Economics Letters 45 (1994) 213-216 0165-1765/94/$07.00 0 1994 Elsevier Science B.V. All rights reserved 213 Tariff behavior in countries: Further Alok K. Bohara* zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Department of Economics, University of New Mexico, William H. Kaempfer five European evidence Albuquerque, NM 87131, USA Department of Economics, University of Colorado, Boulder, CO 80309-0256, USA Received 22 July 1993 Accepted 23 November 1993 Abstract Temporary movements in the average tariff rate for a country can lead to intertemporal substitution effects. A five-country model of tariff movements is estimated in a generalized system. The tariff formation process follows a random walk for the United Kingdom, Denmark, Sweden, and France, which implies that tariff changes can be viewed as permanent in those countries. However, in Switzerland the cumulative impulse responses to a tariff shock imply not only shock persistence, but shock magnification. /EL classification: F13 1. Introduction In their recent paper, Gardner et al. (1992) analyzed unit root behavior of average tariff rates for five countries-the United Kingdom (UK), Denmark (DEN), France (FRN), Sweden (SWD), and Switzerland (SWZ). The objective of their analysis was to explore the potential trade balance impacts of intertemporal substitution that tariff movements would entail. Specifically, they posit that if the tariff time series for some country was accurately described by a simple random walk process, then tariff changes would be perceived by rational citizens as permanent. They conclude that, ‘therefore, intertemporal considerations suggest no clear-cut presumption as to the trade balance effects of increases in the . . . tariff rates’ (p. 77). By analyzing each of their five series individually, Gardner et al. conclude that all series contain a unit root. However, the analysis suggests that only the tariff rates for Denmark and the United Kingdom are described by a simple white noise process. Therefore, citizens of France, Sweden, and Switzerland would perceive long-run adjustment to tariff rate changes in those countries and make appropriate intertemporal consumption substitutions. The motivation of this paper is to re-examine the unit root issue by considering all five series simultaneously. It makes both economic and statistical sense to undertake such an analysis. Economically, tariffs are best considered as endogenous policy instruments. Policy-makers react to economic stimuli, like the unemployment rate or the rate of economic growth, in deciding whether * Corresponding author. SSDZ 0165-1765(94)00400-V