DECOMPOSING WELFARE CHANGES IN THE GTAP MODEL

This paper develops a complete decomposition of the change in global welfare in the GTAP model. In particular, this money metric change is broken down into component parts, each of which relates to a quantity change interacting with a distortion in the model. This enables the user to assess, for example, how much of the gains from trade reform are attributable to a given commodity and/or a given region. The commodity and region specific changes in allocative efficiency can be further decomposed by transaction/tax instrument. We find that this greatly facilitates the presentation and analysis of results from GTAP. We motivate the derivation of this decomposition with the case of a one region, three commodity, analogue to the GTAP model. This permits us to focus on purely allocative efficiency effects (no terms of the trade changes). Extension to the multiregion model adds the prospect of terms of trade effects on regional EV, and the multiregion decomposition isolates the contribution of tradable price changes to regional welfare. This is demonstrated in a 3 region, 3 commodity example. Finally, we offer a more complete decomposition which takes into account the impact of changes in endowments and technology on regional welfare. This revised (2001) version introduces a number of important changes to the original (1996) paper. In particular, we build on the new final demand structure for GTAP proposed by McDougall (2001). This includes explicit recognition of changes in the marginal utility of income, as well as a per capita decomposition. We also take account of version 5.0 changes in the standard GTAP model, including the introduction of multiple international margins commodities.


Issue Date:
2001
Publication Type:
Working or Discussion Paper
PURL Identifier:
http://purl.umn.edu/28708
Total Pages:
45
Series Statement:
GTAP Technical Paper 5




 Record created 2017-04-01, last modified 2017-08-24

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)