The US trade deficit and the belief that the United States may be subject to unfair competition has led to increased political interest in protectionism, fear of which provoked the Plaza and Louvre accords and the Baker proposals for LDC debt. The argument for protectionism is that tariffs lead to higher demand for domestic goods relative to imports, and thus to higher output and an improved current account. But the positive effect on output may be offset by the income effect of higher prices; the effect of a tariff on output can only therefore be determined by an empirical analysis that incorporates numerical values for the substitution and income effects of price changes.
In Discussion Paper No. 286, W Helkie, Research Fellow Andrew Hughes Hallett, Gary Hutson and Jaime Marquez analyse the implications of a policy package consisting of a tariff on US imports and macroeconomic policies to induce dollar depreciation. They use the Federal Reserve Board's Multi-Country Model for the United States, Canada, the United Kingdom, West Germany and Japan over the period 1986-92. The authors use a dynamic game theory setting to capture the repercussions of US trade restrictions on economic activity in the rest of the world and the possibility of foreign retaliation.
The first simulation allows both the United States and the other countries to impose a general import tariff and also to use that tariff as a retaliatory measure. Tariff levels, averaged over the whole period, are found to be quite low 23.3% for the United States and 1.4% for the rest, in addition to any tariffs already in force. This implies that the United States is able to introduce tariffs that are small enough not to trigger effective retaliation, yet sufficient to help clear its trade deficit. A major consequence is the near elimination of Germany's trade surplus and the halving of Japan's, while the UK current account worsens. The strong effects on trade balances arise from the importance in the model of interest rate differentials as the major determinant of capital flows and thus exchange rate adjustments. A comparison of these results with those of a simulation when tariffs are ruled out of policy-makers' options reveals a rather small difference between the outcomes with and without tariffs. The costs of a tariff war are found to be quite low for all G3 countries.
The authors also explore how tariff policy affects the gains from international cooperation over fiscal and monetary policies. Cooperation benefits the United States and Canada more if tariffs are allowed. The possibility that a country may lack sufficient instruments to correct a large trade deficit without damaging itself and its partners suggests that organizations like GATT and the IMF should concentrate first on creating the conditions for cooperation, tolerating some tariffs as a temporary expedient for countries with severe external deficits.
Cooperation and tariffs lead to similar shifts in the policy mix, i.e. fiscal expansions balanced by monetary contractions, but they have very different effects on policy targets. Cooperation increases US growth and reduces growth elsewhere, whereas tariffs reduce growth in all the G3 countries. Cooperation reduces inflation and worsens trade imbalances, while tariffs produce the opposite effects. In all the scenarios it is dollar depreciation, not protectionist measures, that bears most of the burden of adjustment. Under all assumptions about cooperation and protection, the dollar will have to fall to around DM 1.25 or ¥ 95 by 1992 if the US trade deficit is to be eliminated and its budget deficit halved

Protectionism and the US Trade Deficit: An Empirical Analysis
W Helkie, A J Hughes Hallett, G J Hutson and J Marquez

Discussion Paper No. 286, February 1989 (IT)