Building Trade and Investment Capacity in Myanmar:

Applied Analysis I

Witada Aunkoonwattaka (PhD)

Trade and Investment Division, ESCAP

anukoonwattaka@un.org

29 May 2014, Yangon, Myanmar

Outline

• Setting stage • Evidence-based policy making (EBPM) and applied

analysis

• Tools for applied analysis

• Databases and measurement issues

• Introduction to an integrated platform for applied analysis

Policymaking approaches

Opinion-based Policy Making

• Anecdote

• Experts

• Seniors

• Gut feeling

Evidence-based Policy (EBP) Making

• Best available research evidence

• Evidence from collected data (empirical evidence)

– Theory

– Data

– Tools for data analysis

• Gains

– Transparency

– Accountability

EBPM for policy analysis

• Tracking, monitoring and evaluation of the results of policies that have been put in place – Ex-post analysis

• Making decisions in public area (including on issues recognition, policy choice and sequencing, or forecasting future developments) – Ex-ante analysis

4

Components of EBPM

Policy questions: Ex. What will be the potential effects of an FTA? Tools: 1. Making inferences from descriptive statistics and trade indicators 2. Partial equilibrium approaches to estimate the potential effects on an

individual product 3. General equilibrium approaches to estimate the potential effects on the

whole economy 4. Approaches to estimate distribution effects Data sources: - online databases

Policy questions at different stages of policymaking

• Before negotiation of an FTA: An analysis of potential costs and benefits (ex-ante) – What a country can supply to its FTA partners? What it can source from the

partners?

– What are expected impacts on production and employment level and composition, welfare, fiscal balance, etc.

– What are the costs of necessary adjustment policies for the adversely affected sectors?

• After its implementation: An impact assessment (ex-post) – Whether the impacts are within the expected range?

– Whether the expected benefits are fully materialized?

– Whether further adjustment policies are necessary?

6

Points to be raised

• The first stage of EBPM is understanding the needs of policymakers in terms of what policy questions they face

• The second stage is choosing the right tool(s) from the menu. – Often, there is more than one alternative.

– The choice should be dictated by policy questions

– Questions requiring ex-ante and ex-post analyses require different tools

• Problems would arise if researchers choose the tools that are not appropriate to answer your questions. – Policymakers and researchers need to understand the advantages and

disadvantages (limitations) of each tool

7

Available methodologies

• Simple indicators (descriptive statistics)

• (Sophisticated) econometric models (including gravity models, etc.

• Simulation techniques – A partial equilibrium model: sectoral analysis

– A general equilibrium model: economy-wide analysis

• They are complementary with different strength/weakness and different (explicit and/or implicit)assumptions

8

Context of the questions and availability of resources dictate the choice of methodology

• Whether the focus is the impacts at macroeconomic level or industry level ?

• Whether ex-post questions or ex-ante questions are being asked?

• Whether the required data are available?

• How much time and resources (technology, human capital, and money) are available?

9

Analytical tools: some simple approaches

Examples from impact assessment of FTAs

10

Making inferences from descriptive statistics trade indicators

• How much of trade is intraregional? – Will the proposed FTA promote trade between trading partners? Are they

then “natural” trading partners?

• What is the comparative advantage of each member? – Which sectors are likely to have export (import) potential?

• Is export of a particular good regionally oriented? – How strong is a regional bias (if there is one)?

• How complementary is trade between a given pair of FTA members? – To what extent the export pattern of a country matches the import pattern

of a region?

• What is a degree of similarity between partners’ exports? – To what extent a country’s export profile overlaps with other FTA members?

11

Descriptive statistics

• Growth and value of trade

• Product shares

• Market shares

…

12 Note: The list is not exhaustive.

Ex. trade indicators

Note: The list is not exhaustive. 13

14

Advantages

• Relatively easy to implement – Data requirements are minimal

– Computation is straightforward

• Some of trade indicators may be found already computed. For example, – WITS

– International Trade Centre

– World Bank

– UNCTAD

– APTIR

– TRADESIFT

15

Ex. Using descriptive statistics for applied analysis

Calculation using data from WITS

35

40

45

50

55

Ind

ex (

0 t

o 1

00

)

Importing countries (over 97% of exports in 2012)

ASEAN

ASEAN+6

GSP

GSP

ASEAN+6 ASEAN+6

Trade complementary (2012) and trade preference

GSP GSP

ASEAN ASEAN

Calculation using data from WITS

Limitations

• They cannot provide precise numbers that quantify the effect of an FTA on trade, production, consumption, or welfare

• They can be meaningless or misleading if data are unsuitably classified – Trade classifications do not match a country’s production structure

– Data are too aggregated

18

Databases for trade flow and trade policy analysis

19

Databases for trade flow analysis

• Aggregated trade data are available from the IMF Direction of Trade Statistics (DOT)

• DOTs provide bilateral import (c.i.f.) and import data of all products on subscription

• For disaggregated trade data, there exist various nomenclatures to classify products:

a. Harmonized system (HS)

b. Standard International Trade Classification (SITC)

20

Databases for disaggregated trade data

a. UN Comtrade

• Available through the World Bank’s WITS portal

• Covers bilateral trade flows at up to the HS 6 level for almost all countries up to 1962

• All trade values are in thousands of current US dollars

• UN Comtrade also reports volumes (in physical units) so that unit values can, at least in principle, be calculated for each good

b. Base Analytique du Commerce International (BACI)

• Created by CEPII (Centre d'Etude Prospectives et d'Informations Internationales), a Paris-based institute, to reconcile discrepancies between UN Comtrade’s import and export data

• BACI trails UN Comtrade with a two-year lag

21

Databases for disaggregated trade data (ct’d)

c. Trade, Production and Protection (TPP) database • Merges trade flows, production and trade protection data available from

different sources into ISIC Rev. 2 data • Data potentially cover 100 developing and developed countries over 1976-

2004 • Input-output tables make it possible to trace vertical linkages

– Unfortunately those linkages cannot be related to trade because the input-output tables do not distinguish between domestic and imported inputs

d. CEPII Trade, Production and Protection database • Combines bilateral trade from BACI. Production figures are based on the TPP

dataset , and complemented with figures by OECD and UNIDO. Bilateral protection data comes from the MAcMap project.

• Proposes bilateral trade and production covers from 1980 to 2006 and 26 industrial sectors in the ISIC Revision 2. The country's coverage varies with the dataset : - 181 countries (Prod_cepii), - 231 exporters & importers (Trade_cepii) - 241 exporters & 164 importers (Tar_cepii)

22

Databases for trade policy analysis

a. Tariff data

• WTO integrated database (IDB) and Consolidated Tariff Schedules database (CTS): MFN applied tariffs and imports of WTO Members at the tariff-line level which often means 8 digits, sometimes even 10 digits, starting in 1996 – WTO Tariff Analysis Online (TAO)

• CEPII and IFPRI's MAcMap database

• AMAD (Agricultural Market Access Database)

• WITS provides access to trade and tariffs databases: – The WTO's IDB and CTS databases

– UN Comtrade

– UNCTAD's TRAINS database

– Global Preferential Trade Agreements Database

23

Databases for trade policy analysis (ct’d)

b. NTM data

• WITS (21 countries, 19 developing + EU and Japan)

• World Bank Temporary Trade Barriers Database

• AMAD (also data on tariff-quotas)

• TPP database

• World Bank TBT database

• Data on ad-valorem equivalents of NTMs from Kee at al. (2009)

• WTO databases

– SPS

– TBT (and also this database on Specific Trade Concerns)

– Subsidies

24

Measurement issues

Nomenclatures

• Trade data have been recorded according to several product nomenclatures depending on the database and year:

– Standard International Trade Classification (SITC)

– Harmonized Commodity Description and Coding System (HS)

a. Harmonized System (HS)

• Last revised in 2012

• Four harmonized levels, by decreasing degree of aggregation:

– Sections (21 lines)

– Chapters (99 lines, also called "HS 2" because chapter codes have two digits)

– Headings (HS 4, 1243 lines)

– Subheadings (HS 6, more than 5000 lines including various special categories)

• Levels beyond HS 6 (HS 8 and 10) are not harmonized so the description of product categories and their number differs between countries

• HS2 and HS4 are the same in all revisions

27

a. Harmonized System (HS) (ct’d)

• Main drawback of the Harmonized System

• Originally designed to organize tariff collection rather than to organize economically meaningful trade statistics

• Traditional products like textile and clothing (Section XI) are over-represented in terms of number of subheadings compared to newer products in machinery, vehicles and instruments (Sections XVI, XVII and XVIII)

28 Share in number of HS 6 lines

Share

in w

orld tra

de

Textile & clothing

Machinery

Vehicles

Chemicals

Base metals

Share in number of HS 6 lines

Share

in w

orld tra

de

Textile & clothing

Machinery

Vehicles

Chemicals

Base metals

b. Standard International Trade Classification (SITC)

• SITC Revision 4 in 2006

• Five levels:

– Sections (1 digit, 10 lines)

– Divisions (2 digits, 67 lines)

– Groups (3 digits, 262 lines)

– Subgroups (4 digits, 1’023 lines)

– Basic heading (5 digits, 2’970 lines)

• Concordance tables between HS 6 2007 subheadings and SITC Rev. 4 basic headings (in both directions) are available on the UN Statistics Division webpage

29

Production data and classification systems

• In some instances, trade data have to be combined with production data, for which there also exist various nomenclatures:

a. SIC, ISIC and CPC (managed by the United Nations)

b. NACE (EU) and NAICS (North American countries)

c. BEC (United Nations) and Rauch classification

30

Mirror data

• Export data, which is typically not (or marginally) part of the tax base, is monitored less carefully by customs administrations than import data

• Even when the object of analysis is exports, one should in general prefer import data from partner countries, a technique called "mirroring“

• However, mirroring should be avoided if one suspects that the value of imports is deliberately underestimated by traders (tariff avoidance) or the product is declared under a product heading with a lower tariff (tariff misinvoicing)

• Import data is also subject to further reporting errors • Under automated custom systems, data is increasingly entered in computer systems

directly by employees of transit companies, but progress on reducing input errors is slow in many developing countries

31

Missing values and zero trade flows

• It is generally difficult to tell true “zero trade” from unreported trade or entry errors – Very often lines with zero trade are omitted by national customs rather than reported

with a zero value

– Sometimes, the missing data can be complemented by mirroring (this is done by the IMF DOTs)

– Sometimes the nature of the data suggests entry errors rather than zero trade; for instance, when a regular trade flow is observed over several years with a zero in between. In that case (only), interpolation is valid

• Industry or country averages may not be very meaningful in the presence of many missing observations, because they will then correspond to different time periods, or contain different countries in different years

– In this case, one should try to use a consistent sample

32

Trade volumes and “unit values”

• UN Comtrade provides not only trade values, but also volumes. They are seldom used because: – They cannot be aggregated – They are badly monitored by customs because most tariffs are ad valorem

• However, sometimes the researcher is interested in calculating prices (“unit values”): trade value divided by volume

• The result is however often tricky to interpret for two reasons: 1. “Composition problem”: what will be observed will not be the price of a good

but an average price of several (unobserved) sub-goods – Wider categories worsen composition problems

2. Because measurement errors in volumes are in the denominator, they can have nonlinear effects. Suppose for example that a very small volume is mistakenly entered in the system. The unit value will become very large and thus seriously bias subsequent calculations – Narrow categories are likely to have small volumes and thus to be sensitive to this

problem

33

WITS

An integrated platform to access trade and trade policy databases

What does WITS offer?

• WITS gives access to trade and tariff related statistical information

• WITS calculates some trade indices at for a country

• WITS includes a partial equilibrium simulation tool and data analysis tools

Data coverage in WITS

Login to WITS http://wits.worldbank.org

WITS Functionality

Data Retrieval and Analysis (3)

Data Retrieval and Analysis

Data availability

Useful readings

• Andriamananjara, Cadot, and Grether (2013). Tools for Applied Goods Trade Policy Analysis: An Introduction, in Arvid Lukauskas, Robert M. Stern, and Gianni Zanini (eds), Handbook of Trade Policy for Development. Oxford University Press.

• ESCAP (2009). Trade Statistics in Policymaking - A Handbook of Commonly Used Trade Indices and Indicators.

• Plummer, Cheong, and Hamanaka (2010). Methodology for Impact Assessment of Free Trade Agreements. Manila, ADB.

• World Bank (2011). User’s Manual-WITS. • World Bank (2013). Online Trade Outcome Indicators- User’s

Manual. World Integrated Trade Solution (WITS). • WTO-UNCTAD (2012), A Practical Guide to Trade Policy Analysis.

Thank you. For more information on WITS,

please visit https://wits.worldbank.org