摘要：The fact that trade, and especially agricultural trade, remains far from free is puzzling, since the superiority of free trade receives overwhelming support among economists (Gawande and Krishna, 2003; Rodrik, 1994). International trade scholars have devoted significant resources to forecasting the impact of trade policy reforms (for a survey of these studies see Congressional Budget Office, 2005), and their findings are relevant inputs for officials negotiating a trade agreement as well as for politicians in charge of ratifying it. The perceived gains and losses attached to different outcomes have been identified as a significant factor shifting the outcome of a trade agreement negotiation (McMillan, 1990). Less well-known are the factors that would impact the result of a negotiation round involving agriculture. While political scientists have devoted significant resources and developed numerous theories to explain the pattern of bargaining outcomes, the number of empirical assessments remains limited. Among the factors to blame for the low number of empirical evaluations is the scarcity of information regarding negotiation strategies and outcomes (Odell, 2002). The lack of research in this area is puzzling, since countries negotiating regionally and multilaterally have many times struggled to satisfactorily conclude negotiations due to a lack of consensus on agriculture. The potential gains that a reduction in agricultural protection worldwide could generate, and the limited knowledge on the ways negotiation outcomes can be affected, are the main motivations for this study. As stated by Hoekman and Anderson (1999) in reference to reaching an agreement on agriculture in the current WTO Doha Round of negotiations, “The potential gains from liberalization are large”, but “Getting there is another matter” (p. 175). This study focuses on U.S. trade negotiations with Western Hemisphere countries, with particular attention to sensitive agricultural products, about which the most heated bargaining occurs. Arguably as a consequence of the lack of progress at the multilateral arena, the U.S. has embarked on bilateral trade negotiations with several countries, but with a particular emphasis with Western Hemisphere partners. Thus, over the last 10 years it has ratified trade agreements with Chile, Peru, Costa Rica, Dominican Republic, El Salvador, Guatemala, Honduras, and Nicaragua (the last six as part of the DR-CAFTA Agreement), and signed but not yet ratified agreements with Colombia and Panama. Following Trefler (1993), a simultaneous equation model is developed to estimate the correlation between the change in market access (ΔMA) as a result of free trade negotiations, and a number of political economy (PA) and negotiation context (NC) variables suggested by the endogenous tariff theory to be associated with the level of protection granted to a sector. The change in market access in the context of this study is represented by changes in import tariffs and tariff-rate-quotas negotiated as part of the preferential trade agreements (PTAs). While we acknowledge the increasing importance of non-tariff barriers (other than TRQs) as a way to afford protection from imports, PTAs usually do not include specific outcomes on NTBs, and therefore gauging the change in NTBs resulting from a PTA is not possible. Negotiation process variables (NP) primarily bargaining strategies, are seen by many scholars as important intervening variables to explain the pattern of market access achieved in bargaining. By focusing on sensitive agricultural products, this study controls for (NP) variables, since negotiations on sensitive agricultural products between the U.S. and its developing Western Hemisphere partners are assumed to have entailed the use of strong bargaining strategies on both sides. Political economy variables subject to be used are economic size, concentration ratios, import penetration ratios and their change, capital/labor ratios, and sectoral to total labor ratios, all argued to be positively related to the level of protection granted to a sector. Among feasible negotiation context variables are relative power of nations, foreign direct investment in the partner’s economy, unemployment rate, and tax structure (dependence on import policies as a source of revenues). The level of imports (IMP) is modeled to be a function of comparative advantage variables (CA) and the change in market access ΔMA. ΔMA = f(PA, NC NP) IMP = f(CA, ΔMA) The findings from the quantitative assessment are to be complemented with findings from selected case studies, thus expanding our understanding of the factors affecting the outcome of the bargaining activity over sensitive agricultural products in the context of preferential trade agreements.
