The anatomy of cooperation: An introduction to regional trade

Regional and Preferential Trade Areas (RTAs & PTAs respectively) are all the rage today, having grown incredibly quickly since the 1990s, which saw prominent agreements like NAFTA (the North American Free Trade Agreement), signed in 1994, and Mercosur, the PTA formed by Argentina, Brazil, Paraguay and Uruguay in 1991. A telling fact is that every country in the world except Mongolia (and some Pacific Islands) is now in some trade agreement or other (the count in July 2013 was 575).

While this implies that India too is active on this front, it barely indicates the number of agreements we are already in or currently negotiating, which total over 30.

This may be puzzling to those who realise that all members of the WTO are committed to an eventual reduction of trade barriers, and who are aware of the crucial Most Favoured Nation (MFN) clause that requires uniformity in treatment of trading partners. Interpreting this to question why PTAs exist at all, and how they affect the multilateral trading process, certainly makes this a valid query and is one I intend to explore in what follows. However, it should be noted that the MFN issue has been accounted for in various WTO articles like Article XXIV and via the ‘Enabling Clause’.

A brief note is apposite here: the literature on PTAs is impossibly vast for me to even attempt to summarise cogently. So, I must provide a selective, highly subjective account that I believe contributes to an understanding of this issue. Finally, for those unfamiliar with the term, ‘terms of trade’ is defined as the ratio of export prices to import prices.

Welfare impact of PTAs

The dominant line of research has focused on preferential liberalisation of trade in goods, e.g. via a reduction in intra-bloc tariffs, while allowing for either autonomy in setting tariffs on non-bloc members (a ‘Preferential Trade Area’) or a common external tariff (a ‘Customs Union’). The question then becomes whether members of a PTA can unambiguously benefit from changing intra-bloc tariff structures. It is quite difficult to obtain an unequivocally positive answer.

It is not too hard to understand why this is the case. Suppose a country enters into a preferential agreement with another. The reduction in tariffs between the new bloc members may lead to a shift in sourcing of goods from inefficient domestic producers, who thrive on protection, to another country in the bloc (referred to as ‘trade creation’). Preferential liberalisation might be inefficient if it diverts trade away from a non-bloc member that has a lower cost of production than the new trade partner (referred to as ‘trade diversion’).

Of course, it is not just differences in trade quantities post tariff change but also the quantum of the tariff change itself, among other things, that determines if welfare increases or not. These factors (determining welfare impact) are numerous and include: elasticity of import demand and export supply (which capture the magnitude of surpluses under the new tariff regime); the size of the union (related to elasticity and whether a country is a price-taker rather than a setter); whether a country’s import demand can be sated by other bloc members or would need to be met partly from outside the bloc (which would result in intra-bloc terms of trade effects, and an effective transfer of tariff revenue to other bloc members who benefit from a shift in production); the degree of substitutability between imported and exported goods (as tariffs fall within the bloc, so does the imported good’s price and hence the domestic demand for goods imported from outside the bloc); market structure, i.e. deviations from the competitive benchmark etc. All these results are derived in models, in varied settings, and have normative value, but the plausibility of assumptions made needs to be tested.

Empirical literature, employing both cross-country and sector-wise approaches, reveals that trade creation effects are quite dominant. Interestingly, if we take the other route, of fixing external trade quantities and allowing external tariffs to vary, a famous result (due to Kemp and Wan) obtains, where preferential liberalisation makes a customs union no worse, and sometimes better off. This is hardly surprising, given that we have switched off the trade diversion effect altogether (which requires external tariffs to decline).

Of course, non-bloc countries could change their tariff structures in response as well, to alter the outcome. This is a relaxation of the ceteris paribus assumption of other things being equal. Bloc members too might change their external tariffs after entering a trade agreement. This differs between PTAs (which have tariff autonomy) and customs unions (where members coordinate tariff policy). So, PTA members might want to counter trade diversion effects by lowering external tariffs.

The extra competition faced from other bloc members’ firms lowers market shares, so raising tariffs could lead to a revenue transfer effect as described above. There is some empirical backing for this line of thought. An interesting argument is that if PTA members exchange non-trade concessions, they might hesitate to lower external tariffs, as the preference then granted to partners would be relatively lower, which would perhaps threaten those concessions.

Reasons for formation

Why would RTAs be signed at all, if welfare might not improve for bloc members, as the theory at least suggests? And why have there been so many RTAs since the 1990s?

One reason could be political economy considerations. Lobbies, representing the interests of producers who face competition from imports, could contribute to policymakers’ parties in an attempt to influence trade policy design. So, if they contribute enough to outweigh welfare losses or gains that entering an FTA would entail, inefficiency could result. If we assume governments look after domestic firms’ interests as captured by their profits, FTAs could result in gains from access to other bloc members’ markets, but losses from applying the corresponding logic to other bloc members. With trade diversion, trade can be captured from non-bloc members as well, favouring FTA formation. So, political economy models suggest that trade-diverting agreements may be more likely to arise.

The revenue transfer effect mentioned above might dilute this effect, as protection rents enjoyed by domestic lobbies will be reduced as well, lowering their incentive to alter policy. It is also quite common for geographically proximate countries to form RTAs, like India does as well (SAFTA, China etc.). Some have argued that countries whose bilateral trade is a significant fraction of their trade totals would form ‘natural blocs’, but others believe these blocs might not improve welfare.

There are also several non-trade explanations behind the proliferation of RTAs. One is the classic ‘time inconsistency’ problem, where members might want to renege on liberalisation promises made, particularly when there are negative shocks to the economy that would tempt the government to provide (unanticipated) support to affected industries, under import substitution demands or in cases like retrospective taxation. An RTA can then serve as a commitment device that ties the member’s hand. The larger the costs of leaving an RTA (e.g. punitive costs, general exit costs), the stronger will this effect be.

A reasonable question to ask is why multilateral accords like the WTO might not do an identical or better job. A plausible explanation could be that the onus of punishing recalcitrant members is better defined in an RTA, where countries facing costs may be better identified, rather than in the larger WTO. Free-rider effects may also be valid here.

Closely linked to this is the role of investment as an incentive. As policymakers are well aware, uncertainty about the future is hardly conducive to receiving more investment. By joining an RTA, a country can signal both its trade outlook and perhaps its attempt to reform domestic economic policy. Further, the preferential treatment afforded to businesses and investors in a PTA will be enticing as well.

There might also be a political time-consistency effect at work here. The (liberal) party that enters an RTA can tie the hand of its possibly protectionist rival on the issue of liberalisation, particularly if the costs of exiting are high, and also if voters are informed. There might also be a signalling explanation, so that what matters in a situation of uncertainty regarding policy outlook are not just the provisions of the agreement but the decision to enter it in the first place, especially if it is costly to do so. For example, a fledgling democracy might want to signal its liberalising intent by joining an RTA, which, might not always benefit it, perhaps to attract investment.

An explanation that seems quite relevant is the role of customs unions (which are more cohesive) in increasing the bargaining power of a bloc, by making the ‘outside option’ less damaging. It might also jolt the multilateral system into more conclusive action, as perhaps occurred in 1994 with NAFTA and GATT.

There is an insurance motivation as well, so countries can deal with adverse shocks through slightly protectionist measures without inviting reprisals. Finally, it seems quite likely that RTAs also have a strategic role, in bolstering ties between possibly disputing countries and making adverse reactions less likely.

RTAs and Multilateralism

A concern has been whether RTAs might impede progress in multilateral accords. This has been explored in political economy models, with firm profit concerns affecting policy. In that case, opening up further would erode rents enjoyed by domestic firms, and they would lobby against such an action. As mentioned earlier, the rent destruction effect might mean the lobbying channel is not too active in the first place, which would naturally affect this conclusion. RTAs can sometimes block feasible multilateral agreements.

One of the benefits of access to a larger market is greater product diversity for consumers. This effect is stronger in a multilateral system than in an RTA, but the multilateral agreement could result in adverse distributional effects (through the endowments of factors of production) that might overwhelm the beneficial effect of diversity. Regionalism might even create its own demand, if countries invest in partner-specific areas that were not optimal prior to the agreement, but become so afterwards, lowering incentives for a multilateral thrust.

Sometimes customs unions might serve as a bargaining tool to propel multilateral agreements forward, as discussed above. This might be because they tend to have a higher external tariff, so the cost of not reaching a multilateral agreement is larger.

Another factor is openness. If membership is open, outsiders who are worse off from the status-quo RTA would prefer to join, leading to a gradual liberalisation wave. The MFN clause too has a role to play, as it encourages free riding and weakened impetus toward liberalisation on the part of countries. Allowing for preferential treatment could then make countries take the necessary dose. Finally, PTAs can be a benign force towards multilateral integration in a scenario where political factors make multilateral accords difficult to arrive at. The actions of a few members who form the PTA might have spillover effects on the provisions of multilateral agreements by defining a more feasible boundary.

One could look at the other side of the coin, i.e. how multilateralism affects regionalism. This strand of research examines the impact of multilateral liberalisation on the stability of PTAs. Under high tariffs, multilateral agreements will be preferred to PTAs as outside exports are likely to be very small, and so are terms of trade effects under a PTA, whereas benefits from non-bloc liberalization would be present. However, if initial tariffs are lower, PTAs are preferred to multilateral agreements as the former can take advantage of terms of trade benefits that are absent in a scenario where every country liberalises.

Another channel is through progressive liberalisation in developed countries (the ‘North’), which would lead to greater economic integration in the North, and a reduction in external tariffs. This in turn would help bring some industries to less developed countries (the ‘South’), making North-South reciprocal trade agreements more likely.

The empirical evidence on regionalism and multilateralism suggests a positive correlation of incidences, i.e. more RTAs are signed during multilateral negotiations than at other times. Furthermore, a study of the impact of US behaviour during the Tokyo and Uruguay GATT rounds on its subsequent RTA policy reveals a systematic complementarity, with items discussed and liberalised under GATT receiving the same treatment in RTAs.

The Indian experience

India has been particularly prolific in the PTA domain. Many of the reasons for RTA formation could apply, from possible assessments suggesting high trade creation effects to non-trade explanations like enhanced bargaining power and signalling, and even strategic motives. A major explanation has been the boost such agreements could provide to domestic exporters, who often benefit from preferential access.

However, it seems more likely that partner countries would benefit from the opening up of the domestic economy, indicating the importance of trade diversion in any analysis. The low share of India and some of its partners, like those in the SAFTA, in world trade alongside the considerable protection already present in these countries has led Arvind Panagariya and other analysts to conclude that trade diversion is the likely outcome, and that such agreements would do little to enhance bargaining power unless they incorporated more partners and accounted for a larger world share of trade.

Of course, most of the analysis has been couched in terms of the familiar tariff and welfare analysis. However, other knotty issues like labour mobility and services trade too play an important role. A critical factor not discussed yet is Rules of Origin (ROOs), which are complex, bureaucratic, often ad hoc regulations that prevent PTAs with differing external tariffs from importing and reselling within the bloc (‘transshipment’), often through a domestic content requirement. It is widely accepted that these are immensely distorting, as they force firms to alter production structure to demonstrate compliance and entail bureaucratic costs. The scepticism regarding benefits to India from FTAs has in certain cases been attributed to ROOs, like in the India-Japan CEPA. A major critique of PTAs launched by Jagdish Bhagwati is that the same intermediate good often receives diverse tariff treatments, leading to a mad race around bloc member countries to export the finished good at the lowest price, which he terms the ‘Spaghetti Bowl Effect’ (from the cat’s cradle like shape of trade flows).

The Economic survey of 2012-13 has argued that diversifying trade baskets in RTAs to include technologically intensive goods in this era of sluggish growth, along with a need to eliminate the inverted duty structure (where intermediate goods are taxed more than finished products) makes some sectors uncompetitive. It remains to be explored whether PTAs serve as a ‘building bloc’ that facilitates liberalisation or a ‘stumbling bloc’ that hinders it, as well as whether India is pursuing a judicious strategy by entering into so many agreements.

References for the curious reader

In the interest of brevity, this article has just skimmed through the various strands of literature. I strongly believe a clearer understanding can be gained from a diagrammatic exposition of welfare gains or losses. To this end, comprehensive accounts can be found in three review articles:

The first is by Caroline Freund and Emanuel Ornelas in the Annual Review of Economics (2), 2010, titled ‘Regional Trade Agreements’. Relatively technical accounts, nonetheless excellent, are ‘Regional Economic Integration’ by Richard Baldwin and Anthony Venables in the Handbook of International Economics, Vol. 3, 1995; and ‘Preferential Trade Liberalisation’ by Arvind Panagariya in the Journal of Economic Literature, 2000. The last two also chart the historical progress of the sub-discipline. A fine account of non-trade explanations for RTAs is by Raquel Fernandez and Jonathan Portes in the World Bank Economic Review, 1998. The Indian experience section has relied on newspaper reports, economic surveys of the Ministry of Finance, and articles by Panagariya and co-authors, and T .N. Srinivasan and co-authors. An updated list of India’s PTAs is available on the Department of Commerce website.
 

Lalit Contractor has a MPhil in Economics from Oxford University and is curious about Economics and its interactions with politics and society