Trade Policy under Asymmetric Information

Abstract

We consider optimal trade policy for a large country with private information. We show that the optimal tariff leads to a signaling equilibrium with higher tariffs and lower welfare than under complete information, whereas the optimal import quota replicates the complete information equilibrium and thus is superior to the tariff. We also show that, with the tariff, the country may be better off being uninformed. Finally, we show that if the importing nation cannot commit to its tariff, the use of futures contracts together with the dynamically consistent tariff leads to the same equilibrium as under complete information with commitment.

Full text

Page 1

IOWA STATE UNIVERSITY


Department of Economics
Working Papers Series

Ames, Iowa 50011
Iowa State University does not discriminate on the basis of race, color, age, religion, national origin, sexual orientation, sex, marital status,
disability, or status as a U.S. Vietnam Era Veteran. Any persons having inquiries concerning this may contact the Director of Affirmative
Action, 1031 Wallace Road Office Building, Room 101, 515-294-7612.
Trade Policy under Asymmetric Information
Harvey E. Lapan, Tigran Melkonyan
September 2002

Working Paper # 02007

Page 2

Trade Policy under Asymmetric Information

Harvey E. Lapan and Tigran Melkonyan*


December 2001



Corresponding author :

Harvey E. Lapan
Department of Economics
283 Heady Hall
Iowa State University
Ames, IA 50011
hlapan@iastate.edu
phone : 515-294-5917
fax : 515-294-0221








* The authors are, respectively, Professor of Economics, Iowa State University; and Assistant Professor of
Agricultural and Resource Economics, University of Maryland.

Page 3

Trade Policy under Asymmetric Information



Abstract

We consider optimal trade policy for a large country with private information. We show that the
optimal tariff leads to a signaling equilibrium with higher tariffs and lower welfare than under
complete information, whereas the optimal import quota replicates the complete information
equilibrium and thus is superior to the tariff. We also show that, with the tariff, the country may
be better off being uninformed. Finally, we show that if the importing nation cannot commit to
its tariff, the use of futures contracts together with the dynamically consistent tariff leads to the
same equilibrium as under complete information with commitment.

1

Page 4

Trade Policy under Asymmetric Information

I. Introduction
One of the most prolific literatures in international economics concerns the relationship
between tariffs and quotas1. As the literature shows, while the two tools can be equivalent in a
deterministic, competitive setting, there are a number of scenarios under which they differ. One
such circumstance is when firms possess market power, since the residual demand (or supply)
curve firms face differs under the two instruments. Another case in which the two tools differ is
in the presence of uncertainty. If the policy-maker faces uncertainty when she sets the policy
tool, and if the tool cannot be made contingent upon the random variable, then the two policies
give rise to different probability distributions of outcomes. While a normative comparison of the
two tools cannot be made without specifying the economic model and the reason that
commercial policy is being used, the general consensus is probably that tariffs are superior to
quotas. The superiority of tariffs, in most cases, arises because price and quantity decisions can
respond to the realization of the random variable more fully under a tariff than under a quota.
What has not been widely discussed is the fact that the level of the policy tool may be
influenced by private information. For example, consider a government that can impose import
restrictions on a good (e.g., oil), and assume the optimal level of these restrictions depends upon
government reserves. Assume that under full information the optimal tariff decreases (or the
import quota decreases) as oil reserves increase. If all parties know the oil reserves, the
government can set the policy tool without considering how it affects expectations concerning
reserves, and thus the two policy instruments will be equivalent (assuming all the usual caveats
apply). However, if oil reserves are private information, exporters may use the level of the tariff
or quota to make inferences concerning these reserves. In this case, the importing government,
when setting the tariff (or quota), will take into account not only the objective being pursued in
using this instrument, but also the impact of the level of the instrument on exporter’s beliefs.
Thus, either the tariff or the quota may serve as a signal about the level of reserves (also
called the type) of the importing country. However, the impact on exporting nations of the
actual type of the importing country will differ under the two policy regimes and thus the two
policy tools will give rise to different equilibria. With the increased role of governments in

1 The comparison of price-based tools and quantity-based tools is certainly not restricted to the international
economics literature. One of the more contemporary areas where this comparison is also germane is in the
environmental literature where taxes and command and control regulations are compared.

2

Page 5

modern economies, and especially with the growing importance of Communist, or formerly
Communist countries (e.g., China and Russia) in the world trading system, it should be apparent
that this discussion is more than academic. It is clear that some governments, at the same time
they are either setting trade policy or entering into direct agreements with foreign buyers or
suppliers, also have private information which would be relevant in forecasting price (and other
aggregates). One need look no further than commodity markets to see that announcements of
past purchases (or lack of purchases) by foreign countries can affect prices. Furthermore, even
in the case of a market-oriented country like Canada, the Wheat Board may simultaneously
possess government powers for affecting trade plus private information (such as wheat stocks).
Moreover, the strong push toward tariffication in recent GATT (and WTO) negotiations is
largely predicated upon the belief that the impact of tariffs is more transparent than that of
quotas. Thus, it would seem important to compare the informational implications of each tool.
Collie and Hviid (1994) is one of the first attempts to analyze the signaling role of tariffs.
They use a partial equilibrium model in which a nation imports from a monopolist supplier, and
the domestic government uses import tariffs to recapture some of the monopoly rents. However,
since the government possesses private information about domestic demand, the tariff may signal
the type of consumer to the foreign monopolist. As a result of this signaling effect, the tariff will
be higher and the country will be worse off than under complete information. Note that in their
model the tariff is not the first best instrument under complete information.
In this paper we continue the analysis of the signaling role of different trade policy
instruments. Since any normative comparison of policy instruments must provide a rationale for
the use of policy, we work within the context of a large country that uses the tariff, or quota, to
improve its terms of trade and domestic welfare. The basic model is the standard two good
international trade model, where one large country (US) uses trade policy to improve welfare,
whereas all other countries (ROW) pursue free trade. The specific model we use is simplified
enough to be tractable, yet rich enough to demonstrate equivalence of tariffs and quotas under
full information, and the inferiority of quotas to tariffs under uncertainty. Similarly to Collie
and Hviid (1994), we show that the tariff in the signaling equilibrium of the asymmetric
information game exceeds that under full information and that both exporting and importing
nations are worse off than under complete information. We also show that if the policy-active
government uses a quota the resulting outcome is the same as under full information. Thus,

3