Wednesday, December 25, 2019

Decomposing the Gains from Trade through the Standard Gravity Variables


 

Decomposing the Gains from Trade through the Standard Gravity Variables


One sentence summary: After controlling for proximity, FTAs contribute the most to the welfare gains from trade among other gravity variables following the Uruguay Round.

The corresponding paper by Hakan Yilmazkuday has been accepted for publication at International Economic Journal. Free access to first 50 copies is available here.
 
The corresponding working paper is available here.

 
Abstract
Using the implications of a trade model, this paper measures the gains from trade through the standard gravity variables. Theoretically, it is shown that such gains can be calculated by using the estimated coefficients of these variables in a gravity regression, together with the bilateral expenditure shares of countries investigated. Empirically, the results show that the total actual gains through all gravity variables in the world have increased from about 1% in 1950s to about 5% as of 2015 that can be decomposed as 3.5% through proximity and 1.5% through other gravity variables. Gains through free trade agreements (FTAs) have started dominating among these other variables starting from 1990s, following the Uruguay Round. Across countries, the total gains of OECD countries are about 1.5 times those of others, whereas the total gains of European countries are more than 10 times those of Pacific countries. Calculations based on the future potential gains from trade through policy-oriented gravity variables further suggest that there is room for an additional 0.8% or 0.4% of a welfare gain in the world through having free trade agreements or using common currencies, respectively.


Non-technical Summary
The gains from international trade has been investigated for decades. It has been shown in the literature that these gains can be measured by percentage changes in trade costs and the terms-of-trade, which can be summarized by using home expenditure shares of countries and the trade elasticity. Nevertheless, when welfare changes at the world level are considered, the terms-of-trade effects across countries effectively cancel out so that the welfare gains from trade calculations reduce to the knowledge of reductions in effective trade costs.

Based on this background, this paper proposes calculating the welfare gains from trade through reductions in effective trade costs measured by the standard gravity variables. Among these, gravity dummy variables such as proximity, common language or contiguity are mostly fixed as they represent either the geographical location or the historical characteristics of countries, whereas policy-oriented variables such as free trade agreements (FTAs) or common currencies are subject to changes over time through trade policies. Therefore, for policy evaluation, it is important to investigate the contribution of each gravity variable to the reduction in trade costs and thus to the welfare gains from trade.

This paper achieves such an investigation by decomposing the welfare gains from trade into those through each standard gravity variable. In particular, the following questions are asked:

  • What are the gains from proximity?
  • What are the gains from trading with countries through a free trade agreement?
  • What are the gains from trading with countries using the same currency?
  • What are the gains from trading with contiguous countries?
  • What are the gains from trading with countries with a colonial relationship?
  • What are the gains from trading with countries that speak the same language?

These questions are answered by using the implications of a trade model, where the actual welfare gains are calculated by comparing the current situation of countries with a hypothetical case in which none of the countries benefit from these gravity variables. Accordingly, welfare gains from trade through each gravity variable is theoretically shown to depend on the estimated coefficients of these variables in a typical gravity regression, together with the bilateral import shares, subject to the knowledge of the trade elasticity. The implications of the trade model is estimated by using a typical gravity regression to obtain the corresponding coefficients of the gravity (dummy) variables, and they are normalized by the trade elasticity, which is shown to be nothing more than a scale factor in this investigation while having a comparison across countries and across time. These coefficients are further combined with the bilateral imports data and the current value of gravity variables to obtain the actual welfare gains from trade through each gravity variable.

A similar strategy is used to investigate the potential gains from trade through the policy-oriented gravity variables. In particular, the following additional questions are asked:

  • What are the potential gains from trading with countries through an FTA?
  • What are the potential gains from trading with countries using the same currency?

These additional questions are again answered by using the implications of the trade model, where, this time, the potential welfare gains are calculated by comparing the current situation of countries with a hypothetical case in which they have FTAs or common currencies with all of their trade partners. This is achieved by combining the estimated coefficients of the gravity (dummy) variables (subject to their normalization by the trade elasticity) with the bilateral imports data and one minus the current value of gravity (dummy) variables (of FTAs or common currencies).

The empirical results based on a gravity regression covering the period 1948-2015 suggest that the actual gains from trade in the world through all gravity variables have increased over time from about 1% in 1950s to about 5% by the year of 2015. The latter (for 2015) ranges between 6% and 4% for OECD and non-OECD countries, 17% and 5% for landlocked and coastal countries, 11% and 1% for European and Pacific countries, and 3% and 8% for the United States and Germany, respectively.




When the actual gains are decomposed into their components, the total gains from proximity in the world have increased over time from about 1% in 1950s to about 4% by the year of 2015, whereas the total gains from other gravity variables have increased to about 2% during the same period. The latter (for 2015) ranges between 2% and 1% for OECD and non-OECD countries, 5% and 1% for landlocked and coastal countries, and 4% and 1% for South Asian and South American countries, respectively.


Among the gains through gravity variables other than proximity, the contribution of FTAs has started in late 1950s in the world, and they have dominated among these other variables starting from 1990s, following the Uruguay Round. The same domination has been experienced by OECD countries starting from late 1980s, whereas non-OECD countries, Japan or China had to wait until 2000s. In comparison, despite the increasing contribution of FTAs 2000s, the United States or India have not experienced such domination as of 2015, suggesting that there is potential room for further gains from trade through these policy-oriented variables.


Based on this suggestion, this paper has further calculated the potential gains from trade due to policy-oriented gravity variables that are calculated by comparing the current situation of countries with a hypothetical case in which countries have FTAs or common currencies with all of their trade partners. The corresponding results have shown that the world economy can gain about 0.8% more through FTAs and 0.4% more through common currencies as of 2015. The potential gains from FTAs are about 0.6% for Germany, and 0.9% for China and Japan, reflecting the fact that Germany is already gaining more from trade through FTAs compared to these countries. The potential gains from trade through using common currencies are the highest for Southeast Asian or landlocked countries, suggesting that they can compensate for certain geographical and historical restrictions through using common currencies with their trade partners.

Overall, the actual gains from trade through the standard gravity variables in the world are about 5%, whereas the potential gains from trade through the policy-oriented gravity variables are about 1%, suggesting that future FTAs and currency unions could easily boost the world welfare through the gains from trade. This investigation in this paper can easily be expanded by focusing on alternative gravity variables or the sectoral heterogeneity in estimated coefficients of gravity variables, which we leave for future research.

The corresponding paper by Hakan Yilmazkuday has been accepted for publication at International Economic Journal.
 
The corresponding working paper is available here.