Since the inception of the World Trade Organization (WTO) in 1995, member countries have been heavily relying on the organization's dispute settlement procedure (DSP). Exploiting a new database on WTO litigations between 1995 and 2014, this paper describes disputes initiated over this period and identifies potential sources of bias concerning the participation of developing countries. The analysis builds on three different models to determine country i's probability of initiating a dispute against country j. Either it depends only on the two countries' structure of trade, that is the number of products exported by i to j (a situation we refer to as the rules-based model), or it is also affected by country i's or country j's specific characteristics (the unilateral power-based model), or it is also affected by bilateral economic and trade relations between countries i and j (the bilateral power-based model). We find that country i's structure of trade with j plays an important role in explaining the probability that i initiates a dispute against j under the DSP. Furthermore, country i's legal capacity and both countries' political regimes also affect this probability. However, we do not find that bilateral relationships between i and j, such as participants' capacity to retaliate against each others have an impact on dispute initiation.

Does smuggling improve economic welfare? This paper provides a theoretical model of illegal trade and determines the impact of smuggling on economic welfare. We focus on Costa Rica's recent prohibition of avocados imported from Mexico. Using unique data on trade, production, and the price of Costa Rican and Mexican avocados, we find that the quantity of avocados smuggled into Costa Rica on an annual basis ranges from 4,668 to 10,232 metric tons, representing up to four times the quantity of locally produced avocados. Furthermore, we demonstrate that smuggling is necessarily welfare-improving compared to the ``no-smuggling situation''. Compared to the ``free-trade situation'', smuggling does not always compensate for the negative effects arising from the restrictive trade measure. In most cases, welfare is lower after the prohibition and smuggling than prior to the implementation of the prohibition. In some cases, however, smuggling results in a gain in terms of trade that offsets the harmful effects associated with the trading cost of smuggling. We find that such situations occur when the trading cost of smuggling is low, and thus when enforcement of the prohibition by public authorities is weak.