with a positive RCA coefficient in developing members indicates that the latter are facing more stringent PSRO in sectors where they exhibit comparative advantage, while the impact is insignificant in developed economies. PCIi denotes the product complexity index (PCI). We calculate the average value of the index over 2016‒2019 to build this variable. The product complexity index is developed by Harvard Kennedy School’s Growth Lab to identify asymmetries between RCEP developed and developing members. A higher PCI value indicates a higher degree of sophistication and diversity of the productive know-how needed to manufacture the specific product. Typically, the most sophisticated products can only be manufactured in a few countries. Higher values taken by the index reflecting a higher degree of sophistication associated with more lenient PSROs, may suggest that PSROs were negotiated in favor of developed RCEP members. In this case, we expect 𝛿𝛿2 to be statistically significant and have a negative sign. 𝑋𝑋_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑 and 𝑋𝑋_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖𝑎𝑎𝑎𝑎 respectively denote the value of exports from developing RCEP members to economically advanced members and the value of exports from developed to developing RCEP members. Industries in economically advanced RCEP countries may try to impose stricter PSROs on products where exports from RCEP developing economies may considerably rise with RCEP implementation. To capture the trade dynamics, we include the total value of exports from developing to advanced RCEP countries and the total value of imports from advanced to RCEP developing countries. Industries in developed RCEP countries may try to impose stricter PSROs on products where exports from RCEP developing economies are expected to considerably rise with RCEP implementation. To assess this potential use of rules of origin as a protectionist tool, we introduce the variables 𝑋𝑋_𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑and 𝑋𝑋_𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑎𝑎𝑎𝑎 to proxy the potential trade rerouting from the rest of the world once the agreement is in place. 𝑋𝑋_𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑎𝑎𝑑𝑑𝑑𝑑and 𝑋𝑋_𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑎𝑎𝑎𝑎 respectively denote the value of exports from developing RCEP members to the rest of the world, and the value of exports from developed RCEP members to the rest of the world. A higher value corresponds to a higher threat of rerouting, as argued by Portugal-Perez (2011). We expect 𝛿𝛿10 (or 𝛿𝛿11) to be positive if importcompeting industries in developed countries (or developing countries) influence the adoption of stricter PSROs for goods highly exported to the rest of the world and that could be easily rerouted within the region as a result of the agreement’s implementation. Four different variables are included specifically for the PRC, namely the PRC base rate (𝑃𝑃𝑅𝑅𝐶𝐶_𝑀𝑀𝐼𝐼𝐶𝐶𝐼𝐼𝐼𝐼𝐼𝐼𝐶𝐶𝐼𝐼𝑖𝑖), the PRC’s exports to RCEP members (𝑋𝑋_𝑃𝑃𝑅𝑅𝐶𝐶_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖), the PRC’s imports from RCEP countries ( 𝑀𝑀_𝑃𝑃𝑅𝑅𝐶𝐶_𝑅𝑅𝐶𝐶𝑅𝑅𝑃𝑃𝑖𝑖 ), and the PRC’s revealed comparative advantage variable (𝑅𝑅𝐶𝐶𝑅𝑅_𝑃𝑃𝑅𝑅𝐶𝐶𝑖𝑖). The variables are built by using the same sources and methods as the corresponding 15 variables for the developing and developed country groups and can be interpreted in the same way. As additional control variables, two dummy variables (Consumptioni and Intermediatei) are introduced to reflect the position of the good in the value chain. The classification by Broad Economic Categories is used for this purpose.26 Consumptioni takes the value 1 if the good is a consumption product and 0 otherwise; Intermediatei is equal to 1 if the product can be classified as an intermediate product and 0 otherwise. We expect a positive sign on consumption goods that are more likely to incorporate foreign inputs could suggest. The model also includes dummies sector fixed effects (𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝑗𝑗) for each product group as defined by the World Integrated Trade Solution to control for sectoral dynamics and manufacturing process specificities that could explain PSRO stringency. Table 2 provides descriptive statistics of the variables used in this study. Table 2: Descriptive Statistics Variable Obs Mean Std. Dev. Min Max Dummy=1 if RCEP more stringent than AJCEP, 0 otherwise 5205 .052 .222 0 1 Dummy=1 if RCEP more stringent than AANZFTA, 0 otherwise 5205 .342 .475 0 1 Dummy=1 if RCEP more stringent than ACFTA, 0 otherwise 5205 .173 .378 0 1 Dummy=1 if RCEP more stringent than AKFTA, 0 otherwise 5205 .241 .428 0 1 Dummy=1 if RCEP more stringent than ATIGA, 0 otherwise 5205 .409 .492 0 1 Dummy=1 if RCEP more stringent than CPTPP, 0 otherwise 5205 .318 .466 0 1 Dummy=1 if RCEP less stringent than AJCEP, 0 otherwise 5205 .398 .49 0 1 Dummy=1 if RCEP less stringent than AANZFTA, 0 otherwise 5205 .074 .262 0 1 Dummy=1 if RCEP less stringent than ACFTA, 0 otherwise 5205 .292 .455 0 1 Dummy=1 if RCEP less stringent than AKFTA, 0 otherwise 5205 .181 .385 0 1 Dummy=1 if RCEP less stringent than ATIGA, 0 otherwise 5205 .119 .324 0 1 Dummy=1 if RCEP less stringent than CPTPP, 0 otherwise 5205 .314 .464 0 1 Score 1 5205 .261 .245 0 1 Score 2 5118 1.549 1.471 0 6 Rank 1 5124 1.544 .706 1 5 Rank 2 5199 1.685 .518 1 3 Animal 5205 .065 .246 0 1 Chemicals 5205 .151 .358 0 1 Clothing 5205 .052 .221 0 1 Food products 5205 .041 .197 0 1 Footwear 5205 .009 .095 0 1 Fuels 5205 .008 .091 0 1 Hides and skins 5205 .013 .114 0 1 Machinery and electrics 5205 .148 .355 0 1 Metals 5205 .108 .311 0 1 Minerals 5205 .02 .141 0 1 Miscellaneous 5205 .068 .252 0 1 Plastic and rubber 5205 .041 .197 0 1 Stone and glass 5205 .037 .19 0 1 Textiles 5205 .101 .302 0 1 Transportation 5205 .025 .156 0 1 26 United Nations Statistics Division. Classification by Broad Economic Categories (BEC) Revision 4. https://unstats.un.org/unsd/classifications/Family/Detail/10. Continued on the next page 16 Variable Obs Mean Std. Dev. Min Max Vegetable 5205 .068 .251 0 1 Wood 5205 .045 .208 0 1 Consumption 5205 .227 .419 0 1 Intermediate 5205 .582 .493 0 1 Product complexity index 5205 .006 1.016 -3.004 2.471 Baserate, developed countries 5205 3.577 8.058 0 160.06 Propensity to trade deflection to developed RCEP countries (TD ad_dev ) 5205 .674 7.222 0 153.435 Propensity to trade deflection to developing RCEP countries (TD dev_ad ) 5205 5.133 4.446 0 31.667 Revealed comparative advantage developed countries 5177 .907 2.518 0 85.558 Revealed comparative advantage developing countries 5142 1.383 3.725 0 104.3 Exports RCEP from developing to developed 5205 .04 .275 0 9.996 Imports RCEP to developing from developed 5205 .048 .398 0 14.781 Exports to ROW from developed 5205 .348 2.508 0 81.837 Exports to ROW from developing 5205 .626 3.525 0 160.217 Tariff reduction in year 1 5205 2.811 3.395 0 59.281 Base rate PRC for RCEP 5205 9.936 7.366 0 65 Exports from PRC to RCEP 5205 .114 .615 0 23.915 PRC imports from RCEP 5205 .132 1.419 0 56.428 PRC revealed comparative advantage 5205 1.207 1.368 0 6.786 PRC = People’s Republic of China, ROW = rest of world. Trade agreements: AJCEP = ASEAN-Japan Comprehensive Economic Partnership, AANZFTA = ASEAN-AustraliaNew Zealand FTA, ACFTA = ASEAN -People’s Republic of China FTA, AKFTA = ASEAN-Republic of Korea FTA, ATIGA = ASEAN Trade in Goods Agreement, CPTPP = Comprehensive and Progressive Agreement for Trans - Pacific Partnership, RCEP = Regional Comprehensive Economic Partnership. Source: Authors. 17 VI. Results Pairwise comparisons Tables 3 and 4 report estimation results for pairwise comparisons between RCEP and the six other agreements. Table 3: Estimation Results for Pairwise Comparisons – Restricted Model (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) AJCEP AANZFTA ACFTA AKFTA ATIGA CPTPP VARIABLES RCEP - Stricter RCEPMore lenient RCEP - Stricter RCEP - More lenient RCEP - Stricter RCEP - More lenient RCEPStricter RCEP - More lenient RCEP - Stricter RCEP - More lenient RCEP - Stricter RCEP - More lenient Consumption -0.175 -0.267*** 0.156* -0.132 0.354*** -0.860*** 0.012 -0.754*** 0.067 -0.523*** 0.264*** -0.159* (0.137) (0.087) (0.093) (0.156) (0.125) (0.087) (0.127) (0.088) (0.092) (0.143) (0.087) (0.088) Intermediate -0.219** -0.394*** 0.130* 0.004 -0.270** -0.223*** 0.189** -0.666*** -0.159** -0.029 0.502*** -0.277*** (0.095) (0.066) (0.077) (0.083) (0.106) (0.064) (0.093) (0.065) (0.074) (0.076) (0.068) (0.068) Base rate, developed countries 0.005* 0.004** -0.001 0.003 0.000 -0.002 0.005** -0.003 -0.002 0.020 0.003 0.004* (0.003) (0.002) (0.002) (0.005) (0.003) (0.003) (0.003) (0.003) (0.002) (0.017) (0.003) (0.002) Exports RCEP from developing to developed 0.154* 0.052 0.471*** 0.128 0.261** -0.971*** 0.220* 0.055 0.198* 0.110 0.078 -0.179 (0.092) (0.082) (0.161) (0.147) (0.106) (0.312) (0.121) (0.105) (0.102) (0.143) (0.090) (0.171) Imports RCEP to developing from developed -0.248*** 0.020 -0.448** -0.141 -0.160** -0.191 0.004 -0.206 -0.246*** -0.058 -0.083 0.005 (0.074) (0.051) (0.213) (0.124) (0.079) (0.186) (0.079) (0.130) (0.089) (0.092) (0.080) (0.081) Constant -1.603*** -0.433*** -1.363*** -2.564*** -1.013*** -0.283** -1.397*** -0.240* -0.514*** -0.706*** -0.056 -0.411*** (0.176) (0.114) (0.134) (0.350) (0.142) (0.122) (0.138) (0.125) (0.113) (0.120) (0.112) (0.113) R2 0.334 0.223 0.303 0.174 0.473 0.132 0.550 0.158 0.439 0.117 0.206 0.178 Observations 4,371 4,889 5,158 4,057 5,028 4,596 5,158 4,623 5,205 4,021 4,936 4,893 Sector dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes AJCEP = ASEAN-Japan Comprehensive Economic Partnership, AANZFTA = ASEAN-Australia-New Zealand FTA, ACFTA = ASEAN-People’s Republic of China FTA, AKFTA = ASEAN-Republic of Korea FTA, ATIGA = ASEAN Trade in Goods Agreement, CPTPP = Comprehensive and Progressive Agreement for Trans-Pacific Partnership, RCEP = Regional Comprehensive Economic Partnership. Notes: Estimates are obtained by using a probit model. Robust standard errors are reported in parentheses. * means significant at the 10% level, ** means significant at the 5% level, *** means significant at the 1% level. Source: Authors. Estimation results for restricted models containing only the dummy variables for product groups, the base rate of developed countries and trade variables are presented in Table 3. The results show that the larger the value of exports from developing to developed RCEP members, the stricter the PSRO, while the reverse is true for exports from developed to developing countries. In other words, RCEP PSRO tends to penalize developing members, making it more difficult for companies to comply and use the agreement. There is no consistent finding for the mean base rate of developed RCEP countries, with a large number of coefficients found to be statistically insignificant. 18 Table 4: Estimation Results for Pairwise Comparisons – Full Model (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) AJCEP AANZFTA ACFTA AKFTA ATIGA CPTPP VARIABLES RCEP - Stricter RCEPMore lenient RCEP - Stricter RCEP - More lenient RCEP - Stricter RCEP - More lenient RCEPStricter RCEP - More lenient RCEP - Stricter RCEP - More lenient RCEP - Stricter RCEP - More lenient Consumption -0.218 -0.404*** 0.042 -0.271 0.274** -0.702*** 0.034 -0.733*** 0.043 -0.722*** 0.230** -0.158* (0.141) (0.093) (0.097) (0.173) (0.130) (0.092) (0.132) (0.097) (0.097) (0.163) (0.092) (0.095) Intermediate -0.241** -0.399*** 0.167** 0.021 -0.230** -0.268*** 0.129 -0.648*** -0.148* -0.032 0.544*** -0.296*** (0.096) (0.067) (0.079) (0.088) (0.116) (0.067) (0.098) (0.067) (0.076) (0.079) (0.069) (0.069) Product complexity index -0.085 -0.002 -0.126*** 0.047 -0.215*** 0.125*** -0.004 0.149*** -0.092*** 0.009 0.028 0.020 (0.056) (0.030) (0.032) (0.051) (0.042) (0.032) (0.044) (0.036) (0.035) (0.041) (0.030) (0.033) Base rate, developed countries -0.017 0.145*** 0.023 0.124*** 0.051*** -0.023* 0.005 0.027* -0.002 0.159*** -0.025* 0.065*** (0.019) (0.015) (0.014) (0.033) (0.017) (0.014) (0.016) (0.016) (0.014) (0.030) (0.014) (0.014) Propensity to trade deflection towards developed 0.027 -0.124*** -0.018 -0.174* -0.050*** 0.038*** 0.006 -0.002 -0.000 -0.272** 0.014 -0.051*** (0.018) (0.018) (0.014) (0.099) (0.018) (0.014) (0.016) (0.018) (0.014) (0.118) (0.015) (0.014) Propensity to trade deflection towards developing 0.005 0.002 0.016** 0.075*** 0.007 -0.030*** -0.014* 0.012* -0.004 0.073*** 0.004 -0.026*** (0.010) (0.006) (0.007) (0.010) (0.008) (0.006) (0.008) (0.007) (0.007) (0.010) (0.007) (0.006) Revealed comparative advantage developed -0.013 -0.018 0.012 -0.026 -0.004 -0.008 -0.002 -0.055*** -0.005 -0.007 0.001 -0.007 (0.011) (0.012) (0.009) (0.030) (0.011) (0.010) (0.009) (0.017) (0.008) (0.014) (0.009) (0.012) Revealed comparative advantage developing -0.002 0.002 0.007 0.004 0.015*** -0.008 0.015*** -0.013** 0.000 0.014** 0.002 0.006 (0.007) (0.005) (0.005) (0.006) (0.006) (0.005) (0.005) (0.007) (0.005) (0.006) (0.005) (0.005) Exports RCEP from developing to developed 0.098 -0.229 0.226 -0.376 0.012 -0.772** -0.062 0.149 0.090 -0.381* 0.160 0.020 (0.105) (0.149) (0.182) (0.242) (0.108) (0.372) (0.114) (0.120) (0.116) (0.211) (0.126) (0.220) Imports RCEP to developing from developed -0.180* -0.017 -0.262 -0.091 0.002 -0.316 0.117 -0.301** -0.172* 0.020 0.114 -0.077 (0.097) (0.089) (0.172) (0.150) (0.094) (0.202) (0.088) (0.136) (0.091) (0.101) (0.077) (0.091) Exports to ROW from developed -0.013 -0.003 0.019 -0.025 -0.027 -0.016 -0.050** 0.050** 0.011 -0.015 -0.056** 0.018 (0.022) (0.012) (0.017) (0.020) (0.018) (0.027) (0.022) (0.020) (0.015) (0.016) (0.024) (0.017) Exports to ROW from developing 0.008 0.039* 0.073*** 0.025 0.036 0.024 0.022 0.007 0.049** 0.048* 0.011 -0.021 (0.022) (0.021) (0.026) (0.024) (0.026) (0.030) (0.017) (0.022) (0.022) (0.027) (0.021) (0.020) Tariff reduction -0.006 -0.118*** -0.014 -0.261*** -0.009 -0.047** -0.017 -0.093*** 0.007 -0.319*** 0.041** -0.054*** (0.017) (0.022) (0.018) (0.041) (0.015) (0.020) (0.017) (0.025) (0.015) (0.038) (0.019) (0.017) PRC base rate 0.012*** -0.006 0.002 -0.010 -0.012*** 0.006* 0.011*** 0.000 -0.004 -0.006 0.012*** 0.003 (0.005) (0.004) (0.003) (0.007) (0.004) (0.003) (0.004) (0.004) (0.004) (0.007) (0.003) (0.004) Exports from PRC to RCEP -0.047 -0.241** -0.432** -0.067 -0.114 -0.295 -0.100 -0.221 -0.325** -0.216 -0.004 0.015 (0.147) (0.121) (0.169) (0.139) (0.164) (0.191) (0.110) (0.150) (0.137) (0.164) (0.121) (0.124) Chinese Imports from RCEP 0.025 0.101** -0.044 0.051 0.020 0.111** 0.142*** -0.025 -0.014 0.034 -0.046 -0.008 (0.021) (0.046) (0.027) (0.035) (0.021) (0.048) (0.035) (0.036) (0.020) (0.025) (0.056) (0.031) PRC revealed comparative advantage -0.003 0.044*** 0.035** 0.093*** -0.015 -0.003 -0.046* 0.033* 0.005 0.056** -0.024 0.060*** (0.028) (0.017) (0.017) (0.027) (0.022) (0.018) (0.024) (0.019) (0.019) (0.023) (0.017) (0.018) Constant -1.661*** -0.377*** -1.655*** -2.855*** -1.218*** 0.016 -1.273*** 0.009 -0.492*** -0.730*** -0.227* -0.222* (0.195) (0.117) (0.143) (0.378) (0.168) (0.136) (0.153) (0.143) (0.125) (0.131) (0.118) (0.116) R2 0.337 0.245 0.309 0.220 0.478 0.144 0.555 0.173 0.440 0.158 0.214 0.189 Observations 4,330 4,812 5,080 4,004 4,950 4,521 5,080 4,550 5,127 3,973 4,858 4,815 Sector dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Continued on the next page 19 AJCEP = ASEAN-Japan Comprehensive Economic Partnership, AANZFTA = ASEAN-Australia-New Zealand FTA, ACFTA = ASEAN-People’s Republic of China FTA, AKFTA = ASEAN-Republic of Korea FTA, ATIGA = ASEAN Trade in Goods Agreement, CPTPP = Comprehensive and Progressive Agreement for Trans-Pacific Partnership, RCEP = Regional Comprehensive Economic Partnership. Notes: Estimates are obtained by using a probit model. Robust standard errors are reported in parentheses. * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level. Source: Authors. The addition of new covariates does not considerably change the conclusions drawn from Table 3, except that much fewer coefficients are statistically significant for the two trade variables discussed above and that the higher the average base rate of developed countries, the more likely the product to face a more lenient PSRO. However, the converse holds for products experiencing larger tariff reduction in year 1 under RCEP. Trade deflection variables, which are added to models contained in Table 4, are not associated with clear trends, with the exception of the propensity to trade deflection toward developed countries. In this latter case, the higher the propensity, the less likely for the product to face a more lenient PSRO in four of the six comparisons. No clear trends are identifiable for the revealed comparative advantage variables, as most of their coefficients are statistically insignificant. While consumption products are found to face stricter PSROs, the regression results for the product complexity index show that more complex goods tend to be less likely to be subject to more stringent PSROs, while being characterized by a higher probability of facing more lenient PSROs in four of the comparisons. Lastly, estimation results indicate that products for which PRC has a more marked revealed comparative advantage are more likely to face more lenient PSROs in five of the six comparisons. Relative stringency scores Tables 5 and 6 report results for models using the relative stringency measures as dependent variables. Table 5: Estimation Results for Relative Stringency Measures – Restricted Model (1) (2) (3) (4) VARIABLES Score 1 Score 2 Rank 1 Rank 2 Consumption 0.221*** 0.483*** 0.521*** 0.239 (0.076) (0.130) (0.170) (0.147) Intermediate 0.149** 0.384*** 0.457*** 0.419*** (0.058) (0.084) (0.132) (0.100) Base rate, developed countries 0.001 0.001 0.000 -0.000 (0.003) (0.008) (0.005) (0.006) Exports RCEP from developing to developed 0.252*** 0.424*** 0.366*** 0.003 (0.072) (0.150) (0.137) (0.111) Imports RCEP to developing from developed -0.182** -0.283 -0.016 -0.187* (0.072) (0.205) (0.119) (0.100) Constant -1.388*** (0.089) Continued on the next page 20 (1) (2) (3) (4) VARIABLES Score 1 Score 2 Rank 1 Rank 2 R2 0.120 0.186 0.391 0.143 Observations 5,205 5,118 5,124 5,199 Sector dummies Yes Yes Yes Yes RCEP = Regional Comprehensive Economic Partnership. Notes: Estimates are obtained by using a fractional response approach for score 1, and ordered logit models for score 2, rank 1 and rank 2. Robust standard errors are reported in parentheses. * means significant at the 10% level, ** means significant at the 5% level, *** means significant at the 1% level. Source: Authors. Estimation results for a restricted model containing only the dummy variables for product groups, the base rate of developed countries and trade variables in Table 5 show that the higher the value of exports from developing to developed RCEP members, the stricter PSROs are. This finding is consistent across three of the four models (score 1, score 2 and rank1), the only exception being rank 2 comparing RCEP with ATIGA and CPTPP. However, in this case, exports from developed to developing RCEP countries are shown to face more lenient PSROs the higher their value. This result is also found in the first regression for the first regression where score is the dependent variable. In terms of product groups, RCEP is more restrictive for consumption and intermediate products.