A Custom Union is a free trade area -from which trade impediments have been removed-to which a Common External Tariff is added. This paper argues that a country can in certain circumstances lose from joining such a CU. Therefore it first shows that a country can suffer from joining a free trade area. Secondly it moves on to find a practical criteria enabling one to assess the likelihood of a positive welfare effect. Finally it addresses  the question of whether a country can lose from joining a system of Common External Tariffs. The aim of this paper is to find out the conditions under which a country may suffer from joining a Custom Union, given its twofold nature.

 

 

Let us consider a small country which is about to join a Free Trade Area. This country has two potential sources of imports : the country which is about to become its free trade partner (A) and the rest of the World (W) which is the cheapest import supplier (Pa>Pw). Let us suppose that initially the home country imposes a trade barrier that raises the local price of all imports by T. Since the country is small, it faces flat import supply curves. We assume that the price charged by W's exporters is lower than the one charged by A's exporters. Competition from imports fixes the price of local good at Pw+T (since Pw+T<Pa+T), and imports at Q3-Q2.

 

 

                 Price                                                  Home supply

 

 

                 Pa+T

                  Pw+T

                                      A         B    C                 D                 E

     T                              Pa

      Pw                                                  F

                                                                                                   Home demand

 

                                              Q1       Q2                            Q3             Q4            Quantity

Next, suppose that the home country joins a free trade area that encompasses A. This discriminatory liberalisation artificially changes the relative competitiveness of goods from A and W. Now goods from A cost Pa, while those from W still cost Pw+T. Naturally home consumers will divert all of their import demand from W towards A. Let us hold a welfare analysis.

Whether such a policy is harming or not for the home country, depends on the nature of the barrier T. Following E.Baldwin, we have to distinguish between rent-creating barriers and cost-creating barriers. Rent-creating barriers are those which entail a revenue for either the importing government (a tariff) or the exporting government (Voluntary Export Restraint). Cost creating barrier (bureaucratic delays, standards, and custom procedures) do not generate any revenue for anybody.

Let us consider the case whereby the increase in import price (T) corresponds to a rent-creating trade barrier captured by the home government - typically a tariff. In this case, the government loses all tariff revenue due to the diversion of trade from W to A. This loss equals D+F.

On the other hand, the lower price benefits home consumers by an amount equal to A+B+C+D+E, while domestic producers are harmed by A+B. The net impact on the home country's welfare equals the two triangles (C+E) minus the rectangle F, that is C+E-F.

Traditional analysis classifies these welfare effects into 'trade creation' and 'trade diversion'. The cost of trade diversion, area F, stems from the fact that after the discriminatory liberalisation (trade liberalisation with A only), the country buys from a higher-cost international supplier. The trade creation gains (C+E) are due to a more efficient allocation of resources. Area C represents the gain due to a more efficient allocation of domestic resources (production factors): the country stop producing goods at high cost, now buy them for cheaper and can use the saved resources to produce other goods which are more worth to consumers. Area E represents the gain in consumer's welfare due to the extra amount of consumption. So far, they were buying too few goods.

It is therefore clear that a country can lose when it liberalises on a discriminatory basis, depending on the relative weight of trade diversion (F) and trade creation.

 We have to bear in mind that F is a loss only because T is due to a rent-creating barrier benefiting the home government (a tariff in our example).

However, there are some rent-creating barriers which do not benefit the home government but the foreign one - as for instance Volunteer Export Restrictions or the policy consisting in forcing foreign companies to raise their prices under the threat of anti-dumping duties (allowed by the GATT). In such cases it is the foreign economic actors who earn the profit equal to area D+F prior to the liberalisation. Thus when the home country enters a free trade area encompassing country A, there is no loss to the home country to offset the gains of lower prices. The gain from joining the CU is therefore C+D+E. The benefit is strictly positive - even though trade is still diverted from a low cost to a high cost producer.

Moreover, if T is due to a cost-raising barrier, a free trade agreement with A does not entail any welfare loss, event though trade diversion still occurs. Let us remind that a cost-raising barrier does not create any rent for anybody, and are exemplified by excessive bureaucratic delays or standard verification procedures. The mere fact of giving up all the procedures entailed by inefficient customs checking, results in net gains for the home country.

 

The conclusion to be drawn from this analysis is that there are two conditions under which a small country may (depending on the relative weight of F and C+E) lose from joining a free trade area

Firstly, this area must not contain the country whose producers are the most efficient - we shall expand on this idea soon. Secondly, the home country by entering a free trade area has to give up a rent-creating barrier that used to benefit its government (a tariff for instance). If both conditions are met, a trade off between trade creation and trade diversion emerges that may end up in net loses.

To put it the other way round, joining a free trade area is necessarily beneficial 1) either if this area contains the most efficient producer 2) or if the home country has only to give up trade barriers that do not generate any revenue for the government.

 

 

Can we find a practical criteria indicating whether a country is likely to suffer or benefit from joining a free trade area ? Lypsey has identified many rules under which a discriminatory liberalisation is likely to increase the national welfare. Clearly, the most enduring and intuitive one concerns the fraction of total imports that come from others members of the preferential area. Let us try to make this crucial point crystal clear.

 

In the case presented above, the home country initially imports none of the goods from the country with which it signed the free trade agreement. Consequently, the discriminatory tariff-cut led consumers to switch from low-cost supplier to high-cost one. If the former barrier T constitutes a rent for the home government (as a tariff does), this opens the door to a possible welfare loss from the bilateral agreement (C+E - F can be positive or negative).

For comparison, suppose instead that the free trade deal was signed with the lowest-cost supplier, let us show that the liberalisation would unambiguously make the home country better off. Consider the following graph :

 

 

 

 

        Pw+T

        Pa+T

           T                          A      B  C               D             E

            Pa

 

                           Q1     Q2                     Q3         Q4

 

Let's us assume that country A has now the lowest-cost producers and proposes therefore the cheapest goods on the world market at a price Pa. The rest of the world (W) can only propose a price Pb superior to Pa. In this case, even before the liberalisation, the home country trade with A and not with W. A discriminatory trade liberalisation will not end up in this case in trade diversion, since a free trade agreement with A will only reinforce the incentive for H to trade with A.

The welfare analysis of the removal of the trade barrier T is straightforward. The shift from Pa+T down to Pa benefits the consumers by A+B+C+D+E while it harms local producers by - (A+B). If T was a rent-creating barrier (a tariff for instance), then the home government loses D, and the remaining welfare gain is C+E. If T was a pure cost-creating barrier (as administrative inefficiency, protectionist standards) or a rent-creating barrier benefiting the foreign government (as a Voluntary Export Restraint) then the gain is even greater : C+D+E. In both cases the home country is necessarily better off.

Here the point is that a country cannot lose from joining a free trade are if this area encompasses the country which has the lowest export price. For a given product, if H was already importing (before the free trade agreement) from his new trade partner, the later was likely to be the cheapest supplier, and therefore H cannot lose (on this given market) from giving up its trade barriers.

Actual trade between countries is made up of thousands of products, so to be complete one should go through thousands of diagrams like the first graph, to check whether the home country would gain or lose from a preferential trade agreement. However, an obvious short-cut is simply to look at how much trade the home country does with its intended partners.

If it already imports a great deal from its partners, then it is likely that its partners were the low-cost suppliers even before liberalisation. Consequently, the discriminatory tariff cuts on import from this countries are likely to lead to little switching (trade diversion) and therefore to be on the whole beneficial.

 

The conclusion of this first part is therefore that a country can lose from joining a free trade area if the countries involved are not its cheapest import suppliers and therefore its biggest import suppliers. In a nutshell, a free trade area is beneficial for countries which are already integrated, in terms of imports and exports .Therefore the fraction of total imports coming from the members of the free trade area, is the main criteria. The higher this ratio, the more beneficial the agreement is likely to be for the home country. 

 

 

The analysis so far has focused on the free trade area dimension of a Custom Union. Its second dimension is the setting up of Common External Tariff vis-à-vis the rest of the world, that is to say a common commercial policy. Let us assume that a country decides to join a free trade area on the ground that the expected welfare effect is positive . It has still to decide whether or not it wants  to join CETs. Can it lose from joining CETs ?

 

The first thing to note is that a free trade area naturally creates a phenomenon of deflection of trade (Balasa, 1961). Deflection of trade occurs when imports from W (the cheapest source of supply) come via the member country with the lower tariff rate (assuming that transport and administrative costs do not outweigh the tariff differential).

The problem with deflection of trade is that it challenges national commercial policies. Indeed, if deflection of trade occurs, then the free trade area effectively becomes a CU with a Common External Tariff (CET) equal to the lowest tariff , with this important feature that none of the CET is commonly agreed upon.

In order to avoid this result, a free trade area will always adopt one of the two following strategies : either to set up Rules of Origins, or set up a true Common Commercial Policy with Common External Tariffs. Therefore the choice for the country willing to join a free trade area is between joining a free trade area with Rules of Origins or a free trade area with CETs (which is a full CU).

This paper argues that a country willing to join a free trade area (on the ground of a welfare calculation), is necessarily better off  with CETs, rather than with Rules of Origins.

The reason for that is that Rules of Origins actually and unintentionally act as cost-creating barriers, harming free trade (Baldwin). Indeed, Rules of Origins specify how customs official determine the origins of particular products in order to know which trade barrier to apply. They can be very complicated, imposing very heavy administrative costs on exporters and importers, to such an extent that some can prefer to pay the tariff rather than costly qualify for zero-tariff treatment.

According to Baldwin, in 1984, a fifth of the EU's exports to EFTA and a quarter of EFTA exports to the EU were subject to the tariff ! Therefore, rules of origins act as de facto cost-creating barriers. As long as the home country think that free trade is in its own advantage, it should be committed to the elimination of Rules of Origins.

CETs are therefore a better mean to avoid deflection of trade in so far as it does not harm free trade within the area. The setting up of a common commercial policy is a mean for given countries to recuperate some sovereignty, a say in the setting up of the tariff that apply in the home country, yet without creating trade impediments within the free trade area.

 

 

The overall conclusion of this paper is therefore that a country can definitely lose from joining a Free Trade Area (first component of a CU) However, as long as it benefits from joining such an area, it cannot lose from joining a subsequent Common Commercial Policy entailing Common External Tariffs.

Whether or not a country should join a CU is therefore only dependent upon whether it benefits from free trade with specific countries. This condition is generally fulfilled as soon as these countries are its main import suppliers.

 

 

 

BACK TO PUBLIC POLICY ANALYSIS