TRADE LIBERALIZATION AND THE EXTENSIVE MARGIN
| DOI | http://doi.org/10.1111/j.1467-9485.2009.00478.x |
| Author | Purba Mukerji |
| Published date | 01 May 2009 |
| Date | 01 May 2009 |
TRADE LIBERALIZATION AND
THE EXTENSIVE MARGIN
Purba Mukerji
n
Abstract
Trade barriers can lead to the disappearance of products and impose huge costs.
Allowing for the realistic possibility that imported products are substituted by
domestic varieties this paper finds that the cost of protection that allows for
disappearance of products, the ‘Romer cost,’ is higher below a tariff threshold. This
threshold depends on the substitutability of domestic for foreign products. This is
important for developing countries where inferior technology leads to poor
substitutability and traditional calculations underestimate the cost. Analysis of new
varieties trade after the Indian liberalization supports the findings in the context of
a developing country.
I Intro ductio n
A significant portion of the increase in international trade that results from
countries lowering trade barriers has been found to be due to trade in ‘new’
goods. New goods are those that were not traded before the liberalization
occurred and constitute what is known as the extensive margin of trade. Studies
of past episodes of trade liberalization suggest that while there is some increase
in goods already traded before the liberalization (the intensive margin) new
goods’ trade accounts for the largest share of the total increase in trade. Kehoe
and Ruhl (2006) document this pattern in countries of North America and
Europe that were involved in significant trade liberalization and a similar
pattern is observed in this study for the Indian trade liberalization starting in
the early 1990s. This leads to important implications for the cost of trade
protectionism especially in the context of a developing country.
The importance of the extensive margin implies that the true cost of not
liberalizing trade may actually be considerably larger than the traditional
deadweight loss of tariffs calculated on products in existing markets since all
goods that could have potentially existed in the market with free trade actually
do not exist due to tariffs. Therefore the true cost includes the negative welfare
Winner of the 2008 Cairncross Prize: Best Paper by a Young Economist, presented at the
Scottish Economic Society Conference.
n
University of San Francisco
Scottish Journal of Political Economy, Vol. 56, No. 2, May 2009
r2009 The Author
Journal compilation r2009 Scottish Economic Society. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main St, Malden, MA, 02148, USA
141
effects that result from the disappearance of goods because of trade restrictions.
These losses can be significant since entire markets disappear, making society
lose the entire amount of welfare that could have been generated in them. In a
model with differentiated inputs exclusively available from abroad, Romer (1994)
illustrates this using a partial equilibrium model where differentiated goods are not
produced at home and can only be imported. He shows that in a typical economi c
model which implicitly assumes that the set of goods in an economy never changes,
the predicted efficiency loss from a tariff is small; on the order of the square of the
tariff rate. However, with the allowance that international trade can bring new
goods into an economy, this loss can be as much as two times the tariff rate. This
loss is refered as the ‘Romer cost’ in this paper.
The problem is analyzed in the context of a general equilibrium (GE) model,
1
allowing differentiated consumer goods to be produced both at home and
abroad. The GE framework allows for the possibility that when there are trade
restrictions, the domestic market spawns new varieties that would’ve not been
profitable had there been more international competition. This is a significantly
different outcome from the partial equilibrium analysis and definitely more
realistic, since it is usually the presence of import competition that produces the
stimulus for imposing tariffs in the first place.
It is shown that even with the compensating role played by the extra domestic
varieties, the Romer cost of welfare due to tariffs continues to be higher than
the conventionally measured deadweight loss only as long as the tariffs are
higher than a certain threshold. Interestingly this threshold depends on the
elasticity of substitution between differentiated products in the utility function.
When elasticity is high, it is easy to substitute domestic for foreign variety. This
diminishes the magnitude of the Romer cost and lowers the threshold of tariff
beyond which the Romer cost is actually smaller compared with the traditional
deadweight loss.
This highlights the importance of the Romer cost especially in the context of
developing countries. The more underdeveloped and technologically backward
an economy the harder it is for import competitors to produce good substitutes
for the importable goods. In this situation the Romer cost will be significant and
the traditional deadweight loss calculations will grossly underestimate the real
cost of protection. For developed countries on the other hand the degree of
substitutability will likely be high in most product markets and the importance
of the Romer cost will therefore be lower.
The rest of the paper is organized as follows: Section II presents detailed
analysis of India’s trade policy liberalization and the resulting growth in the
extensive and intensive margin of trade, Section III covers the theoretical model,
Section IV discusses the alternative measures of the cost of protection, Section V
presents the results and Section VI concludes.
1
Klenow and Rodriguez-Claire (1997) and Mukerji and Panagariya (2007) consider the
Romer cost in a general equilibrium setup. This present paper builds on these models to study
the practical importance of the traditional deadweight loss vs. the Romer cost of protection.
PURBA MUKERJI142
r2009 The Author
Journal compilation r2009 Scottish Economic Society
To continue reading
Request your trial