A Commitment Theory of Subsidy Agreements

Maggi and Rodriguez-Clare (2007) develop a model that combines the “terms-of-trade” rationale for trade agreements with the commitment approach and formally show the complementary nature of the two theories.

Staiger and Tabellini (1999) provide evidence that GATT/WTO rules have helped the US government to make trade policy commitments to its private sector. More recently, Tang and Wei (2009) have found that accession to the GATT/WTO increases the credibility of policy commitments, particularly for countries with poor domestic governance.

WTO (2012) makes this point for non-tariff measures in general, and provides preliminary evidence that governments tend to use these tools in sectors with more stringent (applied) tariff. In a book on the treatment of subsidies in the multilateral trading system, Gary Hufbauer makes the following case for international disciplines on subsidies (quoted in Sykes 2010): “Unbridled and competing national subsidies can undermine world prosperity … Because the concentrated interests of producers command greater political support than the diffuse interests of consumers, national governments find it much easier to emulate the vices of protection than the virtues of free trade. This lesson has prompted the international community to fashion guidelines that distinguish between acceptable and unacceptable national subsidy measures and to codify these guidelines both in bilateral and multilateral agreements”.

Other policies include different forms of subsidies and government transfers, non-tariff barriers (e.g. protectionist sector-specific regulations), contingent measures (e.g. anti-dumping), etc. While this paper focuses on production subsidies, the logic applies to other measures as well. We come back to this point in the conclusions.

For a simple model that captures this idea, see Brou and Ruta (2009).

The model that we employ is more general than the basic MRC setting, as it does not rely on specific functional forms and allows for positive tariff bindings.

See Bacchetta and Ruta (2011) for a collection of key contributions on subsidies and the WTO.

Early works, in the context of trade agreements, include Copeland (1990) and Bagwell and Staiger (2001b).

The choice of this constant returns to scale technology allows for two important features. First, the fixed input (land) allows for diminishing returns to capital, which is important for the equilibrium allocation of capital between sectors. Second, production in the numeraire sector picks up any labor not used in the manufacturing sector and simplifies the analysis.

Since the early work of Arnold Harberger in the 1960s, a large body of literature has been devoted to the measurement of dead weight losses of taxation (for a review, see Auerbach and Hines 2002). In a setting similar to ours, Matschke (2008) assumes that collecting domestic taxes is relatively more costly than collecting trade taxes and finds evidence consistent with this assumption for the US. In her model, imposing a tariff is efficient as tariff revenue lowers the cost of domestic taxation. While the results of our paper do not depend on how costly revenue raising is modeled, our formalization ensures that laissez faire is the first-best policy mix (see Lemma 1). This facilitates the comparison of our findings with the rest of the literature on trade agreements.

These assumptions are admittedly ad hoc. They serve the purpose of setting the stage in favor of protection in the manufacturing import-competing sector, which allows focusing on the driving elements of the model: the distortions created by the lobbying process and the substitutability between the import tariff and the domestic subsidy.

Note that in the Grossman and Helpman (1994) approach adopted in this model, contributions are equally valued by the government and the lobby. While this assumption is reasonable in certain contexts (most notably if contributions consist of cash transfers and there are no limits on lobbying payments), there are policy environments where the non-transferability of utility between the government and the lobby may be a more appropriate representation of the reality. See Drazen and Limao (2008) and Limao and Tovar (2011) for a model where contributions do not enter linearly in the government objective function.

In the literature of lobbying and trade agreements, the absence of ex-ante lobbying is a common preliminary assumption which allows the sharpening of the predictions of the theory (see, among others, Maggi and Rodriguez-Clare 1998, Limao and Tovar 2011). However, it is well understood that ex-ante lobbying may be relevant at least for certain industries (Maggi and Rodriguez Clare 2012). We come back to this issue in the Conclusion.

In fact, the first-order conditions imply a negative subsidy as long as domestic production is positive in the absence of tariffs and subsidies, i.e. , so that the non-negativity constraint on subsidies is binding. This possibility is ruled out because a negative subsidy is attractive in this setting only because it would be a way for the government to raise revenues while “avoiding” costly taxation.

In the working paper version of this article, we introduce the possibility of an externality in the production of the manufactured good. In that case, the socially optimal policy will depend on the capital allocation. See Brou and Ruta (2012, Section 5) for further details.

Note that for , the tariff is always greater under the political equilibrium (i.e. ). This is because .

These results generalize the findings in Grossman and Helpman (2001, chapter 7) on lobbying when a government has at its disposal multiple policy tools. As a caveat, let us stress here the role of the assumption on the transferability of utility between the government and the lobby implicit in condition [3]. As shown in Drazen and Limao (2008) and Limao and Tovar (2011), removing this assumption can generate a different political equilibrium, where lobbying distorts inefficient policies (i.e. policies with higher social cost).

Notice from eq. [9] that the marginal benefit of a subsidy is greater as t is lower.

One technical problem is that, when , so that the subsidy is not used under full discretion, a tariff reduction will result in a positive subsidy only if it is sufficiently large (as can be seen from [9]). We focus on this more interesting case in the remainder of the paper.

Our model includes Maggi and Rodriguez-Clare (1998) as a special case, where there is no labor. Without labor, production in the manufacturing sector depends only on capital and cannot react to changes in policy. As such, a subsidy would not serve as a substitute for a tariff.

An equivalent interpretation is that the government can credibly commit to the socially optimal level of (zero) tariff and subsidy. The organized group would have no incentive to lobby in this case.

Technically, this can be seen by maximizing the aggregate welfare function, evaluated at , with respect to capital. Applying Euler’s Theorem yields .

Recall that so that . The first term is negative for any and the second term is positive. Similarly, . The first term is negative for any and the second term is negative.

As discussed in previous sections, the two policy problems are related. Specifically, if the trade agreement does not solve the policy substitution problem, the government cannot credibly commit to its policy.

For an introduction to GATT/WTO rules on subsidies refer to Sykes (2005) and WTO (2006).

It is also true that if the subsidy offered to exporters would cause injury to foreign producers, a trading partner could impose a countervailing duty. Since the focus of our model is on transfers to the importing sector, injury does not apply.

The original SCM Agreement also contained a “green light” category of subsidies that could not be challenged (e.g. subsidies to research activities, assistance to disadvantaged regions, green subsidies). This category offered an exception to subsidy commitments in certain pre-specified sectors, independently of their trade effects. In our working paper, we discuss a justification for this exception based on the existence of production externalities. The provision, however, expired in 2000 and has not since been renewed by WTO members.

Article 6.3(c) may best be thought of in the context of a large economy, whose subsidies change the world price in addition to the domestic price and trade volume. Article 6.3(d) deals with the volume effects of a subsidy in a third-market. The issues contemplated by Articles 6.3(c,d) of the SCM Agreement do not arise in the model developed in the previous sections which assumes a small open economy.

Of the 456 trade disputes listed in the WTO website for the period 1995–February 2013, 97 (that is, 21.3%) dealt with subsidies. Most of the cases that relate to domestic subsidies involved large producers (e.g. US -cotton; China -apparel and textile products), but there are also disputes that are targeted at smaller producers, such as the recent dispute on Indian support for solar cells and solar modules production. Moreover, WTO subsidy rules are likely to discourage the use of domestic subsidies in the first place (see Rodrik 2010), which implies that few disputes are actually observed in equilibrium.

We make a similar argument in the working paper version.

Properly identifying the sectors that deserve an explicit exception and embed such loopholes in the multilateral trade system may not be obvious. As noted by Sykes (2005), finding the line that divides a legitimate domestic subsidy from a measure that benefits the interest of an organized group at the expense of society is not always a straightforward matter.

Within the commitment approach, there are other explanations of why various non-tariff measures receive different treatment in the WTO system. Specifically, Limao and Tovar (2011) argue that highly inefficient NTMs are poor substitutes for tariffs and, therefore, there is less need for regulation. To the contrary, more efficient NTMs such as subsidies or quotas are more likely to be used to substitute a tariff, which justifies the more stringent constraints in WTO Agreements.

Interestingly, in a system such as the EU, where there are no internal tariffs, exceptions to subsidy rules are common. To the contrary, within the WTO system, which allows for positive tariffs, subsidy exceptions are rare (and currently do not apply). Needless to say that, in the absence of relevant market failures, both higher tariff bindings and exceptions to subsidy rules reduce social welfare.