When developing domestic policies to meet their emissions limitation commitments under the Kyoto Protocol, some Annex I Parties may wish, or be under pressure, to impose less stringent obligations on some industries to improve their competitiveness. The sensitivity of industry location to the stringency of environmental regulation is called ecological dumping. International co-ordination of environmental policies may be needed to reach an economically efficient outcome in which it is impossible to make one country better off without making at least one other country worse off.
Under certain ideal conditions (e.g., perfect competition in all markets) there is theoretically no need for international policy co-ordination (Oates and Schwab, 1988). However, such conditions do not hold if there is imperfect competition in goods markets or unemployment (Rauscher, 1991, 1994; Barrett, 1994; Kennedy, 1994; A. Ulph, 1994; D. Ulph, 1994). If, and to what extent, international differences in environmental regulation have trade or even relocation implications obviously depends on a host of factors. These include country size, availability of alternatives, relative resource endowment, mobility of production factors, competition level, scope for innovation, possibility of border-tax adjustment, chances of retaliation, and redistribution of environmental tax revenues (OECD, 1996a).
Although it is clear that many factors affect the relationship between the stringency of pollution control policies (if implemented unilaterally) and net exports, some authors have carried out rather straightforward empirical tests on the relationship between the two variables. Han and Braden (1996) examined 19 US manufacturing industries between 1973 and 1990 with the help of regression analysis. They found the relationship between pollution abatement costs and net exports to be negative in most of the sample period, but diminishing over time (with elasticities close to zero in many industries). Van Beers and Van den Bergh (1997), using a gravity model of international trade and two measures of environmental stringency, did not find a significant relationship between environmental stringency and total exports for the dirty industries. However, when they focused on the non-resource based, and therefore more mobile, industries only this relationship was significant.
Early empirical research on the impact of environmental policy on trade found little evidence of a measurable relationship, partly because of low environmental taxes and partly through data and statistical limitations. Therefore, many studies have concentrated on simulations of environmental tax regimes. From a survey of these studies, IPCCs SAR (IPCC, 1996) concluded that estimates of the effects of environmental policies (notably carbon taxes) on trade vary wildly, depending on model parameters (such as energy demand elasticities and assumptions regarding the substitutability of traded goods) and the policy scenario examined (extent of reduction in emissions and extent of international co-ordination).
Various partial equilibrium models have been designed to analyze ecological dumping, many using static or dynamic game theory. Early analyses used a Cournot setting, which models long-run competition among firms as a series of strategic capacity or output choices. The general conclusion from these early models is that the optimal tax (or any comparable domestic environmental policy instrument) would be set below marginal damage. As a consequence, environmental policies are designed to try to protect domestic industries. If producers collude, however, the incentive for governments to engage in ecological dumping is reduced (Ulph, 1993).
The ecological dumping conclusion could change completely if governments act strategically in setting taxes, and if there is Bertrand competition (firms compete by choosing the price to charge, rather than the quantity to produce) instead of Cournot competition (Eaton and Grossman, 1986; Barrett, 1994; Conrad, 1996; Ulph, 1996). If, however, producers act strategically or can collude, then the outcome in terms of ecological dumping is not straightforward. Quantity-based environmental regulation, if implemented unilaterally in a duopolistic case with a domestic and foreign supplier, might actually benefit domestic firms at the cost of domestic consumers (Kooiman, 1998). If both governments and producers act strategically, again, the incentive for governments to distort the environmental policy is less than when only governments acted strategically, so that the Bertrand outcome can be similar to the Cournot outcome (Ulph, 1996).
Ecological dumping also has been analyzed with the help of general equilibrium models of international trade involving externalities (Rauscher, 1994). It was shown that in a second-best world for several market structuresmonopoly power of the exposed sector or oligopoly on an outside market (Elbers and Withagen, 1999)ecological dumping might not (always) be beneficial from a welfare point of view. This is contrary to the conclusions of some of the earlier partial equilibrium models.
The most interesting case for analyzing policy co-ordination needs is that in which national commitments have been decided internationally, but individual Parties may, but need not, co-ordinate their national policies to fulfil their commitments. This would be the Kyoto Protocol case, after ratification. Hoel (1997) has addressed this case and argues that governments may tend to subsidize indirectly particular imperfectly competitive industries selling on the international market. To prevent this from happening, an argument can be made in favour of policy co-ordination, which is possible but not required in the Kyoto Protocol, except insofar as the Kyoto mechanisms are concerned.101
Other reports in this collection