Data mining of the UN Comtrade database in
Data mining of the UN Comtrade database in cooperation with Customs Ronald Jansen Chief of the Trade Statistics Branch United Nations Statistics Division / DESA E-mail: Jansen 1@un. org or Big. Data@un. org 1
Matching Imports and Exports Data 2
Reason for bilateral trade asymmetries o Country of Origin /Country of Destination Ø o Valuation CIF /FOB Ø o Adding Country of consignment Imports and Exports FOB Trade System Ø General Trade System for all 3
Country of Origin / Destination China (A) B records Imports of A A records exports to B Hong Kong (B) C records Imports of A Re-exports to C Netherlands (C) Re-exports to D Germany (D) D records Imports of A (country of origin) 4
Country of Consignment China (A) B records Imports of A A records exports to B Hong Kong (B) C records Imports of B Re-exports to C Netherlands (C) Re-exports to D Germany (D) D records Imports of C 5
Imports CIF / FOB Three Methods to obtain Imports FOB: 1. Recording of Cost, Insurance and Freight per transaction 2. Recording of Cost, Insurance and Freight per Shipment (and partition) 3. Sample Freight and Insurance by HS, Partner country and Mode of 6
Trade System A 2006 global survey showed that 50% of countries use General Trade system and 50% Special Trade system q Difference in coverage (free zones, customs warehousing, processing zones) will lead to discrepancies in recording q All countries encouraged to record all elements of General Trade q 7
Harmonization Process (M=X) 1. 2. 3. 4. 5. Reconciliation exercises – finding common ground Reconciling large trade (Chatham House) Use of imports (origin) as breakdown for partner exports Estimation methods (USITC) Customs interest in solving discrepancies 8
SAS Visual Analytics for UN Comtrade 9
10
11
12
13
14
15
16
17
18
19
20
UN Comtrade in the Sandbox 21
Comtrade in the Sandbox Ø UNECE Big Data Project – Results by November 2015 Ø Available data – 2000 -2014 annual HS and Tariff line data Ø IT specialists – ISTAT, Statistics Netherlands, UNSD, OECD Ø Proposals – Regional Value Chain analysis, Trade asymmetries, Unit-value indices calculations, Trade flow estimations (missing data and forecasting) 22
Regional Value Chain analysis q Replicating – Network Analysis of World Trade (De Benedictis et al. , 2013) a) Global and local centrality measures b) Sectoral Trade Networks • Commodities? (Bananas; Olive Oil [Casieri et al. ]) • Industries? (De. Backer & Miroudot; Sturgeon & Memedovic) c) Restricting to Intermediate Goods trade d) Focusing on Geo-graphical groups q Building on “Mapping Global Value Chains” a) Intra- versus Extra-group trade in intermediate goods q Building on OECD work on “Regional economic integration” - Yamano et al; De. Backer and Miroudot; a) International I-O approach 23
Trade Networks q Commodities o Bananas, Cement, Movies, Oil, Footwear, Engines (De Benedictis) o Olive Oil (Casieri) q Industries o Agriculture and Food, Chemical products, Motor vehicles, electronics, business services, financial services (De. Backer & Miroudot) o Electronics, Passenger vehicles, Apparel (Sturgeon & Memedovic) o CGGC Duke: Electronics, Aerospace, Medical Devices, Horticulture, Wheat, Fruit and vegetables, 24
Project added-value Industry Mapping ü o o o BEC and intermediate goods ü o ü ü GVC Mapping ISIC sectors (Ti. VA) Other? BEC Revision 5 – Split of Economic Categories and End-use; Goods and Services; differentiating within Intermediate goods – generic and specific intermediates Estimating missing trade flows Analyzing bilateral trade asymmetries
Thank you Ronald Jansen Chief of the Trade Statistics Branch United Nations Statistics Division / DESA E-mail: Jansen 1@un. org or Big. Data@un. org 26
- Slides: 26