Analyses of EU RoHS/ELV Directives Based on an AGE Model with Melitz-type Trade Specification

Modeling non-tariff barriers (NTBs) has long been a challenge for builders of applied general equilibrium (AGE) models, since NTBs are not straightforwardly connectable to economic variables included in a model unlike taxes or tariffs, in addition to the fact that information on NTBs is not easy to collect, sort out complication, or quantitatively evaluate. Non-tariff measures are introduced in order not only to protect local industries, but also to regulate the domestic market. In consequence, NTBs may generate different kinds of economic effects, i.e., protection effects as well as supply- and demand-shifting effects (Fugazza and Maur (2008)). Protection effects may be generated by measures which restrict trade raising cost. Supply-shifting effects may be caused by regulations which specify and affect production processes, such that prevent the sales of hazardous products and create standards to increase compatibility and interoperability. Demand-shifting effects may be brought by rules which affect consumers' behaviors, such that obligate to provide certain information related to the sold commodity. Protection effects can be assessed by two different approaches. One approach uses ad-valorem equivalent (AVE) estimates of NTBs based on the difference between the world price and the domestic price in the importing or exporting country, which has been widely adopted by previous AGE analyses, such as Andriamananjara, Ferrantino and Tsigas (2003) and Fugazza and Maur (2008). Another one focuses on the additional cost of production that firms have to bear in order to export to a specific foreign market. This kind of cost is considered by the seminal work of Melitz (2003) where intra-industry resource allocation among heterogeneous firms plays an important role. The purpose of this study is to show the usefulness of an AGE model with the Melitz-type trade specification in assessing the impact of technical regulations, taking the case of European Union (EU) Restriction of Hazardous Substances/End of Life Vehicles (RoHS/ELV) directives. In addition, we also explore cases when demand-shifting effects incur, changing the importer's preference on variety (LoV). The AGE model used in this study is calibrated to the GTAP 8.1 database for 2007. The original 129 countries/regions and 57 commodities/activities are respectively aggregated to six and five. The regions consist of the European Union (r01), United States of America (r02), Japan (r03), China (r04), ASEAN (r05), and Rest of the World (r06). The five sectors are the Primary Industry (i01), Services (i02), Motor Vehicles & Parts (i03), Electronic Equipment (i04), and Other Manufacturing (i05). Sectors i03 through i05 are assumed to be imperfectly competitive with increasing returns to scale (IRTS), while the other two are characterized by constant returns to scale (CRTS). Sector i01 uses a sector specific factor, such as land and natural resources, in addition to capital, labor, and intermediate goods in its production process, while Sector i02 provides a fraction of its output as the inter-regional shipping supply. Some of the parameters and exogenous variables are determined by the author based on the empirical studies such as done by Ardelean (2006) and Melitz and Redding (2013). To include the Melitz-type trade specification, only the Pareto shape parameter is required as addition information, since the choice of initial numbers of firms or levels of fixed costs will not affect simulation results expressed as deviations from the baseline (Oyamada (2014). Simulations with a special focus on the strength of the importer's love of variety (LoV), the key findings can be summarized as follows: (A) raising the fixed cost to make sales in the EU market brings reasonable results that exports of the targeted commodities (i03 for ELV and i04 for RoHS) to EU from outside regions shrink while the intra-regional trade within EU expands when the LoV is not so strong as Ardelean (2006) suggests; (B) when LoV is strong as assumed in the theoretical model by Melitz (2003), totally opposite embarrassing results are obtained; (C) when the strength of LoV is gradually changed from zero to unity, there is a point around 0.8 where effects of increasing the fixed export cost reverse; (D) consequently, demand-shifting effects incur in the direction to moderate the total impact when the strength of LoV is not so strong.