2. The Assessment Methodology

2.1. Literature Review of Economic Assessment of Preferential Trade Agreements

The formal analysis of the economic impact of preferential trade agreements began with the seminal work of Jacob Viner more than a half century ago (Viner, 1950). Unlike the traditional intuition of trade liberalization being always beneficial, he argued that preferential trade agreements could be beneficial or harmful to participating countries because the preferential nature of these trade deals generates both trade creation and trade diversion effects. In Viner's view, the overall welfare gains in the signatory countries therefore depend on the extent of trade creation relative to trade diversion.

In Viner's spirit, the economic effects of preferential trade agreements can be summarized on the export and import sides (WTO, 2011). There are two effects on the export side:

  • Improved market access – exporters in FTA partner countries gain market access to each other when tariffs in both home and partner countries are removed or reduced.
  • Preferential advantages – the agreement gives exporters in FTA partner countries preferential advantages relative to imports from third countries, which would otherwise not exist if liberalization were carried out in a non-discriminatory multilateral fashion. However, the preferential advantages would be eroded if the partner country formed a new preferential agreement with a third country.

On the import side, the effects are ambiguous:

  • Trade creation effect: This refers to the increase in imports from partner countries induced by liberalization to displace high-cost and less efficient domestic production. This type of increase in imports is economic-welfare improving. Consider a case in which home country signs a preferential free trade agreement with a partner country (See Figure 1). In such situation, imports from the partner country are no longer subject to tariffs, and the domestic price of the good falls to P. At this price, home country will import from the partner country the quantity Q2 - Q1. Assume that home country concludes a preferential trade agreement with the most efficient producer; the agreement results in pure trade creation. The gains from trade creation are measured by the shaded triangles "b", which represents gains in producer surplus, and "d", which represents gains in consumer surplus. Government loses all tariff revenues that were previously collected on imports of the product (depicted as the area "c"). Thus, the overall net effect of the preferential trade agreement for national economic welfare is positive with a gain of "b + d".

    Figure 1: Effects of Tariff Reductions on Economic Welfare

    Graph of supply and demand: x-axis is quantity and y-axis is price. Domestic supply curve is straight line sloping upward and domestic demand curve is straight line sloping downward. World supply curve is a straight line parallel to x-axis and crosses y-axis at price level upper case P which is the world price level. Upper case Q subscript one shows the amount of import supplied and upper case Q subscript two shows the amount of import demanded. Domestic government imposes an import tariff and d

  • Trade diversion effect: This refers to imports diverted to a less efficient partner country from an efficient supplier due to the preferential arrangements, which is welfare decreasing. This effect would not take place if tariff liberalization were carried out in a non-discriminatory fashion.

The effects of preferential trade agreements are not limited to trade creation and diversion as defined in Viner's premise. Subsequent research in the context of the "new trade theory" in the 1980s-1990s and the "new new trade theory" in the 2000s has shed new light on the sources of gains from trade liberalization.

The "new trade theory" developed in the 1980s attempted to explain rising intra-industry trade in differentiated products among countries at similar income levels on the basis of love of variety by consumers and product differentiation by firms operating under conditions of monopolistic competition and facing increasing returns to scale Footnote 1. According to this theory, consumers' preference for variety and their willingness to pay premium for varieties is the key driver for trade in differentiated products between countries. Producers invest in developing niche products in response to consumer's desire for variety and in doing so manage to obtain monopoly profits in the niche market. However, as the market fragments into niche products, producers struggle to attain the scale of production necessary to survive. A larger marketplace can be created by liberalization through preferential trade arrangements and this allows producers to operate at a larger scale for each niche product. Trade liberalization also erodes incumbent producers' market power as it brings foreign competition into the market. The outcomes of trade liberalization are more varieties at lower prices for consumers and a larger market for producers.

The more recently emerged "new new trade theory" places its emphasis on the central role of firms in international trade Footnote 2. It singles out the "reallocation effect" at the firm level as a new source of gains from trade. This is based on the observation that firms are very different even within narrowly defined industries. Some firms can be larger, more productive or more profitable than others. When trade is liberalized, the most productive firms thrive and expand into foreign markets, while least productive firms shrink and even exit the market when facing foreign competition. As a result, average productivity in the industry increases because market shares and resources are reallocated from less-efficient firms to more-efficient firms. This generates a new source of gains from trade, in addition to the gains from comparative advantage/specialization and product varieties/increasing returns identified in the traditional and "new" trade theories.

Before the arrival of the "new trade theory", the effect of trade liberalization was generally conceptualized in terms of the expansion of the volume of trade in products that were already being traded - indeed, insofar as trade impacted on varieties, the theory of comparative advantage implied a narrowing of production palettes as each country focused on producing and exporting the things it was most efficient in producing. The possibility of a change in product mixes, in particular, entries of new products, new markets and new firms in freer trade was simply overlooked. It is now well recognized that reductions in trade costs through tariff reductions and trade facilitation lead not just to the expansion of trade flows for existing products (the so-called "intensive margin of trade") but to the creation of new trade due to the entry of new firms and new products ("extensive margin of trade").

The extensive margin effect emerges clearly under "new new trade theory" because it explicitly takes into account the fact that firms must incur up-front fixed and sunk costs Footnote 3 to establish themselves in international markets. Many firms, including some productive ones, choose not to export because they are uncertain about their ability to export enough to foreign markets to recover these large costs. PTAs facilitate the participation of these firms in international markets by lowering entry thresholds, including importantly by reducing uncertainty about market access through the various non-tariff provisions that PTAs include.

The expansion of extensive margins is particularly important to a country's economic welfare. If a country, particularly a large one, intensively exports more of each existing product, the prices of its product could be lowered in the world market, resulting in a negative terms-of-trade effect. In contrast, if it exports a broad spectrum of differentiated products at higher prices, though not so intensively for each product, it may achieve greater welfare gains. Thus, Romer (1994) argues that the welfare gains of tariff liberalization can be larger when the gains at the extensive margins of trade dominate as compared to the case when only trade in existing varieties is considered.

2.2 Empirical Evidence

Trade Creation and Trade Diversion

Much of the empirical work on the economic impact of trade agreements has been conducted using the gravity model of trade. Without embarking on complicated welfare calculations, this body of literature attempts to address a simple question: whether preferential trade agreements generate additional trade between partner countries by controlling for other factors that also affect trade flows such as size of economy, geographic proximity, similarities in cultures and languages, income growth and others. Gravity models tend to show large trade creation effects. For example, Jeffrey Bergstrand and Scott Baier, working with a dataset of 96 countries, show that on average, trade agreements double trade between partner countries after 10 years.

On the contrary, the trade diversion effects of trade agreements have been empirically shown to be less significant than expected. A recent assessment of trade creation and diversion effects of preferential trade agreements suggests that trade diversion may play a role for some agreements and for some sectors, but it does not emerge as a key effect (Freund and Ornelas, 2010). Studies focusing on the Canada-United States free trade agreement (CUSFTA) also fail to find significant trade diversion effects. Clausing (2001) finds that the CUSFTA increased U.S. imports from Canada, but did not divert U.S. importing from other U.S. trading partners. The CUSFTA study by Trefler (2004) also concurs that the trade diversion effect under CUSFTA was negligible.

Extensive Margin Impacts

Many of the ex post trade agreement assessments focus only on the intensive margin, ignoring the effects of the increased number of traded products or firms in the affected economies. However, a growing body of literature argues that expansion in extensive margins is more important than expansion in intensive margins for welfare improvement. For instance, Kehoe and Kim (2009) find significant evidence of expansion in extensive margins following a decrease in trade barriers in the wake of major trade agreements. The set of previously least-traded products which accounted for only 10 percent of trade before trade liberalization grew to account for 30 percent of trade or more following the liberalization.

Non-tariff effects

A shortcoming in many ex ante trade agreement assessments conducted using simulation models is that they are restricted to assessing the impact of the eliminations of tariffs only. This approach captures the effect of the agreements to a certain extent, but likely underestimates the potential gains from liberalization for the following reasons (Ciuriak, 2007):

  1. Today's "new age" FTAs often go beyond the traditional FTA approach to include services, investment, customs co-operation, facilitation, and other areas of cooperation. Assessment that focuses on the price effect from tariff reductions does not fully capture the broader implications of the economic cooperation agreement. This is particularly true in the era of the growing importance of investment and rising trade in intermediate goods and services, measures to facilitate investment and liberalize services trade in an economic partnership agreement are expected to have a significant effect on two-way trade in goods over and beyond the effect induced by lower tariffs.
  2. An FTA provides greater certainty about market access in partner countries. In the presence of sunk costs, greater certainty leads to a reduction of perceived business risk, and will increase the expected returns on the commercial presence established in the partner country.
  3. The conclusion of an agreement acts like a "wake-up" call to the private sector, drawing attention to the new business possibilities offered by the agreement (these are sometimes referred to as "announcement" or "animal spirits" effects).

Related to these arguments, Baldwin (2012) describes the deeper commitments under modern PTAs as filling the "supply chain governance gap"; these commitments restrict backsliding on contractual commitments and thus address the "hold-up" problem in international sourcing. These commitments expand trade not necessarily by reducing tariffs but by reducing uncertainty. Footnote 4

Empirical literature based on the "new new trade theory" is still evolving. Most of this body of literature focus on producers' productivity gains stemming from the reallocation of resources from less-efficient firms to more-efficient firms within the same industry in freer trade (Melitz and Trefler 2012), rather than providing a comprehensive picture of overall welfare gains. In this regard, a recent contribution by Arkolakis, Costinot, and Rodriguez-Clare (2012) shows that for a wide class of quantitative trade models, including models developed in the spirit of both "new trade theory" and "new new trade theory", the welfare gains from trade can be calculated using just two important variables: (i) share of expenditure on domestic goods; and (ii) elasticity of imports with respect to variable trade costs or "trade elasticity". This approach significantly simplifies welfare calculations on the effects of trade agreements, as it reflects the "all-in" effects of the agreements, both tariff and non-tariff.

Taking into account the theoretical and empirical developments in the assessment literature summarized above, the assessment of the CCFTA will focus on the following three issues:

  1. the extent to which the CCFTA increased bilateral trade between Canada and Chile;
  2. extensive and intensive margin effects of the CCFTA, and
  3. the overall welfare gains of the CCFTA.


Footnote 1

The seminal papers in the development of "new trade theory" are Krugman (1979, 1980, 1981); Dixit and Norman

Return to footnote 1 referrer

Footnote 2

Melitz (2003) and Bernard, Eaton, Jensen and Kortum (2003) are the seminal papers in "new new trade theory", with the Melitz model serving as the framework of choice for subsequent empirical research. Tybout (2003) provides a survey of the extensive firm-level empirical work that provided the underpinnings for the development of the firm-based trade theories. Redding (2011) and Melitz and Trefler (2012) provide recent surveys of this literature. Ciuriak et al. (2011) draw out the policy implications

Return to footnote 2 referrer

Footnote 3

Melitz (2003) and Bernard, Eaton, Jensen and Kortum (2003) are the seminal papers in "new new trade theory", with the Melitz model serving as the framework of choice for subsequent empirical research. Tybout (2003) provides a survey of the extensive firm-level empirical work that provided the underpinnings for the development of the firm-based trade theories. Redding (2011) and Melitz and Trefler (2012) provide recent surveys of this literature. Ciuriak et al. (2011) draw out the policy implications

Return to footnote 3 referrer

Footnote 4

See also Antràs and Helpman (2004, 2008) who provide a theoretical treatment of these issues in the context of "new new trade theory". They adapt and embed the Antràs (2003) model that features an incomplete-contracting, property-rights theory of the boundaries of the firm into a Melitz-type model.

Return to footnote 4 referrer