Preferential trading agreements (PTA’s) are one of the biggest issues facing world trade, they involve groups of countries charging high trade tariffs to the rest of the world while lower on no trade tariffs are charged to the countries within the preferential trading agreement. The European Union and NAFTA (North American Free Trade Association) are examples of these, while the countries within the agreements benefit, the ones outside do not which puts a serious block on some developing countries growth.
I will be making use of economic theory to examine this. For ease of explanation and due to space limitations, I will only be looking at the static economy, it is static in the sense that it is restricted by its assumptions as they appear unrealistic, but analysis of this type does make conceptual understanding far easier. Before we begin, I feel that it is important for me to distinguish between 2 types of PTA, a free trade area which is a group of countries with no trade barriers between themselves and a customs union which is a free trade area with common external tariffs and quotas.
These terms are very important as they are where some of the comparisons lie in the relevant theories, the first of these looking at production effects, trade creation and trade diversion before briefly examining the free trade vs. customs union issue in the same terms. I will then look at these concepts in relation to consumption effects before looking at some of the factors that will in theoretical terms create a greater gain in welfare. I have previously mentioned the assumptions and they are included in the paragraph below.
There are only 3 countries, A, B and C, A and B always form a union and C is the rest of the world, Tariffs are the only instrument of trade policy, there is perfect competition, production costs per unit of output are constant and directly determine retail prices and the economies are static with no expectations. There are also no new goods, no innovation and no depreciation of capital stock. All goods and services are homogeneous and have unit income elasticity’s of demand; there are no non tradable commodities, no intra union trade and no inventories. As far as intervention is concerned, fiscal and monetary operations are ruled out and country C which falls outside the union is assumed not to retaliate.
The most accepted theory on trade creation and trade diversion was supplied by Viner in 1950
Country A B C
Price 35s 26s 20s
Source: Theory of Customs Unions: A General Survey by RG Lipsey
Economic Journal Vol 70 Sept – Dec 1960 p497
If country A levied a Tariff of 100%, it would be enough to protect its domestic market that is producing commodity X. If A formed a customs union with B it would only have to pay 26s instead of 35s., this is trade creation. Now before the existence of the customs union A would have had to pay a tariff on goods imported from B and C, she would always buy goods from C as they are produced at the cheaper price. However with the formation of the customs union with B, the tariff which was say 50% is waived, and A will buy from B as opposed to C who is producing the good more efficiently. This is how Viner described trade diversion, trade is being diverted from a lower to a higher real cost source. What this means is that B is sufficiently tariff protected to eliminate competition from C which will secure A’s market for B’s inefficient industry. However A and B are both producing the product inefficiently and a customs union is formed then B will capture the union market at the expense of A, this is trade creation which results from trade diversion. Viner always regarded trade creation as good and trade diversion bad.
“Where the trade creating force is predominant, one of the members at least must benefit, both may benefit, the combined two must have a net benefit, and the world at large benefits; but only the outside world loses, in the short run at least, and can gain in the long run only as a result of the general diffusion of the increased prosperity of the customs union area. Where the trade diverting effect is predominant, one at least of the member countries is bound to be injured, both maybe injured, the two combined will suffer a net injury, and there will be an injury to the outside world and the world at large”.
If you refer to Fig 2 below, assume that the supply from the producers in the rest of the world is fully elastic at Pw. Also assume that Country A is the high cost producer where it wants to capture its home market by a large tariff. Country B produces at a lower cost and only has a low tariff. Suppose now that A and B form a customs union which has a common external tariff t*, this is determined by an averaging of the two tariffs. Supply and demand settles at Pcu. Now country A will buy all its imports BE from B as Pcu is lower than Pwt*t, production in country A is now OaB and country B then produces ObE’ and of this exports B’E’ equal to BE in country A.
Fig 2: Trade and Production Effects of a Customs Union for Countries A and B
Country A Country B
Source: The Economics of European Integration by W Molle p98
If initially, country A was protected, then there is a positive outcome, the new trade flow BE exists between countries A and B and CD is trade diversion as it replaces the imports that would have otherwise come from C. BC is trade creation and DE is trade expansion, so as BC + DE is larger than CD in this case, there is a positive outcome. However this is not the case if country A started from a free Trade situation, as trade would have been lessened by AB on the producer side and by EF on the consumer side, so BE is being diverted from the most efficient country C to the higher cost country B. If country B started from free trade, the common tariff will stop the trade that may have existed between country B and country C, which means that there will be less trade creation and expansion as the less efficient producers of B take over from the more efficient C. If country B was initially tariff protected then there will be no trade effects as there were no imports flowing into country B .
To get and idea of the magnitude of the welfare aspect refer to fig 3 below and assume that the customs union price is Pcu.
Fig 3: Welfare Effects Of A Trade Diverting Customs Union
Country A Country B
Source: The Economics of European Integration by W Molle p100
For country A’s production, trade creation is represented by the blue Ä KRL trade expansion or consumption is represented by the orange Ä UMN and trade diversion is shown by HLMI, but the production inputs are larger which results in a net gain. In the case of country B, the customs union will cause a rise in price, the consumers get less but have to pay more and that is indicated by Ä VXU. As additional resources were drawn in to produce more of the product, there will also be a production loss Ä RZW, but these costs will surely be outweighed by the extra trade with country A, so country B will benefit in the long run.
Viner’s conclusion was challenged by Cooper and Massell in 1965 when they claimed that the theoretical process involved in determining trade creation and diversion was not useful in the analysis of welfare gain. Firstly they should, before the customs union is formed reduce the tariff level for B and C so it would give the same union price and production, consumption and import changes. Then they should start a customs union from the new price which if you refer back to Fig 3 would be OaPcu, this will ensure that the gains from trade creation ÄKRL and ÄMUN still accrue while the losses from the trade diverting effect will not be there as the new supply curve GKLMN makes sure that imports come from country C at the cost of CDHI. As far as trade creation goes, the new imports will cost less which will lead to gains of KVLH and MNIT. This suggests that an equal tariff reduction policy is more favourable than a customs union as far as world welfare is concerned, the trade diverting effect is eliminated and there is more trade creation. This is reverting back to the concept of a free trade area.
However Aarnt 1968 pointed out that variable terms of trade can make customs unions superior to Cooper and Massell’s alternative which is a rather valid point in terms of the EU and its bargaining power. Wonnacott and Wonnacott 1981 looked at the scenario of the rest of the world also employing tariffs, they looked at the reasons why customs unions are formed and came to the conclusion that as the world is already tariff ridden then a customs union is the best option.
Johnson 1965 came to the conclusion that trade diversion is preferable to trade creation as the supply curve of C becomes more inelastic, although C will lose more, the countries within the union will sacrifice less and less domestic production. He added that both trade creation and trade diversion may lead to greater efficiency through economies of scale. Johnson’s theory is not as accepted as the Viner interpretation on the grounds of conflicting views on the assumptions that he made, the distinctions between social and private costs and benefits .
What we have looked so far is partial equilibrium model which only considers the market for a single good. An extension on the static framework includes a general equilibrium model which considers all markets. In a theoretical frame work Lipsey 1957 considered this scenario using three countries and two goods in the same manner as previously discussed with respect to the three countries, A and B forming a union to stop A trading with C. However Lipsey adopts a slightly different approach in the way that he uses indifference curves in his analysis. This is shown below in fig 3.
Fig 3: Welfare In A Trade Diverting Customs Union
Source: International Economic Integration By M N Jovanovic p34
In a free trade situation A would trade with C and achieve indifference curve II at equilibrium point E, if a non discriminatory tariff was imposed then the price would increase to AT and equilibrium would be achieved at G along I’I’. If now that the government put the tariff money back into the economy then equilibrium must be at point K along AC. If now country A formed a customs union with B, the terms of trade line is AB, now points K and L lie along the same indifference curve I2I2. Although the structure of consumption has changed, the formation of the union has not effected A’s welfare. If point E represents the most favourable point of consumption, the formation of the union will only result in a shift from one second best position L to another sub optimal position K. If this occurs then country A is indifferent as far as welfare is concerned. If A’s terms of trade are worse than OB/OA then a common external tariff will make a worse off and have a negative welfare effect. If those terms are better than OB/OA then trade diversion can be beneficial to country A.
To go one step further, Lipsey developed his model where both goods are produced in the same country and substitution in production is also permitted. This is shown diagrammatically in fig 4 below.
Fig 4: Production and Consumption in Country A Before and After a Tariff
Source: International Economic Integration By M N Jovanovic p35
If country A is initially self sufficient, she will produce at Qa and consume at Ca, this is as both functions are tangential to the price line PA at the same point. If country A started trading in a free trade scenario at a price ratio PP between good X and Good Y then country A will now produce at Q and consume at C, as amount FG of good Y is exported and JK of good X is imported. If country A imposed a non discriminatory tariff on imports, production is at Q2 along the new price line TT and consumption is at C2. If the government decided to put all the tariff money back into the economy then there will be more imports shown by line RR and TT equals tariff revenue remembering that at this stage A is still importing from the most efficient producer C. If now A and B form a customs union which diverts trade away from C, the customs union will be beneficial to A if the new indifference curve is higher than I3, If it is lower than I3 then country is better off imposing the non discriminatory tariff. If however the indifference curve is higher than I3, then A’s welfare would be increased implying that trade diversion is not always bad.
Theoretically Lipsey’s argument is sound but we cannot really make conclusions on counting optimal conditions but further analysis by him in “A Theory of Customs Union: A General Survey” , suggests that the more trade that takes place between the union members in comparison to trade with countries not in the union then the more likely it is that the consumption effects will be positive and the lower the pre union trade by the member countries will lower the possibility of negative consumption effects. But this must depend on the original tariffs between the union countries, if the initial tariff was high then the consumption pattern will be severely distorted, but eventually as more new trade is created the consumer satisfaction per unit of trade will increase. If high tariffs are kept in place against country C then there will be a loss in welfare that results from the reduced imports.
Although Lipsey’s theory shows that trade diversion can be good for welfare in certain circumstances it does not move away from the fact that trade creation is the more favourable option. A customs union is in itself economically a second best option , but it is important for trade creation and in turn welfare to predominate. A low demand for C’s products and a high demand for those produced in the customs union will cause trade creation. The greater the common external tariff, the more trade is created, the flatter the supply and demand curves, the greater the trade creation areas, the more countries that participate in a customs union lessens the chance of trade diversion, this means that the probability of trade diversion will decrease as more countries are added as there are only a finite number of countries in the world. It is advisable that integration occurs between neighbouring countries as they will see the real impact of integration by reduced transport costs. If the countries that form a union have a competitive production structure then there must be better substitute goods so there will be an increase in trade and positive consumption effects. If however there is complementarity between the goods of the countries, then there will be little reduction in imports anyway and the countries within the union will not see as much gain. The more small firms that exist within an industry, the smaller the potential for resistance and the greater the competition within an integrated market.
In this analysis, I have only looked at the static economy, of course in reality it is very different, all of these theories discussed only really hold true due to their assumptions, studies into the dynamic economics of customs unions looks at the effects of terms of trade, economies of scale, increased competition, rapid spreading of technology along with transport costs and geographical spread along with administrative economies and greater collusion between companies leading to higher prices .
These arguments in relation to welfare are very difficult to assess, as it cannot be accurately determined what a countries position would have been like had it not joined the union in the first place, what about future attitudes, future policies and world events, or distinctly political areas like greater political influence or loss of sovereignty. Many of these issues are discussed by politicians and cannot be accurately quantified in economic theory.