Research

Trade in value-added and the welfare gains of international fragmentation 

It is notorious in the literature that international fragmentation substantially increases the welfare gains of trade which are, as well known, inversely proportional to the shares of domestic trade in intermediates and final goods. Most calculations, however, implicitly assume that a shutdown of trade in intermediate goods would leave the share of domestic trade in final goods fixed, which overestimates the results. The varieties of goods available in a country would likely change if local firms lost access to intermediate inputs required to produce their final goods. The share of domestic trade in final goods should, therefore, vary following a shutdown of international fragmentation. A simple mechanism describes how this share changes following the shock. The shutdown would create a substitution between intermediate goods and final goods imports. The loss of access to foreign intermediate inputs by a country’s local producers would be compensated by an increase in final goods imports using these inputs, which would attenuate the initial impact of the shutdown. Using this mechanism, we calculate that international fragmentation represents on average approximately 20 % of the total welfare gains of trade, while it would represent at least 50 % without this adjustment.

Are African exports that weak ? A trade in value-added approach

African countries are known to export less than any other group of countries in the world. Many studies have advanced that the main reason for this is the high level of transport costs due to the poor quality of transport infrastructures on the African continent. We first show that, depending on the estimator used, African countries as an aggregate do not necessarily trade a lower volume of gross exports than other countries on average, even though they clearly underperform in exports of final goods. This underperformance, reflected by the greater impact of bilateral trade costs such as distance on African exports of final goods compared to other countries, is not observed for African intermediate goods flows. Second, we formulate a model for trade in value-added by adapting the standard gravity equation to take into account the structure of value-added exports. The proposed model points up the importance of the indirect trade costs of third countries via which a country of origin’s value-added transits before reaching its final destination. When we control for these indirect trade costs in the value-added trade estimation, the additional impact of the bilateral trade costs observed for African countries’ final goods exports is six times lower.

Working papers

Multinational Production and the Overstated Cost of Input Trade Disruptions

Standard trade models predict that disrupting intermediate input trade would severely reduce welfare, with gains from fragmentation representing roughly 50% of total trade gains. We show these costs are overstated by half when accounting for multinational production. Extending the proximity-concentration framework to incorporate country-specific but freely tradable inputs, we demonstrate that shutting down input trade triggers substitution from affiliate production to final goods exports, as foreign firms circumvent local replication and the border effect. Using cross-sectoral variation in R&D intensity as an exogenous proxy for input specificity within a triple-difference design covering 43 countries and 56 sectors (2000-2014), we find the substitution elasticity increases by 0.42 in high-R&D sectors. Calibrating the model, a complete shutdown of intermediate trade reduces real wages by 5.4% on average, compared to 10.9% in models without multinational substitution, a gap largest for small open economies and smallest for diversified economies.

The True Cost of Reshoring

Standard quantitative trade models predict large welfare gains from international fragmentation. This paper embeds those models in a more general framework where productivity and quality are endogenous rather than fixed, and intermediate inputs are differentiated. Imported intermediates expand production possibilities through variety gains but may also generate efficiency losses when domestic firms use technologically distinct foreign inputs. As a result of input differentiation, domestic production using foreign intermediates substitutes for direct imports of final goods from the same source. When input differentiation and endogenous quality adjustment are taken into account, trade in intermediate inputs and trade in final goods become near equivalents in welfare terms. As a result, the welfare contribution of intermediate input trade is much smaller than in standard models. Quantitatively, welfare losses from a complete shutdown of intermediate input trade are on average about 85 percent smaller than those implied by canonical quantitative trade frameworks.

Work in progress