CEPII, Recherche et Expertise sur l'economie mondiale
Product Level Trade Elasticities


Description


The "Product Level Trade Elasticities" dataset provides trade elasticities at product level. In the online version of the database, positive (2% of products) or non-significant (at the 1% significance level) HS6 trade elasticities have been substituted by the corresponding average HS4 elasticity. Concerned products are flagged. 

The database therefore contains four variables: 
(i) HS6: the HS6 product category, in revision 2007
(ii) Elasticity: the value of trade elasticity 
(iii) Zero: a dummy indicating whether the elasticity from the estimation was non-significant at the 1% level
(iv) Positive: a dummy indicating whether the elasticity from the estimation was positive and significant

The project website also provides:
(i) Elasticities estimated by HS6 (rev 2007) products and by sector in HS4, TiVA, GTAP-10, WIOD (rev 2013), WIOD (rev 2016), ICIO and ICIO by product type (using BEC classification to distinguish intermediate/final good) classifications.
(ii) Elasticities of transport cost to distance
(iii) The bilateral trade and tariff data for each product (2001, 2004, 2007, 2010, 2013 and 2016)
(iv) The source code
(v) The pre-print version of the reference article under Creative Commons license.


Reference document to cite: Fontagné L., Guimbard H. and Orefice G. (2022) Product-Level Trade Elasticities. Journal of International Economics, vol 137

Person in charge & contact: Lionel Fontagné, elasticitiescepii.fr

Licence: Etalab 2.0

Download
>>>


Elasticities in HS4, TiVA and GTAP-10 classifications are available on the project page.

0.1 version of the dataset (csv format):
ProTEE 0.1

Methodology
>>>


We use the most-disaggregated level of information on trade policies and bilateral imports available for the universe of products and importing countries -- the HS6 that covers over 5,000 different product categories -- for a sample of 152 importing countries.

For each HS6 product category we observe the universe of bilateral trade flows between countries, in value, in a given year, and the tariff (preferential or not) applied to each exporter by each importer on the specific product. This is done for 2001, 2004, 2007, 2010, 2013 and 2016.

A great deal of the variation in tariffs is cross-sectional. For this reason, we explain -- for a given importer (and product) -- the cross-country variation in imports via the cross-country variation in tariffs. We start from the prior that the coefficient associated with tariffs -- a variable trade cost -- corresponds to the import-demand elasticity in a structural gravity equation for bilateral trade.

Our approach also recovers the elasticity of shipping costs with respect to distance.

We benefit from the fine grain of our data, and estimate not only product-level (HS6) trade elasticities but also sector-level trade elasticities by pooling the product-level observations within each sector.

Details on the estimation strategy and robustness tests are provided in:
Fontagné L., Guimbard H. et Orefice G. (2022) Product-Level Trade Elasticities. Journal of International Economics, vol 137
F.A.Q.
>>>


1. Where can I find the actual point estimate for non-significant HS6 elasticities?
Answer: point estimates are provided with standard errors on the project page https://sites.google.com/view/product-level-trade-elasticity

2. Where can I find trade elasticities at the sector level?
Answer: Elasticities estimated by sector in HS4, TiVA and GTAP-10 classifications are provided on the project page https://sites.google.com/view/product-level-trade-elasticity

3. How are sectoral trade elasticities estimated?
Answer:Sectoral trade elasticities are estimated by pooling HS products within sectors

4. Where can I find the trade and tariff dataset used for estimations?
Answer: Trade and tariff data are available on the project page https://sites.google.com/view/product-level-trade-elasticity