Financial Deepening for Stronger Growth and Sustainable Recovery © 2023 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions and is partially supported by the Umbrella Facility for Trade trust fund, financed by the governments of the Netherlands, Norway, Sweden, Switzerland and the United Kingdom. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Attribution—Please cite the work as follows: Angella Faith Montfaucon, Massimiliano Cali, Bayu Agnimaruto, Jana Mirjam Silberring, Agnesia Adhissa Hasmand D, Csilla Lakatos, Mochamad Pasha,. (2023) “Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries”, Washington, DC: World Bank. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@ worldbank.org. Cover photo: World Bank Further permission required for reuse. Cover design: Arsianti The report was designed and typeset by Arsianti. Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Table of Contents Abbreviation VI Foreword by UNCTAD and the Government of Indonesia VII Acknowledgements IX Abstract X 1 Introduction 1 2 How does the Collection and Update Process Differ? 5 2.1 Frequency of Updates and Links to Policy 5 2.2 Using Multiple Sources 7 2.3 Customized Database 7 2.4 Key Differences with Existing Data at the time of Compilation 8 3 Usage of the Data 10 3.1 Indicators on the Incidence of NTMs 10 3.2 Indicators of the Costs of NTMs 12 3.3 Identifying Burdensome NTMs: Research Using the Data 13 3.4 Simulating the Impact of NTM Reforms 16 3.5 Cross-Country Comparison 17 4 How the Data is Built 18 Step 1. Sourcing the Regulations and Setting up Series as the Backbone 19 Step 2. Backward and Forward Tracing 19 Step 3. From Regulation to NTM code 21 Step 4. HS Product Extraction 22 Step 5. Building the Regulation-NTM-HS Dataset 25 Step 6. Building the Regulation-HS Dataset 26 Step 7. Building the Regulation-NTM-HS Dataset 27 Step 8. Building the Panel Dataset 27 5 Conclusion 29 References 30 Appendix 31 NOTE: Click THE BUTTONS below to go to THE CHAPTER you wish to read or click HOME BUTTON to go back to Table of Contents page iii Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries FIGURES Figure 1: General Principles on NTMs and Monitoring 1 Figure 2: Example of Data Visualizations on the WITS Website 3 Figure 3: Covid-19 Response Involved Many New NTMs 6 Figure 4: Number of Active Regulations that include Pre-Shipment Inspections (C1) Dropped in 6 November 2021 Figure 5: In 2021, the Government Revoked a Higher Number of Regulations than it Amended 6 Figure 6: TIncrease in the Share of Affected Products from Newly Applied NTMs in 2021 6 Figure 7: Several Government Institutions are Responsible for Issuing NTMs 7 Figure 8: The World Bank Recorded a Higher Number of Cumulative Regulations before 9 the UNCTAD update Figure 9: ...But a Lower Number of Total Active Regulations 9 Figure 10: The Frequency Ratio Varies Between the Different NTM Groups 11 Figure 11: The Coverage Ratio of Compliance With National Standards (SNI) varies by Product 11 End Use Figure 12: Frequency Ratio of Import Approvals (B14) Applied to Specific Green Goods 11 Figure 13: NTMs Have Different AVEs on Intermediate Goods 12 Figure 14: NTMs Have Different AVEs on Renewable Energy Products 12 Figure 15: Importer-Exporters Affected by the 4 NTMs 13 Figure 16: Importing Exporters Represent Notable share of Indonesia’s Exports 13 Figure 17: The 4 NTMs Reduce Firms’ Export Survival 13 Figure 18: The 4 NTMs Reduce the Number of Products and Destinations 14 Figure 19: The 4 NTMs Reduce the Number of Products and Value 14 Figure 20: Firms Exposed to the 4 NTMs have a larger drop in Exports 14 Figure 21: Products Facing Port of Entry Restrictions Had lower Survival During Covid-19 Lockdowns 15 Figure 22: GVC Firms That Faced Port of Entry Restrictions had a Lower Survival Rate Following 15 Covid-19 Figure 23: Monthly Average Export Quantities Dropped More for Firms Exposed to Port of 15 Entry Restrictions Figure 24: IThe Four Trade Reforms Would Boost Indonesia's Economy 16 Figure 25: Eliminating or Streamlining Certain NTMs Could Lower Food Prices 16 Figure 26: Example of the Schematic of a Series 83/M-IND/PER/10/2014 – Mandatory 20 Enforcement of the Indonesian National Standard (SNI) on Concrete Steel Wire for Concrete Construction Purposes Figure 27: An Illustration of Backward and Forward Tracing Using 83/M-IND/PER/10/2014s 20 Figure 28: Interpreting NTM Codes from the Regulations Articles Within a Series 22 Figure 29: Interpreting NTM Codes from a Regulation’s Annexes 22 Figure 30: Extracting HS Codes in the Annexes’ Regulation 23 Figure 31: Extracting HS Codes in the Body Regulation Table 23 Figure 32: Extracting HS Codes When Only Product Names are Stated in Regulation’s Articles 24 Figure 33: No HS Code provided in original Regulation 25 iv Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Figure 34: Extracting HS Codes in the Previous Regulation 25 Figure 35: Building the Regulations-NTM Dataset 26 Figure 36: Building the Regulations-HS Dataset 26 Figure 37: Building the Regulations-NTM-HS Dataset 27 Figure 38: Building the NTM Panel Dataset 28 Figure A1: The World Bank Records a Higher Number of Regulations for Most NTM Groups 33 Figure A2: The World Bank Records a Lower Number of Active Regulations for Most NTM Groups 33 Figure A3: The Number of Regulations Recorded Varies Between NTMs 33 TABLES Table 1: List of custom NTMs in the dataset 7 Table A1: Sample of Regulations which Revoked NTMs in 2021 31 Table A2: Sample of Regulations Related to Covid-19, Implemented in 2020 31 Table A3: List of Repositories for Line Ministries 32 Table A4: AVE by NTM Group 35 Table A5: NTM Classifications 35 Table A6: Extracting HS Codes in the Annexes’ Regulation 37 Table A7: Extracting HS Codes from Interpreting Regulation’s Articles 38 v Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Abbreviations A b b re viation s AVE Ad-valorem Equivalent BPOM National Agency of Drug and Food Control BTKI Indonesian Customs Tariff Book CR Coverage Ratio DGCE Directorate General of Customs and Excise ERIA Economic Research Institute for ASEAN and East Asia EXP Export-related measures FR Frequency Ratio GoI Government of Indonesia HS Harmonized System INP Indonesian National Police INSP Pre-shipment inspection and other formalities MAST Multi-Agency Support Team MoA Ministry of Agriculture MoCI Ministry of Communication and Information Technology MoEF of Environment and Forestry MoEMR Ministry of Energy and Mineral Resources MoF Ministry of Finance MoH Ministry of Health MoI Ministry of Industry MoIT Ministry of Industry and Trade MoMAF Ministry of Marine Affairs and Fisheries MoT Ministry of Trade MoTr Ministry of Transport NTM Non-tariff measures OTH Other measures PC Price control measures QC Quantity-control measures TBT Technical barriers to trade SNI Indonesian National Standard SPS Sanitary and phytosanitary measures UNCTAD United Nations Conference on Trade and Development WBOJ World Bank Jakarta Office WCO World Customs Organization WITS The World Integrated Trade Solution vi Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Foreword Foreword by UNCTAD and Government of Indonesia UNCTAD In our increasingly globalized and industrialized world with fragmented production processes and complex trade networks, concerns about health and safety as well as protection of the environment are rising. Trade related regulations including technical measures are often used to target these concerns and protect health, safety and the environment. These regulations are important for non-trade concerns but also have a significant impact on trade - much higher than tariffs do, in fact. They can increase or decrease trade and trade costs. Most often, they increase trade costs and trade is hampered. Furthermore, there is evidence that more vulnerable groups such as smaller enterprises, women and lower- income countries are disproportionately affected by such regulations. These non-tariff measures (NTMs) are complex by nature and generally not available in a systematic user-friendly way. Lack of transparency alone increases trade costs by about 25 per cent. UNCTAD and the World Bank, together with its other partners in the Multi Agency Support Team on NTMs (FAO, IMF, ITC, OECD, UNIDO, WTO) and experts from countries and organizations such as ERIA, have developed a common language, the International Classification of NTMs, and a standardized approach to collect NTM information to address the lack of transparency. Indonesia was among the first countries in Asia where comprehensive NTM data on all applicable regulations was collected. Together with the government of Indonesia, ERIA and UNCTAD collected data on NTMs for all ten ASEAN countries in 2015 and updated the data in 2018. The data has been published in a systematic manner and is easily accessible for policy makers, researchers and traders. For example, policy makers use the data for trade negotiations and regulatory cooperation (which could, according to UNCTAD research, reduce trade costs also by about 25 per cent), researchers use it to assess the impact of NTMs on trade and vulnerable groups, and traders, particularly small and medium size enterprises, use it to get intelligence about potential export markets. The World Bank office in Indonesia, together with the government of Indonesia, has now updated this information and even widened the time coverage to include regulations that were in force before 2015 but no longer applicable in 2015. This is remarkable and significant for at least two reasons. First, considerable effort was necessary to update and complete the data sufficiently for it to truly be a “panel data series”. This is important for researchers as it allows them to assess the impact of NTMs on a much more detailed level than if they had access only to cross-section data. Second, the government of Indonesia is a role model in supporting transparency. Transparency supports the importers and exporters of Indonesia and contributes to attracting investments. Furthermore, the data is useful also for internal and external coordination and negotiation processes. I congratulate the World Bank office in Jakarta and the government of Indonesia for this outstanding work and contribution to the global transparency agenda. UNCTAD looks forward to continuing to work with the World Bank and Indonesia and use the data for sustainable development. Ralf Peters, Head, Trade Analysis Branch, Division on International Trade and Commodities, UNCTAD. vii Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Foreword Government of Indonesia Currently, the government continues striving to accelerate the process of national economic recovery. Constructive efforts were being enforce through the Job Creation Law, now the Job Creation In Lieu of Law (Perppu). The government keep on formulating appropriate strategies and policies so that Indonesia can optimally achieve its potential. Especially amidst global economic conditions that are still struggling, increasing competitiveness to support domestic resilience is one of the main strategies. Quoting the World Bank›s Indonesia Economic Prospect report on December 2022, foreign investment has been responding positively to the reforms. Looking ahead, complementary reforms under the Perppu of Job Creation with further trade reforms could result in larger multiplier effects on investment and growth. In the medium to long term, economic transformation policies and boosting export growth are expected to create more sustainable economic growth. One of the strategies to encourage export growth is to ensure that Indonesia›s Non-Tariff Measures (NTMs) policies are supporting the export and industry competitiveness. NTMs are often used to achieve public policy goals, such as food safety, consumer protection, and environmental protection. For this reason, it is important to know the implications of this policy for national economic development, so that the Government can maximize the positive impact of NTMs while reducing the costs of NTMs. Therefore, the availability of data and information related to NTM policies in Indonesia is very important to support the process of formulating policies based on scientific evidence and rational arguments (evidence-based policy making), so as to produce optimal output. The data is a product of the World Bank and the Government of Indonesia›s long-standing engagement on trade policy which was expanded through the Coordinating Ministry for Economic Affairs (CMEA). Through CMEA, trade policy dialogue was established with many various government stakeholders including the Ministry of Trade, Ministry of Finance, Ministry of Industry, Ministry of National Development Planning, and also Financial System Stability Committee. Due to this engagement and the government›s own development agenda, there was progress in reforming some measures through the implementation of the Job Creation Law in 2021, notable of which was a removal of pre- shipment inspections. We hope the on-going dialogue and the data will continue to help identify measures that may need to be improved or reformed. We also hope this data-set will be a useful reference in discussions and policy analyses in Indonesia. The collaboration between the World Bank and the Government of Indonesia (CMEA) elaborating this data-set has been a very fruitful experience that we hope to maintain in the future. FERRY IRAWAN Acting Deputy Minister for Macroeconomics and Financial Coordination, Coordinating Ministry for Economic Affairs *** As part of the engagement with the World Bank, DG Customs provided data to World Bank that helped identify the effect of non-tariff measures on Indonesian traders. We also engaged in several active and productive discussions with the World Bank on the rationales of several non-tariff measures and their implementation to see the plausible implications. This is important as not all non-tariff measures are distortive. Therefore identifying those few that need reform is crucial. As customs, this also ensures more efficiency in our ports regarding documentary requirements, helping reduce implementation cost to the government and to traders. We look forward to continuing working with the World Bank to enhance the competitive of Indonesia and smooth port procedures. Furthermore, we are pleased to be part of a process that shows Indonesia as an example for other countries to follow. Rudy Rahmaddi Director of Customs and Excise Information and Technology, Ministry of Finance viii Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Acknowledgements Acknowledgements This manual and related materials for the Indonesia non-tariff measures (NTM) data publication were prepared by a core team consisting of Angella Faith Montfaucon (Economist, World Bank and team lead), Bayu Agnimaruto (Research Analyst, World Bank) and Jana Mirjam Silberring (Junior Trade Analyst, World Bank). The team also led coordination efforts with the United Nations Conference on Trade and Development (UNCTAD). Contributions and comments were received from, Mochamad Pasha (Economist, World Bank), Agnesia Adhissa Hasmand D (Consultant, World Bank) and Massimiliano Cali (Senior Economist, World Bank), who presided over the initiation and early stages of developing the non-tariff measures data for Indonesia. The authors acknowledge the contribution of Wisnu Harto Adiwojoyo, Nabil Rizky Ryandiansyah, Aufa Doarest (Private Sector Specialist, World Bank) and Muhammad Hazmi Ash Shidqi (Consultant, World Bank) who worked together with the core team and contributors of this manual in periodically assembling and updating the data. This manual also benefited from the reviewer comments of Csilla Lakatos (Senior Trade Economist for Indonesia and Timor-Leste, World Bank), Siddhesh Vishwanath Kaushik (Senior Data Scientist, World Bank) and Samuel Munyaneza (Economic Affairs Officer, UNCTAD). Csilla Lakatos also supervised the work. Special thanks to the Monetary and Financial Affairs Division (Deputy I) of the Coordinating Ministry of Economic Affairs (CMEA) for Indonesia, especially Ferry Irawan (Acting Deputy Minister, CMEA), Thasya Pauline (Associate Policy Analyst, CMEA), M. Fahmi Priyatna (Economic Analyst, CMEA) and Anggara Gupta (Activity and Budget Manager, CMEA), who co-organized and hosted the launch event of the Indonesia NTM database release in February 2023. Thanks to Siddhesh Vishwanath Kaushik (Senior Data Scientist, World Bank) and who worked closely with the core team to release the data through the various platforms. Thanks also to Indonesia’s Directorate General for Customs and Excise (DG Customs) of Ministry of Finance, especially Rudy Rahmaddi, who provided customs-level data that enabled more detailed analysis of non- tariff measures in Indonesia. The team benefited from the overall guidance of Satu Kahkonen (Country Director for Indonesia and Timor-Leste, World Bank), Lars Christian Moller (Practice Manager, East Asia and Pacific, World Bank) and Habib Rab (Lead Economist for Indonesia and Timor-Leste, World Bank). Deviana Djalil (Program Assistant, World Bank Jakarta) provided excellent administrative support and coordinated the organization of the data launch event. The dissemination was organized by Jerry Kurniawan, under the guidance of Lestari Boediono Qureshi. The manual was designed and typeset by Arsianti Arsianti.   ix Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Abstract Abstract As import tariffs have been declining over the past decades, non-tariff measures (NTMs) have become the most frequently used measures in trade policy. The increasing use of NTMs in global trade has highlighted the need for timely, high frequency and accurate data in order to better understand the implications that NTMs have on products, firms and the economy. This manual describes the first high-frequency panel dataset built by the World Bank on the universe of NTMs applied by a country, i.e. Indonesia. The manual includes a comprehensive overview of the purpose, building procedures and usage of the data for Indonesia. The dataset expands on and improves on existing data on Indonesian NTMs collected by other institutions (UNCTAD and ERIA) by covering a broader source base, customizing the data, and by increasing the frequency of updates. By documenting the data collection and transformation process, the manual hopes to facilitate the construction of similar datasets in other countries. x Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 1. Introduction 1. Introduction In recent years, governments NTMs include “policy measures other than ordinary customs tariffs that can have increasingly used non- potentially have an economic effect on international trade in goods, changing tariff measures (NTMs) to quantities traded, or prices or both” (UNCTAD 2022). Unlike tariffs, NTMs are achieve a variety of policy objectives.¹ typically used to address potential negative externalities of trade and thus, many NTMs aim primarily at achieving public policy objectives such as food safety or consumer protection. However, even without protectionist intent, they can entail costs which are unnecessarily high to achieve their intended objective. Firms might face administrative burdens, information transparency issues, inconstancy or discriminatory behavior, lack of sector-specific facilities, lack of recognition, time constraints and/or informal payments. Identifying problematic NTMs is key before making a case for the elimination or modification of an existing government regulation. Since the demand for This increases the need for reliable updated data that help identify the extent protection against health of application and the impact of NTMs on the economy. The key to maximizing and environmental hazards is benefits of NTMs is to reduce the cost of compliance with necessary NTMs and expected to increase in future years, so are the number of eliminate unnecessary NTMs. A useful benchmark to identify problematic NTMs NTMs (Cadot et al. 2022). is whether or not they comply with the key principles of the trading system put in place by the World Trade Organization (WTO): do not discriminate among trading partners; and do not create unnecessary obstacles to trade². It is therefore crucial to understand these dynamics to ensure consistency with basic principles of implementation (Figure 1). Figure 1: General Principles on NTMs and Monitoring Non-tariff measures should be Non-tariff measures Non-tariff measures transparent, should not be more should be based on consultative and Non-tariff meaures trade-restrictive than relevant international Monitoring and timely, resulting in Non-tariff measures Non-tariff measures should be necessary to meet a standards, and should assessment could be predictable, coherent, should be consistent should not arbitrarily periodically reviewed legitimate objective, be developed in considered as a and non- with WTO/FTA or unjustifiably to ensure that they and where appropriate, accordance with the possible tool to assess discriminatory commitments and discriminate against meet the intended should focus on WTO TBT/SPS consistency with these application; and obligations. imported products; policy objective and outcome, rather than Agreements/ principles. information about remain relevant. mandating prescriptive non-tariff measures Committee approaches; should be publicly available. Source: APEC Cross-cutting Principles on NTMs 1 As defined by the Multi-Agency Support Team and the Group of Eminent Persons on Non-Tariff Barriers. 2 https://www.wto.org/english/tratop_e/tbt_e/tbt_e.htm 1 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 1. Introduction However, NTM data is First, NTMs cannot usually be tracked from a single source as they are introduced challenging to collect for by many different government entities making it hard to monitor. Second, multiple reasons. regulations may mention product names without specifying the product codes, meaning that the product name needs to be assigned to the relevant product code to produce an accurate data. Third, the regulations are frequently renewed and adapted, making it hard to keep track of all changes over time so as to build a proper panel dataset of measures. The main global source of The data set covers over 100 countries, containing more than 10,000 different NTM data is the UNCTAD regulations and almost 60,000 distinct measures based on the UNCTAD data set³. International Classification of NTMs (UNCTAD 2022). It was collected through official sources, including national laws and regulations and provides annual snapshots by selecting the cut-off date for each year. The data can be accessed through multiple portals like the TRAINS Portal⁴, the World Integrated Trade Solution (WITS)⁵ or the Global Trade Help Desk⁶. The UNCTAD data is used by researchers to study the impact of NTMs, and the broad coverage of the data enables researchers to make comparisons across countries on the NTM application. However, as it is a snapshot Additionally, updates are usually once every few years (especially given the at a specific point in time, coverage of a wide number of countries), making timely tracking of policies it does not enable to track the evolution of NTMs, nor challenging. For Indonesia, the data is available based on a 2015 and 2018 to perform a time-series snapshot, collected in collaboration with ERIA. This data provides the starting analysis of NTMs. point for the dataset construction described in this manual. This manual presents the The manual presents the mechanism through which data is collected, analyzed, process through for which compiled, updated, and used for Indonesia. It also provides a comparison of the NTMs have been hand- collected, compiled and data with other existing datasets as well as the steps for building the data to tracked to generate the World allow its replication in other countries. Additionally, examples on how the data Bank NTM Database for has been applied for policy analysis is provided, including in-depth monitoring Indonesia, closely following and assessment. UNCTAD guidelines. The World Bank NTM The data uses highly disaggregated NTM classification, i.e., 3-digit, with the highest Database improves on frequency available, monthly series, applied to a granular product classification, existing data through more HS-10 digit. The NTM (and product) classification is the most appropriate for frequent updates, a wide variety of sources and a policy analysis as it identifies individual measures introduced or modified by each routine for forward and agency and policy is usually made at a granular product level (as further discussed backward tracing that enables below). The data is updated annually and spans between 2008 and 2021 at the the creation of a panel data. time of writing and release. The monthly panel data construction is based on 652 regulations in Indonesia spanning 13 government ministries and agencies that are responsible for the measures.⁷ 3 https://unctad.org/topic/trade-analysis/non-tariff-measures 4 https://trainsonline.unctad.org/home 5 https://wits.worldbank.org/ 6 https://globaltradehelpdesk.org/en 7 Note that this number will be different in the UNCTAD TRAINS data and WITS, since the data is at an annual frequency using 31st Decem- ber of each year as cut-off date and the panel has a monthly variation. This means that any regulations that were changed within the year (e.g. enacted in February and revoked in September of the same year) would be in the panel data (World Bank) but not in the annual data in UNCTAD. 2 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 1. Introduction The data is organized as a A total of 89 individual 2-digit, 3-digit, and 4-digit NTMs are listed in the panel at the NTM-product- dataset, of which 15 are customized NTMs. Among these 33 are sanitary and month-year level with the phytosanitary measures (SPS), 24 are technical barriers to trade (TBT), 3 are individual NTMs (and tariffs) as dummy variables. pre-shipment inspection and other formalities (INSP), 11 are quantity-control measures (QC), 1 is a price control measure (PC), 2 are other measures (OTH), and 15 are export-related measures (EXP) as of December 2021, the latest update. For each NTM classification a value of 1 is assigned to the specific product- month-year pair if that NTM is in effect, and 0 otherwise. The regulations for each NTM measure as well as the issuing ministry/government agency are also available in a separate NTM-product-regulation-month-year data file which serves as the basis of NTM-product-month-year panel construction. The NTM- product-regulation-month-year data contains the series (base regulations), amendments, and revoked regulations along with the products code and NTM code based on in each regulation. The data and related resources The raw data, containing regulation-product-measure over time (1,946,891 can be found online. observations as of December 2021), the panel data (measure-product panel, 1,472,688 observations over time), and all related material can be downloaded on the World Bank Development Data Hub⁸. Additionally, data visualizations are available on the WITS website⁹, where the user can apply different filters (examples in Figure 2) and download the respective data10. Finally, the data is also available in the UNCTAD format on the UNCTAD TRAINS website11 and WITS, and can thus be used in a multi-country analysis with other countries ‘available data found in TRAINS. The UNCTAD format is annual with cutoff date of December each year and the World Bank data was transformed into this format then UNCTAD added this to the Indonesia data in TRAINS. Figure 2: Example of Data Visualizations on the WITS Website TBT: Technical barriers to trade OTH: Other measures SPS: Sanitar y and 35.9 % 13.1 % phytosanitary measures 8.7 % INSP: Pre-shipment inspection and other formalities QC: Quantity 11.6 % control measures 5.9 % 0.0 % 35.9 % Note: Shares are share within each group Source: WITS database 8 https://datacatalog.worldbank.org/search/dataset/0063543/indonesia_nontariff_measures 9 https://wits.worldbank.org/tariff/non-tariff-measures/en/ntm-datavisualization 10 http://wbmswitsqa201.worldbank.org/tariff/non-tariff-measures/en/ntm-about 11 https://trainsonline.unctad.org/home 3 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 1. Introduction The potential users of the These audiences have different needs and will potentially consume the data in Indonesia NTM data include different ways. Academics can use the data to run econometric and other analyses. government policy makers, Policy makers may consume the data through indicators which are computed international financial institutions (including the based on the raw data and summarize key messages on prevalent NTMs and most World Bank, and relevant affected products, which may warrant for an in-depth review of the regulatory multilaterals), the academic space in the identified areas. These indicators are calculated consistently across all research community and the products and all types of NTMs in the data and are presented at the country private sector. level as well as by different products and product groups. Thus, they facilitate benchmarking across products, NTMs, and across time. Policy makers may also benefit International organizations may use the data to understand and guide their from the data via reports or private sector development and structural reform operational work. Finally, the policy and analytical pieces that summarize the results data may provide a way to easily trace the regulations and issuing institutions for from the data. products of interest to the private sector. The rest of this manual is Chapter 2 elaborates the difference in the data collection process between organized as follows. UNCTAD and the World Bank and shows how these differences translate into differences in the data. Chapter 3 discusses how the data can be used and shows examples of existing analysis and limitations of the data. In Chapter 4 the manual elaborates more on how the data was built. Chapter 5 concludes. 4 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 2. Data Collection and Update 2. How does the Collection and Update Process Differ? While the construction of the World Bank NTM database has followed the guidelines of UNCTAD for the data collection process, key differences also exist aimed to facilitate the use of the data for policy analysis. These include the frequency of updates, the frequency of the data, the sources used, customized NTMs and finally the link to policies such as COVID-19. 2.1 Frequency of Updates and Links to Policy The data is updated Annual updates and backward and forward tracing ensure that only active annually, ensuring only regulations remain in the data and revoked regulations are also timely identified active regulations remain and marked as such (for a sample of regulations which revoked NTMs in 2021 see and inactive regulations are timely identified. Table A1). The policy direction is monitored and creates a reliable panel dataset that is time varying. Each annual update also provides the opportunity to check the data for consistency, possible duplicates and other aspects to continually improve its quality. The annual updates also For instance, since 2020, the data tags NTM-product measures linked to ensure the identification of regulations that were put in place in response to COVID-19 (Figure 3). Other NTMs that were put in place relevant tags are also added as needed. For instance, in 2020 Indonesia enacted in response to transitory shocks or structural reforms. the Omnibus Law for Job creation, subsequent to which hundreds of regulations were implemented including ones on trade policy. The data identifies changes in regulations and subsequently the NTM and products linked to the broader reform and therefore can be traced more easily and if need be, analyzed. For example, in 2020, 12 regulations were related to Covid-19 (Table A2), which were linked to 38 specific NTM codes and 647 HS-10 products (Figure 3). The data is constructed at This high-frequency is necessary given the frequently changing nature of the the monthly level which regulations and allows mapping to equally high-frequency trade data. Tracking is appropriate to capture of regulations therefore allows for timely tracking of new policies and reforms frequent policy changes. especially in an environment with frequent regulatory changes. Figure 4 shows the monthly change in the number of active regulations for a specific NTM. The data allows the user to The high frequency dataset also makes it possible to track the specific product get a sense of new/amended/ and NTMs that the newly applied and/or revoked regulations affect (Figure 6). revoked regulations over time and thus the scope of As a specific example, in November 2021, the government enacted significant regulatory changes, whether reforms to regulations regarding the pre-shipment inspections, where there was more are being added or the largest decline in the number of regulations, and total regulations were at revoked (Figure 5). their lowest (122 regulations by the end of 2021 (Figure 7). 5 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 2. Data Collection and Update Figure 3: Covid-19 Response Involved Many New Figure 4: Number of Active Regulations that include NTMs Pre-Shipment Inspections (C1) Dropped in November 2021 40 35 35 30 30 Number of Regulations Number of NTMs 25 25 20 20 15 2019 15 2020 10 10 2021 5 0 5 Feb Mar Apr Jun Oct Mar Jul 0 2020 2021 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Source: World Bank Jakarta Indonesia NTM Database and Statistics Indonesia Trade data Figure 5: In 2021, the Government Revoked a Figure 6: Increase in the Share of Affected Products from Higher Number of Regulations than it Amended Newly Applied NTMs in 2021 30 7% Became applied NTM completely revoked 17 16 18 Number of amendment/new and 20 12 15 13 Share of affected products based on new and revoked 9 6% 10 4 2 6 revoked regulations 5% 0 4% NTMs -10 -3 -20 3% -14 -30 2% -40 1% -50 -42 0% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Source: World Bank Jakarta Indonesia NTM Database and Statistics Indonesia Trade data 2.2 Using Multiple Sources The ERIA-UNCTAD 2015 When ERIA-UNCTAD 2018 was released, it was checked to see if there are any NTMs was used as a starting new/missed regulations. Regulations from multiple sources including repositories point for backward and of line ministries (the detailed list of repositories is available in the annex in Table forward tracing. A3 and more on how this is done in the next section), hukumonline.com12, and peraturan.go.id13. Using added and multiple NTM-related regulations have been spread out among a total of 13 government sources ensures regulations institutions (ministries, agencies and some are issued as a presidential decree) in are less likely to be missed Indonesia. Although most import and export provisions are issued by the Ministry and forward and backward tracing is more accurate. of Trade, each institution/ministry has its own mandate to provide additional technical provision on certain products (for example technical recommendation and certification for horticulture product come from the Ministry of Agriculture, technical recommendation for iron and steel come from the Ministry of Industry). This makes monitoring NTMs difficult since there is no dedicated government 12 hukumonline.com is a subscription-based site managed by law experts which also provides historical regulation evolution in Indonesia. 13 peraturan.go.id is a site managed by Ministry of Law and Human Rights to accommodate all enacted regulations in Indonesia 6 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 2. Data Collection and Update institution to carry out NTM regulatory review and stocktaking. This dataset provides a consolidated source of regulations that spans across line ministries, agencies and presidential decrees which makes monitoring easier. The most prominent institutions are MoT, MoMAF and MoA (Figure 7). Figure 7: Several Government Institutions are Responsible for Issuing NTMs 200 MoTr 180 MoT MoMAF 160 Number of regulations MoI 140 MoH 120 MoEMR 100 MoEF 80 MoCI 60 MoA 40 INP 20 GoI 0 BPOM LMs 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Note: MoT = Ministry of Trade; MoMF = Ministry of Marine Affairs and Fisheries; MoTr = Ministry of Transport; MoI = Ministry of Industry; BPOM = National Agency of Drug and Food Control; MoIT = Ministry of Industry and Trade; MoH = Ministry of Health; MoEF= Ministry of Environment and Forestry; MoEMR = Ministry of Energy and Mineral Resources; GoI = Government of Indonesia; MoA = Ministry of Agriculture; INP = Indonesian National Police; MoCI = Ministry of Communication and Information Technology. Source: World Bank Indonesia NTM dataset 2.3 Customized Database For the data to be relevant While the data uses the existing 2019 version of NTMs, some additional codes to the Indonesian context, are created by adding a letter to the normal MAST codes to further specify the several customizations measure. For instance, an “R” for recommendation letter is added to B14 import were made to some of the measures. approval. This signifies that B14R is the mandatory recommendation letter for import approval. Another example is export approval being issued only if the exporters have a recommendation letter from a ministry or agency related to the sector. In that case the code will be P11R. This extension allows to qualify the potentially more burdensome measure than the standard one due to the recommendation letter requirement. The full list is provided in Table 1. Table 1: List of custom NTMs in the dataset NTM Description A14R Mandatory recommendation letter for getting import approval A15IP Specific import license for producer importer A15IT Specific import license for registered importer A15R Mandatory recommendation letter for getting import license B14R Mandatory recommendation letter for getting import approval B15IP Specific import license for producer importer B15IT Specific import license for registered importer B15R Mandatory recommendation letter for getting import license C32 Requirement to pass through a very specific port of customs P11R Mandatory recommendation letter for getting export approval P12R Mandatory recommendation letter for getting export license Source: World Bank Indonesia NTM dataset 7 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 2. Data Collection and Update 2.4 Key Differences with Existing Data at the time of compilation Data As described above, there are A comparison was made with the UNCTAD-ERIA data on which the World key differences in the data Bank data was based. While these differences are important, ultimately the two collection for the World Bank datasets are different in that one is a cross-section at a point in time (and a time NTM dataset compared to previously existing data. dimension taken from in force date), while the World Bank data is a time-series panel by design with revoked regulations updated more frequently. Crucially, the World Bank data was transformed into the UNCTAD format and Indonesia data in UNCTAD TRAINS has already been updated with the World Bank data. Therefore, the differences refer to comparisons made with the data at the time when the data was being assembled and this should be kept in mind in the brief discussion below. Overall, the World Bank data This may be due to different reasons: includes a higher number of 1) Differences in frequency of updating the data (annual for World Bank, stock regulations. taking exercise for UNCTAD). 2) While UNCTAD and World Bank are relying on ministerial/agential repositories (e.g., MoT repository), the World Bank uses additional sources such as hukumonline.com and peraturan.go.id. Hukumonline.com provides historical data of regulations which enable users to track the previous and recent version of a regulation. Contrarily, not all ministries provide real time updates on their own repositories. 3) As is the case in many countries, there is no single agency in Indonesia to track NTMs, which can also lead to different number of regulations. 4) World Bank is collecting data and monitoring regulation-measure-product combinations that are active and no-longer active for a time varying dataset, while UNCTAD is a stock taking exercise at a single point in time. Figure 8 shows that in 2018 the Apart from a few exceptions, the World Bank has a lower number of active World Bank recorded a higher NTMs for all broad NTM groups. This analysis can be broken down further to the number of total regulations but lower number of total specific NTM level. For pre-shipment inspections and port of entry restrictions, active regulations (Figure 9) the World Bank records a higher number of regulations compared to UNCTAD. compared to UNCTAD. For compliance with national standards (SNI) on the other hand, the reverse was observed. The lower number of total The 2018 UNCTAD data captures regulations as of March 2018 while the World active regulations recorded Bank data covers active regulations at the end of the year, therefore any regulation by the World Bank in 2018 links back to point (1) and changes in the time period would not reflect the UNCTAD data (i.e. if some were (4) above, i.e. more frequent revoked later in the year, they would appear as still active in the UNCTAD data updates and tracking but not in the World Bank data, which would lead to a higher number of ‘active’ of active (and revoked) regulations in the former). Indeed, upon analyzing the 185 regulations (Figure measure-product pairs based 9) in the UNCTAD data for Indonesia in 2018, 92 were not the latest active as on regulatory updates. of the end of 2018. For these reasons, the World Bank data has a lower number of active regulations. On the other hand, the cumulative number of NTMs is higher for World Bank, partially due to the multiplicity of data sources used, since cumulative is simply the total number of regulations, regardless of whether they are active or not. The difference in stock-taking regulations might have led to different HS-NTM mapping and interpretation of the regulations. A more detailed analysis on this again based on the 2015 and 2018 UNCTAD-ERIA data is provided in Annex A.4. 8 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 2. Data Collection and Update Figure 8: The World Bank Recorded a Higher Figure 9: ...But a Lower Number of Total Active Number of Cumulative Regulations before the Regulation UNCTAD update 900 WBOJ UNCTAD 190 WBOJ UNCTAD Number of active regulations – all 800 Total number of regulations 185 (cumulative) – all NTMs 700 180 600 500 175 NTMs 400 170 300 165 200 160 100 0 155 2015 2018 2015 2018 Note: A breakdown by broad NTM groups or specific NTMs can be found in the Annex Figure A1, Figure A2 and Figure A3. This was comparison done during compilation of the data, before UNCTAD TRAINS added the World bank NTM data to the Indonesia data therein. Source: World Bank Jakarta NTM Database for Indonesia and UNCTAD NTM Data 9 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage 3. Usage of the Data The database can be of These include researchers, policy makers, governments, NGOs and the general interest to a wide variety of public. The data can help answer research questions, analyze the current policy potential users. landscape, enable a systematic review of NTMs, and provide a basis for decisions on policy changes among other things. There are three types of indicators that define the potential economic impact of NTMs: (i) the extent to which NTMs are applied across ranges of products; (ii) the costs they may entail to the trade procedures; and (iii) the value of the externality the NTM aims to address, e.g. how much ensuring the safety of imported meat is worth in terms of consumer welfare. The NTM dataset allows to develop direct measures of indicators (i) and (ii), while estimating (iii) would require additional data that is not immediately available, such as the health and economic costs of disease outbreak from imported meat and the change in probability of an episode of disease outbreak due to the application of the NTM. As such the analysis based on this NTM data would ideally be complemented with one capturing more fully the potential benefits of NTMs, which goes beyond the scope of this manual. 3.1 Indicators on the Incidences of NTMs Useful indicators based on The coverage ratio is the share of import value affected in that year as a share of a broad range of product total imports of that category. The frequency ratio is the number of products that categories can be calculated the measure applies to in that year as a share of total number products in that by users. category. Figure 10 show how the coverage ratio evolved over time for all product groups in Indonesia. The coverage ratio and These can reveal more meaningful analysis in terms of the underlying regulations frequency ratio can also and implications on products. Figure 11 shows the frequency ratio by differentiating be calculated over time for specific NTMs and specific the use of goods, namely consumption goods, capital goods and intermediate products, such as green goods as well as specific measures (compliance with national standards, SNI). goods and technologies. Figure 12 shows that import approvals affect a varying share of green good product categories including nearly 40 percent of energy efficiency technologies and products. Other NTMs such as requirements to pass through a specific port of customs (C3) affect as much as 30 percent of Environmentally Preferable Products based on End-Use or Disposal Characteristics in 2021 (more indicators can be viewed on WITS data visualization14). Trade in green and Eliminating or modifying certain NTMs can play a crucial role in supporting environmental goods international green competitiveness and unlocking the potential for the trade has become increasingly and climate change nexus. However, the next section will show which of these important in recent years. NTMs impose a cost on these goods. These useful indicators can also be viewed using available data visualizations online15 which are based on the panel data. Visualizations may also be used in WITS using the UNCTAD TRAINS version which includes this data. 14 https://wits.worldbank.org/non-tariff-measures/visualization/indonesia-ntm-data.html 15 https://wits.worldbank.org/non-tariff-measures/visualization/indonesia-ntm-data.html 10 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage Figure 10: The Frequency Ratio Varies Between the Figure 11: The Coverage Ratio of Compliance With Different NTM Groups National Standards (SNI) Varies by Product End EXP INSP OTH PC 18% 60% QC SPS TBT 16% Import Value Affected SNI (B7) 50% 14% Share of affected product - 12% Frequency Ratio 40% 10% 30% 8% 20% 6% Capital Goods 4% Consumption Goods 10% Intermediate Goods 2% 0% 0% 2014 2015 2016 2017 2018 2019 2020 2021 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Note: SPS = Sanitary and phytosanitary measures; TBT = Technical barriers to trade; INSP = Pre-shipment inspection and other formalities; QC = Quantity-control measures; OTH = Other measures; EXP = Export-related measures; PC = Price control measures. Source: World Bank staff calculations from World Bank NTM Database and trade data sourced from BPS Figure 12: Frequency Ratio of Import Approvals (B14) Applied to Specific Green Goods Resources and Pollution Management Noise and Vibration Abatement Gas Flairing Emission Reduction Waste Management, Recycling and Remediation Air Pollution Control Efficient Consumption of Energy Technologies and Carbon… Environmental Monitoring, Analysis and Assessment Equipment Clean Up or Remediation of Soil and Water Cleaner or More Resource Efficient Technologies and Products Heat and Energy Management Natural Risk Management Renewable Energy Management of Solid and Hazardous Waste and Recycling… Waste Water Management and Potable Water Treatment Energy Efficiency Environmentally Preferable Products based on End-Use or… Water Supply 0% 20% 40% 60% 80% 100% Frequency Ratio of Import Approvals (B14) applied to specific green goods Note: Frequency Ratio measured in percentage, 2021. Source: World Bank staff calculations 11 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage 3.2 Indicators of the Costs of NTMs A systematic analysis of the AVEs of NTMs involve estimating the uniform tariff that will result in the same effects of NTMs involves trade impacts on the import of a product due to the presence of the NTMs. evaluating their tariff ad- The AVE of an NTM is often interpreted as measuring the distortion imposed valorem equivalents (AVEs). by the NTM to the domestic economy. However, as some NTMs are imposed to address market failures, due to the presence of externalities or public goods, simply interpreting AVEs as measuring distortions would be misleading. To be able to draw meaningful insights, an in-depth review on whether the measures are justified on the products they affect is also needed. For instance, the tariff equivalent of complying with standards may be high, but these may be applied to medical products which would cause public health challenges if for instance these were counterfeit. The results show varying For instance, AVEs of 13% for all technical barriers to trade (TBT classifications levels of costs for different aggregated). Licensing for specific use (E112) have an estimated AVE of 30% for measures. intermediate goods imports (Figure 13). Further, it is possible to break down the analysis by product type or focus on green goods. Results show that there has been an increasing number of NTMs on green goods in recent years with some imposing significant costs on renewable energy products (Figure 14). This suggests that reforming some of these measures could enable access to green technologies from foreign markets (Lakatos et al. 2021b). A list of all statistically significant and positive AVEs on all goods can be found in Table A4. The methodology for the AVE estimation can be found in the annex A.5. Identifying problematic NTMs is key before making a case for the elimination or modification of an existing government regulation, but AVEs may provide insights on the implementation efficiency even when measures are justified. Figure 13: NTMs Have Different AVEs on Figure 14: NTMs Have Different AVEs on Renewable Intermediate Goods Energy Products B14 - Authorization requirements for importing certain products C3: Requirement to pass through specified port of customs C1 - Pre-shipment inspection B84 - Inspection requirement B83: Certification requirement B85 - Traceability requirements E113: Licensing linked with local B21 - Tolerance limits for residues of or production contamination by certain substances A15: Authorization requirement A86 - Quarantine requirement for importers for SPS reasons E112 - Licensing for specified use -10% 10% 30% 50% 70% 0% 10% 20% 30% 40% AVE on renewable energy products AVE on Intermediate Goods Note: Estimated based on a sample from 2008-2021. Source: World Bank staff calculations from World Bank NTM Database. Green goods based on GTN list of green goods 12 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage 3.3 Identifying Burdensome NTMs: Research Using the Data Using the data, Cali et al. While some specific NTMs increase the quality of the products on which they are (2021) show that the average applied to, others act as barriers to trade. Identifying these measures is where the effect of all NTMs masks data is most useful. The World Bank identified four specific measures (out of the significant heterogeneity. 89 in the data) to be among the most burdensome for Indonesian firms, while not appearing necessary to achieve non-protectionist public policy objectives (World Bank, 2022). These are pre-shipment inspections (PSI), restrictions on port of entry, import approvals and mandatory certification with Indonesian National Standards (SNI). For example, analysis shows that importing exporters are more exposed to some of these NTMs relative to pure importers (Figure 15) and these represent a significant amount of export value (Figure 16) as over two-thirds of exports is generated by importing exporters (Cali et al. 2022). Figure 15: Importer-Exporters Affected by the 4 NTMs Figure 16: Importing Exporters Represent Notable share of Indonesia’s Exports All importers Importer-exporter 50% % of Importing Exporters Exposed 70% Share of firms exposed based on 61% whether the effected product is 60% % Export Value represented Share of traders affected by the 60% 54%55% 40% 50% 4 NTMs in Indonesia 40% 30% 39% 40% imported 31%31% 20% 30% 20% 10% 10% 0% 0% PSI (C1) Specific Port Product Import Import Pre-shipment SNI Port of entry of Entry (C3) quality or Approvals Approvals Inspections restrictions performance (B14) (B7) Note: measured as % of firms in the respective group. Note: measured as % of exporter-importer firms. Source: Cali et al. (2022) Source: World Bank staff calculations from World Bank NTM Database and DG Customs Data A study by Cali and They find that NTMs reduce firms’ Figure 17: The 4 NTMs Reduce Firms’ Export Survival Montfaucon (2021) incentives to continue to export. empirically tests how import 0 Firms that face these NTMs % change in survival time for a 1% increase in NTM exposures competition affects economic -2 performance by studying have a shorter life span in the -1.9 these four NTMs. export market, as they become -4 -3 less competitive globally and -6 can no longer continue taking -8 advantage of foreign markets -10 -8.5 (Figure 17). Additionally, NTMs -12 negatively affect the export -11.7 -14 performance for Indonesia: National Import Pre-shipment Port of entry a 10 percent addition in the standard Approvals inspections requirements basket of goods affected by certification NTMs, leads to over 5 percent Note: World Bank staff estimates using all Indonesian exports, drop in values and quantities 2014-18 from DG Customs and Excise, MoF. of exporting companies, by the Source: Cali and Montfaucon 2021 most conservative estimates (Figure 18). 13 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage Furthermore, NTMs Second, they affect exports to new destination markets, therefore making it negatively affect exports of more difficult to take advantage of market access opportunities or change export new and different products, destinations (Figure 19). Overall, the negative effects exceed the negative impact therefore making it more difficult to switch resources of import tariffs. to new products or add different products that would make Indonesia more competitive in a changing global economy. Figure 18: The 4 NTMs Reduce the Number of Figure 19: The 4 NTMs Reduce the Number of Products and Destinations Products and Value 0 0 percent increase in NTM exposure Percentage value/quantity for a 1 percent increase in exposure Percentage change after a 1 -0.04 -0.2 -0.06 -0.1 -0.08 -0.09 -0.09 -0.4 -0.11 -0.15 -0.6 -0.5 -0.2 -0.6 -0.6 -0.2 -0.8 -0.7 Values Products Destinations -0.8 -0.8 -0.8 -1 Quantities -0.9 -0.3 Pre-shipment Port of entry National Import Pre-shipment Port of entry SNI Import inspections restrictions standard approvals inspections Approvals certification Note: World Bank using all export data from 2014 to 2018 sourced from DGCE, MoF. Source: Cali and Montfaucon 2021 Figure 20: Firms Exposed to the 4 NTMs have a Cali et al. (2022) also use the The authors use the larger drop in Exports NTM data to study the effect depreciation of the yuan and Import of a foreign demand shock differentiate between firms SNI approval Pre-shipment Port of entry on exporters and particularly that face NTMs versus firms 0 how trade policy affects the that do not. The results show level of NTM exposure relative to Response of exports to a mean ability of firms to adapt in -2 -1 firms without any exposure these situations. that firms that face NTMs -4 on their inputs suffer from -6 a larger negative impact on -8 their export values (Figure 20). -8 -10 -10 -12 -12 Effect of Covid-19 Lockdowns -14 on Products Affected by Note: World Bank staff estimates using Indonesian exports to NTMs Japan, 2014-18 from DG Customs. Source: Cali et al. (2022) As explained in Chapter Making use of this feature, Majune and Montfaucon (2023) study how the four 2.1, the dataset also allows NTMs impacted the effect that Covid-19 related lockdown measures had on trade. the user to analyze the interactions of NTMs and They found that intermediate imports subject to NTMs were more negatively other policies. affected than those that were not exposed. Those exposed to port of entry restrictions and pre-shipment inspections for instance, were more negatively affected by external lockdowns, which seem to have worsened the supply side factors for imports. Results from the product- This was especially true for imports subject to port of entry restrictions and import country relationship survival approvals (Figure 21). Domestic lockdowns also adversely affected intermediate analysis show that the failure rate of imports subject to imports subject to NTMs, especially measures requiring physical inspection NTMs had lower survival and testing (such as certification with Indonesian product standards or Standar rates during both domestic Nasional Indonesia (SNI)) and import approval processes. and external Covid-19 lockdowns. 14 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage Effect of Covid-19 on GVC Firms Affected by NTMs At the firm level, Ghose and Firms facing any type of NTMs, including the four NTMs, on average have lower Montfaucon (forthcoming) survival rates compared to firms who do not face any NTMs (Figure 22). Further, use the NTM data to analyze the effect that Covid-19 had among GVC firms, firms who faced port of entry restrictions, on average faced on GVC firms that faced the reductions in export quantities and number of transactions, consistent with the four NTMs compared to evidence of major port congestion during Covid-19 (Figure 23). those that did not. These examples show how Identifying potential flaws in the setup of certain NTMs enables policy makers to the World Bank NTM data adjust their regulations accordingly and in turn facilitate trade and growth. In the can assist in answering policy next section, the World Bank simulates the impact of regulatory changes in the relevant questions. four NTMs on the Indonesian economy. Figure 21: Products Facing Port of Entry Restrictions Had lower Survival During Covid-19 Lockdowns Note: Domestic lockdown indicates the months when Indonesia imposed containment measures (from January 2020). External lockdown is the period partner countries imposed lockdowns (from January 2020). Source: Majune and Montfaucon (2023), authors’ compilation using BPS data Figure 22: GVC Firms That Faced Port of Entry Figure 23: Monthly Average Export Quantities Restrictions had a Lower Survival Rate Following Dropped More for Firms Exposed to Port of Covid-19 Entry Restrictions Note: The graph reports results from the event study, where the dependent variable is the monthly average export quantities (a) and monthly average export transactions (b). NTM firms are firms that were exposed to C3 during Feb-Nov 2019. Non-NTM firs are otherwise. Source: Ghose and Montfaucon (forthcoming), Panjiva 15 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage 3.4 Simulating the Impact of NTM Reforms Using estimated AVEs (see Focusing on the same four NTMs discussed in the previous section (pre-shipment section 3.2), these are inspections, restrictions on port of entry of imports, import approval requirements plugged into a computation and mandatory certification with Indonesian product standards), reforms in all General Equilibrium model (CGE) to estimate economy- four measures would boost GDP by 5 percent. The simulated reforms are turning wide impact of NTM reforms import approval into automatic licenses expect for products with quotas, turning for Indonesia. third-part SNI into self-certification except for high-risk products and eliminating pre shipment inspections and port of entry restrictions. The results further show that these reforms would not undermine legitimate concerns and could increase Indonesia’s total exports by 10 percent and investment by 27 percent over the medium- to long run (Figure 24). Analysis also shows that eliminating problematic NTMs and streamlining inefficient NTMs could greatly reduce food prices (Figure 25) Figure 24: Eliminating or Streamlining Certain NTMs Could Lower Food Prices Pre-shipment 0.1 0.3 Inspections 0.3 0.2 1.2 Port of Entry Restricitons 1.4 1.2 8 GDP 2.3 Import Approvals 3.1 Exports 2.8 11.4 Imports 1.2 SNI 4.8 4.5 Investment 8.2 5 All Four Reforms 10.1 9.3 27.4 0 5 10 15 20 25 30 Percentage change (relative to baseline) Note: Measured in percent change relative to baseline, 2008-2018. SNI (B7), Import approval (B14), Port of entry Restrictions (C3), Pre-shipment Inspections (C1) Source: Lakatos et al. (2021a) Figure 25: Eliminating or Streamlining Certain NTMs Would Lower Food Prices Impact of eliminating or streaming certain NTMs on food prices 10% reduction in cost compliance of product -1.9 registration (A81) 10% reduction in cost compliance of certification -4.27 requirement (A83) 10% reduction in cost compliance of inspection -2.45 requirement (A84) Eliminate pre-shipment inspection (C1) -19.59 Eliminate specific port requirements (C3) -11.54 Eliminate monopoly import regime (H11) -39.9 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 Note: Sample period 2008-2017. Source: Cali et al. (forthcoming) 16 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 3. Data Usage 3.5 Cross-country comparison The above examples are all While the data is ideal for this type of in-depth country analysis, it is trickier for on Indonesia. it to be used for cross-country data unless similar data exists for the country one would like to compare with. The data has been added to existing data for Indonesia in UNCTAD TRAINS16 following the UNCTAD format and therefore can be used in cross-country studies, and this partially addresses this limitation. However, since currently the data in its panel form only exists for Indonesia, it cannot be strictly used in its product-measure-month (or even year) in a cross- country analysis. The aim of the release and this detailed manual is that the effort could be replicated in other countries while also contributing to evidence-based policy making in Indonesia. The second limitation is that Additionally, the team responsible for the data collection needs to have adequate the data collection involves local knowledge to understand the NTM implementation processes and the significant investments of content of the regulations. They also need to be familiar with NTM code description time and resources to collect data from a variety of sources and functions. To ensure high quality data, there is a need for consistent human and update it regularly. resources. To help train the staff, UNCTAD provides online training for different audience and for different purpose (see UNCTAD website17). 16 https://trainsonline.unctad.org/home 17 https://unctad.org/topic/trade-analysis/non-tariff-measures/NTMs-policy-support/online-training 17 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building 4. How the Data is Built This section goes into the Building on the UNCTAD -ERIA NTM 2015 stock-take, the World Bank engaged in step-by-step detail on how a labor-intensive exercise of compiling all the relevant regulations on NTMs issued the NTM data was built. in Indonesia by various ministries from 2008, with the help of the Global Trade Alert dataset18. For each regulation, 2-digit, 3-digit, 4-digit NTM codes (based on UNCTAD International Classification of NTMs 2019) were assigned, emanating from that regulation and applied it to the HS codes included in the regulation and/or HS codes corresponding to the description of affected products. The Indonesia NTM dataset The dataset uses a consistent HS 2007 code for goods classification, and the consists of NTMs that are 2019 version of the UNCTAD International Classification of NTMs (UNCTAD 2019) applied in Indonesia from for the NTM codes. This comprehensive NTM database is built by the World 2008-2021. Bank on the basis of data collected by the ERIA in collaboration with UNCTAD. The UNCTAD-ERIA data identifies the stock of NTMs applied by Indonesia in 2015 (and then updated in 2018). The data is reviewed extensively to identify the right coding based on the guidelines from UNCTAD and made time-varying by tracking the individual NTMs applied to each product before and after 2015. A detailed step by step procedure is provided below. According to the UNCTAD Only those measures backed by official mandatory regulations are to be collected guidelines for collecting NTM and classified. There are six important steps provided by UNCTAD for collecting data, for the purpose of the NTMs data, detailed below from point a) to point f): classification, a measure is a mandatory trade control requirement enacted by an 1. Obtain the source data official (UNCTAD 2021)19. a. Identify sources of information b. Identify regulations from each document or source 2. Classify and register the information c. Identify and classify measures within each regulation d. Identify and classify affected products for each measure e. Identify and classify affected countries for each measure f. Identify and classify objectives for each measure, whenever possible Each measure is likely to All of them must be registered. After the data collector registers all relevant affect certain products and information (NTMs and HS codes for the products affected by the measures), countries, and there may also be objectives mentioned the data collection supervisor will validate the accuracy of registered measures explicitly in the text. and codes. The data will then be ready for publication. For creating Indonesia's NTM dataset, a similar process with UNCTAD guidelines was followed, with an additional 2 steps to transform the data into a panel, resulting in a total of eight steps as detailed below: 18 https://www.globaltradealert.org/data_extraction 19 As defined by the Multi-Agency Support Team and the Group of Eminent Persons on Non-Tariff Barriers. 18 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Step 1. Sourcing the Regulations and Setting up Series as the Backbone The NTM regulations are a. Line Ministries (LMs) repositories (the detailed list of repositories is available sourced from various in the annex A.3) sources including: b. hukumonline.com c. peraturan.go.id d. and ERIA-UNCTAD NTM 2015 dataset Some of the initial series are The ERIA-UNCTAD NTM 2015 series serves as the "backbone regulation" or starting taken from ERIA-UNCTAD point to trace backward and forward each applied product-NTM. The series does NTM 2015 dataset (also with the help of the Global not necessarily act as an initial regulation due to its nature of having one or more Trade Alert dataset20), while preceding regulations. Otherwise, the series could serve as a standalone and/or others not in this initial initial regulation if there is no preceding/following regulation. data are sourced from the LM repositories and other regulation resources. It should be noted that Changing the initial series for the purposes of our data does not change the data UNCTAD-ERIA updated the or content in itself since the data is traced forward and backward (as explained in NTM data and the latest the next step) and thus inevitably including of the 2018 updated regulations and one is 2018, where some corrections were made21. regulations that may have been omitted in the 2015 stock take. Step 2. Backward and Forward Tracing The date of entry into force In doing that it is crucial to track the individual regulations over time. The first of the regulation is used to check in that respect is whether each regulation was the first such regulation convert the stock of existing introducing the specific NTM on that specific product, or if it was updating a regulations applied at one point in time (as per the previous regulation. In the latter case (which incidentally is the most common original ERIA data) into a case in our data backward tracing is needed to complete the NTM series. At panel dataset. the same time, tracking the same regulation forward is also needed to ensure consistency of the data over time. For example, consider However, since the data spans from 2008 to 2021, backward and forward tracing Figure 26, which shows from 2008 to 2021 are also needed for all regulations related to Mandatory that regulation 83/M-IND/ PER/10/2014, i.e. Mandatory Enforcement of SNI on Concrete Steel Wire for Concrete Construction Purposes. Enforcement of the The result of this is we find that the initial regulation started in October 2011 Indonesian National Standard (which began from the 42/M-IND/PER/4/2011 regulation), and the most updated (SNI) on Concrete Steel Wire for Concrete Construction regulation is PERMENPERIN 35 2019 from October 201922. Purposes, is available in the initial data. 20 https://www.globaltradealert.org/ 21 https://wits.worldbank.org/ 22 Note that the ministry who issued both of these regulations id the Ministry of Industry. The structure of the title is different (M-IND vs PER- MENPERIN) due to a change in the ministry over the time period. 19 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Figure 26: Example of the Schematic of a Series 83/M-IND/PER/10/2014 – Mandatory Enforcement of the Indonesian National Standard (SNI) on Concrete Steel Wire for Concrete Construction Purposes Note: The illustration shows an example how regulations precede and follow each other. Through backward and forward tracing, it becomes clear how one regulations supersedes the next. Source: Author illustration based on Ministry of Industry repository and hukumonline.com Further consider Figure 27 This states that the regulation revoked a previous regulation, 42/M-IND/ which displays regulation PER/2/2012. If the backward process is continued, 42/M-IND/PER/4/2011 is 83/M-IND/PER/10/2014 obtained as the initial regulation related to Mandatory Enforcement of SNI on article 15. Concrete Steel Wire for Concrete Construction Purposes. As for forward tracing, all the regulations that updated the 83/M-IND/PER/10/2014 regulation are investigated. As a result, we find that 28/M-IND/PER/7/2017 and PERMENPERIN 35 2019 updated the 83/M-IND/PER/10/2014 regulation. Hence, the backward and forward tracing would conclude here and the series for the 83/M-IND/ PER/10/2014 regulation is considered as completed. Figure 27: An Illustration of Backward and Forward Tracing using 83/M-IND/PER/10/2014 Note: The illustration shows an example of how a regulation is traced. Source: Author illustration based on Ministry of Industry repository and hukumonline.com 20 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Step 3. From Regulation to NTM Code The next step is to apply the For each regulation, 2-digit, 3-digit, 4-digit NTM codes (based on UNCTAD NTM(s) code by translating International Classification of NTMs 2019) are assigned. The intention to use every article in each UNCTAD 2019 Version handbook is to get streamlined 2019 NTM nomenclature, regulation. regardless of the promulgation date of the regulations. Reviews, adds, and amends, are conducted whenever appropriate. It is useful to classify NTMs To that end, the international classification developed by the Multi-Agency Support in groups/types as they Team (MAST), an inter-organization group chaired by UNCTAD, is followed. This comprise a large variety of different measures enacted classification includes broad groupings, comprising different NTMs according by different parts of the to their typology, e.g., Sanitary and Phyto-Sanitary Standards (SPS), Technical government. Barriers to Trade (TBT) and pre-shipment inspection and other formalities (INSP) (see annex Table A5 for a full list of all classifications). It also includes more refined measures, classified at the 2- and 3-digit level, which typically match specific measures introduced by the individual regulations. In fact, each ministerial or agency regulation can introduce or modify more than 1 NTM (at the 3-digit level). This level of classification is therefore the appropriate one for policy advice and that is what is used in the analysis that has used the data (See Chapter 3). As outlines in Section 2.3, These are listed in Table 1. In addition, the in-force dates and NTM codes for each for the data to be relevant regulation-NTM pair in the NTM dataset by UNCTAD-ERIA were reviewed and to the Indonesian context, customized NTMs beyond revised (if needed). It makes this exercise and resulting data differ from UNCTAD- UNCTAD NTM 2019 Version ERIA since: handbook are also added. a. In some cases, there is an incorrect in-force date for regulation-NTM code pair in the UNCTAD-ERIA data. b. Possibility of different interpretations of articles between UNCTAD and the World Bank on which specific NTM code the text corresponds to. c. Aim to build a higher frequency of NTM dataset, which is a monthly panel at the product level. NTMs are identified by Figure 28 shows an example using the 83/M-IND/PER/10/2014 regulation series. interpreting the regulation In this case, the regulation 88/M-IND/PER/10/2011 is the initial regulation of the articles within a series. A 83/M-IND/PER/10/2014 series which only contains the requirements to comply series may carry the same NTM without the regulation with the national standard (SNI, coded as B7). However, 83/M-IND/PER/10/2014 issuing new article(s). updated and revoked a previous regulation, which, included import licensing (coded as B15), thus adding measures to Concrete Steel Wire for Concrete Construction Purposes. Therefore, all the regulations which update the 83/M-IND/ PER/10/2014 regulation will have both B7 and B15 NTM codes in the regulation- NTM data. In some cases, the type of measures may be buried in the annexes while unavailable in the main text. Figure 29 is an example of a regulation called PERMENDAG 20 2021, whose annexes contain three NTMs: B14 (import approval), B15 (import licenses), and C3 (port of entry restrictions). 21 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Figure 28: Interpreting NTM Codes from the Regulations Articles Within a Series Note: The illustration shows an example of how each regulation within a series is interpreted to identify the introduced NTMs. Source: Author illustration based on Ministry of Industry repository Figure 29: Interpreting NTM Codes from a Regulation’s Annexes Note: The illustration shows an example of PERMENDAG 20 2021 where measure descriptions are in the Regulations annex. Source: Author illustration based on Ministry of Trade repository Step 4. HS Product Extraction The Harmonized System Various versions of it are used internationally. HS code is used by customs (HS) Code is a standardized authorities worldwide to identify products when assessing duties and taxes numerical method of and for gathering statistics. It has been regularly updated since 1988. The HS classifying traded products. is administrated by the World Customs Organization (WCO). Following WCO, Indonesia has been using: 1. HS version 2007 for regulations that were enacted in 2007-2011 2. HS version 2012 for regulations that were enacted in 2012-2016 3. HS version 2017 for regulations that were enacted in 2017-2021 4. HS version 2022 for regulations that enacted in 2022 The international HS assigns In Indonesia, for the period 2007-2016, products use HS Code at the 10 digits level, specific six-digit codes for with the first six digits being the international HS number. Since 2017, Indonesia varying classifications and commodities. has been using HS Codes at the 8 digits level. The NTM data converts products to the same HS version of 2007, which is the version of the initial year of the data 2008, for consistency. The World Bank uses Indonesia Customs Tariff Book 2007, 2012, and 2017(Buku Tarif Bea Masuk Indonesia 2007, Buku Tarif Kepabeanan Indonesia 2012, and Buku Tarif Kepabeanan Indonesia 2017). 22 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building When extracting HS product extraction from Indonesian regulations, there are four possible cases that can be encountered: a. Case 1: HS codes are Figure 30: Extracting HS Codes in the Annexes’ carried in the annexes of the Regulation regulation. For the first case, many of the regulations carry HS codes in the annexes' regulations. Standard HS extraction procedure from PDF to excel sheet is feasible. For example, this is the case when we once again use the regulation PERMENDAG 20 2021 for illustration in Figure 30. Note: The Figure shows an example of how the HS codes can be carried in the annexes’ regulation. The translated table can be found in the annex (Table A6). Source: Ministry of Trade repository, regulation PERMENDAG 20 2021 b. Case 2 – HS codes are carried Figure 31: Extracting HS Codes in the Body in the body of the regulation. Regulation Table HS codes being carried in the body of the regulation is a common case in Indonesian trade regulations. Standard HS extraction procedure to excel sheet is feasible. An example of this case can be seen in article 2 of a regulation on SNI, 28/M-IND/PER/7/2017 (Figure 31). Note: The Figure shows an example of how the HS codes are carried in the body of the regulation. Source: Ministry of Industry repository, regulation the 28/M-IND/PER/7/2017 23 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building c. Case 3 – HS codes are not explicitly stated in the regulation. In the third case, products are mentioned in the regulation, but the HS codes are not. These products then need to be matched on the basis of their descriptions, with the Indonesian Customs Tariff Book (BTKI). Figure 32 shows an example of this case, where article 2 of regulation PERMENKOMINFO 13 2021 only stated subscriber station and base station as the goods that the NTM applies. In this case, HS codes 8517.11.00, 8517.12.00, 8517.18.00, 8517.61.00, 8471.30.20, 8471.30.90, 8471.50.10, and 8471.50.90 are included in the data (at 8 digit level of HS). This case where product and no HS codes are given however, is only a few of the regulations.23 Figure 32: Extracting HS Codes When Only Product Names are Stated in Regulation’s Articles Note: The Figure shows an example of when the HS codes are not explicitly stated. The translated table can be found in the annex (Table A7). Source: Ministry of Communication and Information Technology repository, regulation PERMENKOMINFO 13 2021 d. Case 4 – HS codes are not provided in the regulation, but the preceding regulation carries them over Usually, this scenario happens when the following regulation only updates the validity period of the previous regulation without any other substantial changes. Since a particular regulation is updating a previous regulation, the HS codes are carried over by the preceding regulation. For example, it can be seen in Figure 33 that 88/M-IND/PER/10/2011 did not provide any clear HS codes in the regulation. Nevertheless, since it is known that 88/M-IND/PER/10/2011 updated a previous regulation (42/M-IND/PER/4/2011), the HS codes in 88/M-IND/PER/10/2011 will be the same as those in 42/M-IND/PER/4/2011 (Figure 34). 23 One of the issues faced in other countries is the product name sometime is too generic like Fishes or Vegetables and this means adding a lot of HS code at the 8 or 10 digit level. While our data does not maintain a repository to identify mapped names to HS codes of such a case, these are far and apart and will be included in subsequent updates. 24 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Figure 33: No HS Code provided in original Figure 34: Extracting HS Codes in the Previous Regulation Regulation Note: The Figure shows an example of when the HS codes are not Note: The Figure shows an example of when the HS codes of a provided in the regulation. preceding regulation can be carried over. Source: Ministry of Industry repository, regulation 88/M-IND/ Source: Ministry of Industry repository, regulation 42/M-IND/ PER/10/2011 PER/4/2011 Step 5. Building the Regulation-NTM-HS Dataset From step 1 until step 3, we Therefore, a table to present a proper dataset for further analysis can be created. get the regulation series and Figure 35 provides an example of the table format for creating the data. This their respective NTM codes. enables the data collector or user to better analyze the NTMs and may provide some basic statistical analysis. In Figure 35, the column At the same time, the regs column indicates the standardized regulation name seri provides us with a traced backward-forward within a unique series. The inst column shows which standardized series name as the backbone of each institution issued particular regulations. The infc column provides the regulations relevant regulation-NTM pair. in force date. The artc and verb columns provide information on regulation articles in verbatim for each applied/carried-over and abolished/not yet applied NTM. Columns ntms19 and desc19 show us the NTM codes (2019 nomenclature) and their descriptions from the UNCTAD handbook. Finally, stat column indicates whether the assigned NTM code is applied (coded as 1) or not yet applied / no longer applies (coded as 0) in the time period. 25 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Figure 35: Building the Regulations-NTM Dataset Note: seri = Standardized series name as the backbone of each relevant regulation-NTM pair; regs = Standardized regulation name that have been traced backward-forward within unique series; inst = Issuing institution; infc = in force date; artc/verb = Articles and in their verbatim manner that referred to each applied/carried-over and abolished/not yet applied NTM; ntms19/desc19 = 3-digit NTM codes (2019 nomenclature) and their descriptions; stat = Assigned NTM status of applied or not yet applied. Source: World bank Indonesia NTM database Step 6. Building the Regulation-HS Dataset Steps 1 to 4 enables us to Similar to step 5, in this step, tables that consist of regulations and HS codes that obtain the regulation series are affected by particular regulations are built. By having Figure 36 table format, and HS codes that are affected the reader could investigate which HS product codes are affected by which by a particular regulation. regulations in Indonesia. The reader also may conduct basic statistical analysis. In Figure 36, the regs column The nomen column provides information about HS nomenclature based on indicates the standardized WCO that is applied in particular regulations. The hs column shows extracted HS regulation name that has codes from each regulation. And finally, the ntms19 columns provide us with the been traced backward- forward within unique series. NTM codes (2019 nomenclature) from the UNCTAD handbook. Figure 36: Building the Regulations-HS Dataset Note: regs = Standardized regulation name that has been traced backward-forward within a unique series; nomen = HS nomenclature based on WCO; hs = Extracted HS codes for each regulation; ntms19 = Identified NTMs from regulation-NTM dataset that applied to each HS codes from each regulation. Source: World bank Indonesia NTM database 26 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Step 7. Building the Regulation-NTM-HS Dataset The Regulation-NTM-HS This will yield a dataset which we have named "NTM_HS_Regulation.dta" which dataset is built by merging is available for download at the World Bank Data Development Data Hub24. The regulation-NTM and regulation-HS datasets Regulation-NTM-HS dataset consists of regulation series, NTM codes, and HS obtained in steps 5 and 6. codes that are affected by particular regulations. Figure 37 shows a preview what the “NTM_HS_Regulation.dta” consists of: series, regulation, in force dates, year, month, NTM, HS codes, and NTM statusFor each regulation-NTM-HS pair, the abolished and not yet applied NTMs are coded 0 in status. While the applied and carried-over NTMs are coded 1 in status. The importance of keeping both 0 and 1 codes of status is to capture the monthly variance of the NTM dataset. Figure 37: Building the Regulations-NTM-HS Dataset Note: seri = Standardized series name as the backbone of each relevant regulation-NTM pair; regs = Standardized regulation name that have been traced backward-forward within unique series; infc = in force date; year = in force year; mont = in force month; ntms19 = 3-digit NTM codes (2019 nomenclature) and their descriptions; HS10_07 = Indonesia HS 2007 10 digits code; stat = Assigned NTM status of applied or not yet applied. Source: World bank Indonesia NTM database Step 8. Building the Panel Dataset The last step is to transform This will yield a dataset which we have called “NTM_HS_Panel_2008-2021.dta” the “NTM_HS_Regulation. and can also be downloaded at the World Bank Data Development Data Hub. dta” data into the NTM-HS- Year-Month format, using Figure 38 shows us a sample of the contents of “NTM_HS_Panel_2008-2021.dta”. a panel data frame with streamlined NTM 2019 and HS 2007 nomenclature. 24 https://datacatalog.worldbank.org/search/dataset/0063543/indonesia_nontariff_measures 27 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 4. Data Building Figure 38: Building the NTM Panel Dataset Note: HS10_07 = Indonesia HS 2007 10 digits code; HS10_07_desc = Description Indonesia HS 2007 10 digits code; year = in force year; mont = in force month; tariff = applied tariff; d_A11 = Dummy prohibitions for sanitary and phytosanitary reasons; d_A12 = Dummy for geographical restrictions on eligibility; d_A13 = dummy for systems approach. Source: World bank Indonesia NTM database 28 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries 5. Conclusion 5. Conclusion The World Bank NTM data The high frequency data covers NTMs based on regulations issued by a multitude for Indonesia offers extensive of agencies to facilitate an accurate and comprehensive time series of NTMs. The information on NTMs in data is hand-collected through different sources and checked rigorously through Indonesia. backward and forward tracing. The data can help users understand the current policy landscape, calculate frequency and coverage ratios, ad-valorem equivalents, answer research and policy questions and understand the implications of specific policies. The data collection process We hope that following a similar methodology will enable other countries to can be adjusted to facilitate increase their understanding of NTMs and therefore contribute to the policy NTM data collection in other debate. Additionally, we hope that the data on Indonesia can be used to analyze countries. NTMs in ways that other existing data does not permit, and we encourage researchers, policy makers and other interested parties to use the data to contribute to Indonesia’s policy debate and global understanding of NTMs and welcome any feedback users may have. 29 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries References References Cadot, O., Malouche, M., and Sáez, S. 2012. Streamlining Non-Tariff Measures – A Toolkit for Policy Makers. The World Bank, Washington, D.C. https://openknowledge.worldbank.org/handle/10986/6019 Cali, M., Le Moglie, M., and Presidente, G. 2021. Gain without Pain? Non-Tariff Measures, Plants’ Productivity and Markups. Policy Research Working Paper Series 9654. The World Bank, Washington, D.C. https:// openknowledge.worldbank.org/handle/10986/35567 Cali, M., and Montfaucon, A.F.L. 2021. Non-Tariff Measures, Import Competition, and Exports. Policy Research Working Paper Series 9801. The World Bank, Washington, D.C. https://openknowledge.worldbank.org/ handle/10986/36387. Cali, M., Ghose, D., Montfaucon, A.F.L., and Ruta, M. 2022. Trade Policy and Exporters’ Resilience: Evidence from Indonesia. Policy Research Working Paper Series 10068. The World Bank, Washington, D.C. https:// openknowledge.worldbank.org/handle/10986/37489 Cali, M., Pasha, M., Darko, F.A., Sumarto, S., Hidayat, T., and Tiwari, S. forthcoming. Nutritional Impacts of Trade Policies: Evidence from Indonesia. World Bank Working Paper. Ghose, D., and Montfaucon, A.F.L. forthcoming. GVC Firms, Covid-19 and NTMs in Indonesia. The World Bank. Lakatos, C., Montfaucon, A.F.L., and Agnimaruto, B. 2021a. Green Goods and Technologies Trade in Indonesia - A firm Level Perspective. Lakatos, C., Montfaucon, A.F.L., and Agnimaruto, B. 2021b. The role of Trade Policies in Indonesia's Green Transition. Majune, S., and Montfaucon, A.F.L. 2023. Trade Policies and Sea and Air freight: The Impact of COVID-19 Lockdowns on Imports and Exports. Policy Research Working Paper Series 10271. The World Bank, Washington, D.C. https://openknowledge.worldbank.org/handle/10986/38497 UNCTAD. 2022. Non-Tariff Measures from A to Z. United Nations Conference on Trade and Development, New York and Geneva. UNCTAD. 2021. Guidelines for the Collection of Data and Official Non-Tariff Measures. United Nations Conference on Trade and Development, New York and Geneva. UNCTAD. 2019. International Classification of Non-Tariff Measures. United Nations Conference on Trade and Development, New York and Geneva. World Bank. 2022. Trade for Growth and Economic Transformation. Indonesia Economic Prospects (IEP). The World Bank, Jakarta. https://www.worldbank.org/en/country/indonesia/publication/indonesia-economic- prospect 30 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix Appendix A.1 Sample of Revoked NTMs 2021 Table A1: Sample of Regulations which Revoked NTMs in 2021 Institution Revoking In force Revoked NTM NTM Description regulation date regulation Code MoA PERMENTAN 16 18-05-21 18/PERMENTAN/ A15 Authorization requirement for 2021 OT.140/4/2009 importers for SPS reasons MoMAF PERMENKP 1 15-01-21 8/PERMEN- A82 Testing requirement 2021 KP/2020 MoMAF PERMENKP 1 15-01-21 8/PERMEN- A83 Certification requirement 2021 KP/2020 MoT PERMENDAG 19 19-11-21 13/M-DAG/ P169 Conformity-assessment measures 2021 PER/3/2012 n.e.s. MoT PERMENDAG 19 19-11-21 32/M-DAG/ P169 Conformity-assessment measures 2021 PER/5/2017 n.e.s. MoT PERMENDAG 19 19-11-21 33/M-DAG/ P169 Conformity-assessment measures 2021 PER/5/2015 n.e.s. MoT PERMENDAG 19 19-11-21 72/M-DAG/ P169 Conformity-assessment measures 2021 PER/12/2013 n.e.s. MoT PERMENDAG 19 19-11-21 PERMENDAG 1 B14 Authorization requirements for 2021 2018 importing certain products MoT PERMENDAG 19 19-11-21 PERMENDAG 1 B15 Authorization requirements for 2021 2018 importers Source: World Bank Indonesia NTM dataset A.2 Sample of Covid-19 related regulations Table A2: Sample of Regulations Related to Covid-19, Implemented in 2020 Institution Regulation In Forced Date MoT PERMENDAG 10 2020 7/2/2020 MoT PERMENDAG 23 2020 18/3/2020 MoT PERMENDAG 27 2020 18/3/2020 MoT PERMENDAG 28 2020 23/3/2020 MoT PERMENDAG 34 2020 1/4/2020 MoT PERMENDAG 37 2020 2/4/2020 MoT PERMENDAG 57 2020 19/6/2020 MoT PERMENDAG 74 2020 25/10/2020 MoA PERMENTAN 13 2020 2/4/2020 MoI SURAT EDARAN MENPERIN 5 2020 7/4/2020 MoI SURAT EDARAN MENPERIN 6 2020 7/4/2020 Source: World Bank Indonesia NTM dataset 31 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix A.3 Repositories Table A3: List of Repositories for Line Ministries LMs Abbreviation Repository The National Agency for Drug and Food Control BPOM https://jdih.pom.go.id/ Government of Indonesia GoI https://peraturan.go.id/pp.html Indonesian National Police INP https://jdih.polri.go.id/ Ministry of Agriculture MoA https://jdih.pertanian.go.id/ Ministry of Communication and Information MoCI http://jdih.kominfo.go.id/ Ministry of Environment and Forestry MoEF https://jdih.menlhk.go.id/ Ministry of Energy and Mineral Resource MoEMR https://jdih.esdm.go.id/ Ministry of Health MoH https://jdih.kemkes.go.id/ Ministry of Industry MoI http://jdih.kemenperin.go.id/ Ministry of Marine Affair and Fishery MoMAF https://jdih.kkp.go.id/ Ministry of Trade MoT http://jdih.kemendag.go.id/ Ministry of Transportation MoTr https://jdih.dephub.go.id/ Source: World Bank Indonesia NTM dataset A.4 Differences between the UNCTAD and World Bank data at the time of compilation Note that these represent differences prior to the UNCTAD update of 2023 which has added the World Bank data to the Indonesia data in UNCTAD TRAINS. Once the data was assembled, further checks were made to compare to the UNCTAD-ERIA data of 2015 and 2018 on which the World Bank data was based. To test this out, difference of frequency ratio (FR) and pairwise correlation of the HS-NTM mapping in each year was estimated. It’s important to point out that while FR differences and correlations may not provide a full explanation, they give us a glimpse how the World Bank and UNCTAD interpreted and mapped HS-NTMs differently when compared with the UNCTAD data at the time. The differences between the World Bank and UNCTAD further demonstrate how the World Bank updated and expanded on the UNCTAD data that was available for Indonesia. Now that the UNCTAD data includes this data however, these differences may no longer be as relevant. 32 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix Figure A1: The World Bank Recorded a Higher Figure A2: The World Bank Recorded a Lower Number of Regulations for Most NTM Groups Number of Active Regulations for Most NTM Groups 450 140 WBOJ World Bank UNCTAD Number of regulations 400 120 UNCTAD Number of regulations 350 100 300 250 80 200 60 150 40 100 20 50 0 0 SPS SPS EXP INSP OTH TBT EXP INSP OTH TBT PC QC PC QC SPS SPS EXP INSP OTH TBT EXP INSP OTH TBT QC QC 2015 2018 2015 2018 Note: SPS = Sanitary and phytosanitary measures; TBT = Technical barriers to trade; INSP = Pre-shipment inspection and other formalities; QC = Quantity-control measures; OTH = Other measures; EXP = Export-related measures; PC = Price control measures. Note that one regulation may have various NTM codes therefore this chart may not necessarily add up to all the total active regulations. This was comparison done during compilation of the data, before UNCTAD TRAINS added the World bank NTM data to the Indonesia data therein. Source: World Bank Jakarta NTM Database for Indonesia and UNCTAD NTM Data Figure A3: The Number of Regulations Recorded Varies Between NTMs 100 WBOJ UNCTAD 80 Number of regulations 60 40 20 0 C1 C3 B14 B7 C1 C3 B14 B7 2015 2018 Note: C1=PSI; C3=Specific Port of Entry; B14=Import Approvals; and B7=Product quality or performance. Source: World Bank Jakarta NTM Database for Indonesia and UNCTAD NTM Data A.5 AVE Estimation Methodology The ad-valorem equivalents (AVE) of NTMs are estimated by comparing the trade effect of NTMs to the one from tariffs. Specifically, it is theorized that the total effect of NTMs is a product of trade elasticity and the AVEs. The following regression specification is then estimated using the EA countries’ import data as follows: 33 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix Eq.1 is the first step to get the coefficients of NTM and tariff. They are β_j and β_1, respectively lnV_it is the log import value of commodity i (HS 10) at year t. tariff_it is the ad-valorem tariff of commodity i at year t. NTM_ijt is a dummy that takes value of 1 if NTM of interest j affects commodity i at year t. NTM_ikt is a dummy that takes value of 1 for all other NTMs k that affect commodity i at year t. α_i is the product dummy that serves as a control for other product characteristics α_t is the year dummy that serve as a control for shocks to a given year ε_it is the error term. Eq.2 is the second step that will give us the estimated of unique AVE for each NTM j. The AVE is defined as the ratio between estimated coefficient of NTM j and estimated coefficient of ad-valorem tariff, both of which we already derived from Eq.1. Essentially, this allows turning NTMs into “tariff units” since NTMs are regulatory text which are represented by a dummy variable. The AVE from Eq.2 is only feasible and calculated if the estimated coefficient of β_j and β_1 are statistically significant. For product groups, the estimation is done at HS-10 product level if the product is within that product category or group (sub samples). Due to the differences in the NTM data used compared to previous studies, AVEs may be different from other existing estimates in the literature. A.6 Ad-Valorem Equivalent Result Table A4: AVE by NTM Group NTM Group NTM Code and Description AVE SPS: Sanitary and phytosanitary measures A42; Hygienic practice during production 56% A64; Storage and transport conditions 55% A83; Certification requirement 13% A86; Quarantine requirement 34% TBT: Technical barriers to trade B15; Authorization requirements for importers 12% B84; Inspection requirement 10% B85; Traceability requirements 21% 34 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix INSP; Pre-shipment inspection and other C1; Pre-shipment inspection 21% formalities Non-automatic import licensing, quotas, E316; Prohibition of used, repaired or remanufactured goods 22% prohibitions, quantity-control measures and other restrictions Source: World Bank staff calculations from World Bank NTM Database and BPS Data A.7 Full List of NTM Classifications Table A5: NTM Classifications NTM NTM Broad Specific Description Classification Group WB NTM Sanitary and SPS A Chapter A deals with sanitary and phytosanitary Phytosanitary measures. The chapter outlines measures such as Measures those restricting substances, ensuring food safety and preventing the dissemination of diseases or pests. Chapter A also includes all conformity-assessment Technical Measure measures related to food safety, such as certification, testing and inspection, and quarantine. Technical Barriers TBT B Chapter B provides a collection of technical to Trade measures, also called technical barriers to trade. The chapter describes measures relating to product characteristics such as technical specifications and quality requirements; related processes and production methods; and measures such as labelling and packaging in relation to environmental protection, consumer safety and national security. As in the case of sanitary and phytosanitary measures, chapter B includes all conformity-assessment measures related to technical requirements, such as certification, testing and inspection. Pre-shipment INSP C Chapter C, the last chapter in the technical measures inspection and section, classifies the measures related to pre-shipment other formalities inspections and other customs formalities 35 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix Contingent D Chapter D groups contingent measures, that is, those trade-protective measures implemented to counteract the adverse measures effects of imports in the market of the importing country, including measures aimed at tackling unfair foreign trade practices. These include anti-dumping, countervailing and safeguard measures. Non-automatic QC E Non-Technical Measure import licensing, quotas, prohibitions, Chapters E and F feature the “hard” measures that are quantity-control traditionally used in trade policy. Chapter E includes measures and licensing, quotas and other quantity-control measures, other restrictions including tariff-rate quotas. Chapter F lists the price- not including control measures that are implemented to control sanitary and or affect the prices of imported goods. Among the phytosanitary examples are those measures designed to support measures or the domestic prices of certain products when the measures import prices of these goods are lower, to establish the relating to domestic prices of certain products because of price technical barriers fluctuation in domestic markets or price instability in a to trade foreign market and to increase or preserve tax revenue. Price-control F This category also includes measures other than tariffs measures, measures that increase the cost of imports in a similar including manner (para-tariff measures) additional taxes and charges Finance G Chapter G lists the finance measures. The chapter measures outlines measures restricting the payments of imports, for example when the access and cost of foreign exchange is regulated. It also includes measures imposing restrictions on terms of payment. Measures OTH H Chapter H includes those measures affecting affecting competition – those that grant exclusive or special competition preferences or privileges to one or more limited group of economic operators. They are mainly monopolistic measures, such as State trading, sole importing agencies or compulsory national insurance or transport. Trade-related OTH I Chapter I deals with trade-related investment measures investment and groups the measures that restrict investment by measures requiring local content or requesting that investment be related to export in order to balance imports. 36 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix Distribution J Chapters J and K relate to the way products – or restrictions services connected to the products – are marketed after being imported. They are considered non-tariff measures because they could affect the decision to import such products or services. Chapter J, on Non-Technical Measure distribution restrictions, describes restrictive measures related to the internal distribution of imported products. Chapter K deals with restrictions on post-sales services, for example restrictions on the provision of accessory services. Restrictions K on post-sales services. Subsidies and L Chapter L contains measures that relate to the subsidies other forms of that affect trade. support Government M Chapter M, on government procurement restrictions, procurement describes the restrictions bidders may find when trying restrictions to sell their products to a foreign government. Intellectual N Chapter N contains restrictions related to intellectual property property measures and rights. Rules of origin O Chapter O, on rules of origin, groups the measures that restrict the origin of products or its inputs Exports Export-related EXP P Chapter P, the last chapter, is on export measures. The measures chapter groups the measures applied by a country to its exports, inter alia, export taxes, export quotas and export prohibitions. Source: UNCTAD 2019 A.8 Extracting HS Code from Regulation Table A6: Extracting HS Codes in the Annexes’ Regulation No Pos Tarif/HS Description 71.01 Pearls, natural or cultured, whether or not worked or graded, but not strung, mounted or arranged; pearls, natural or cultured, temporarily strung for easy transport. 190. 7101.10.00 - Natural pearls - Cultured pearls: 191. 7101.21.00 -- Not graded 192. 7101.22.00 -- Graded 71.16 Articles of natural or cultured pearls, precious or semi-precious stones (natural, synthetic or reconstructed). 193. 7116.10.00 - From natural pearls or cultured Source: Ministry of Industry repository, regulation PERMENDAG 20 2021 37 Building a Dataset for Non-Tariff Measures and its Usage: The Case of Indonesia and Applicability for Other Countries Appendix Table A7: Extracting HS Codes from Interpreting Regulation’s Articles Article 2 1) Telecommunications equipment and/or mobile telecommunications devices as referred to in Article 1 shall include: a. Subscriber stations; and b. Base stations. 2) Telecommunications equipment and/or mobile telecommunications devices as referred to in paragraph (1) shall be based on the following technological standards: a. Long-Term Evolution; and b. International Mobile Telecommunications-2020. Source: Ministry of Industry repository, regulation PERMENKOMINFO 13 2021 38