Britain's Trade: Has Europe Made a Difference?

The fourteen years since Britain acceded to the European Economic Community have seen fundamental shifts in the pattern of her international trade; exports and imports have become more heavily concentrated on her new partners in the Community. It is only natural that economists and policy-makers should wonder to what extent these changes in trade patterns are due to accession. We cannot simply attribute the entire change in trade patterns to accession; other factors might have brought about changes even if the UK had not joined the Community. We therefore need a theory which specifies the variables which influence trade patterns and quantitative estimates of these influences. This note discusses recent attempts to estimate rigorously the effects of accession to the Common Market on British trade in manufactures.

Imagine. the market for a single good , which Britain trades with the only two other 'countries' in the world - Europe and ROW (the Rest of the World). Suppose also that initially Britain imposed the same tariff on imports from each source, and that after forming a customs union with Europe, the latter's imports are admitted free of duty .The customs union thus reduces the price of the good from European suppliers relative to suppliers in Britain and ROW.

This shift in relative prices can have a number of consequences, some of which were first set forth by Jacob Viner in his 1950 pioneering study of customs unions. First, European supplies might displace British supplies: this represents an increase in international trade, or 'trade creation'. Trade creation, if it occurs, represents a net saving of resources: UK consumers are now buying European goods because they are cheaper, duty free, than those produced in Britain. Second, European goods might displace ROW goods from the British market; this is 'trade diversion', and involves a welfare loss for Britain. Why should this be so? We have assumed that before the customs union, Europe and ROW faced the same UK tariff; because ROW supplied the UK market, the ROW price (excluding the duty) must have been lower than the European price (excluding duty). After the formation of the union, however, Britons may buy from Europe instead, if the if the duty-free European price is lower than the rest of the world duty-inclusive one. Although UK consumers pay less, Britain as a whole is worse off because she no longer receives any tariff revenue. The real resource cost arises because Britain has shifted to a supplier who produces at a higher cost (excluding duty, of course).

Third, consumption might rise as a result of entry into the customs union, because the price paid by consumers will have fallen. Fourth, the lower duties inside the customs union allow European and British suppliers to expand their markets. If there are economies of scale in production, this may lead to lower costs, and these suppliers may capture markets from ROW suppliers. If a supplier captures more of his own domestic market, we have 'trade suppression': if the gain is in a partner's market, we have something like trade diversion. Fifth, policy changes associated with the formation of a customs union also influence trade patterns. Such changes may be autonomous, e.g. the adoption of common transport or energy policies, or accommodating, that is, policy responses to the economic effects of the shifts in trade patterns resulting from accession. For example, a trade deficit induced by accession may lead to compensating shifts in macroeconomic policy and exchange rates. Strictly speaking, these should be taken into account in analysing the effects of accession. Finally, it is often argued that customs unions have 'dynamic effects', for example on confidence and rates of technical progress, which could obviously affect trade flows.

The mercantilist fallacy

The discussion so far has proceeded entirely in terms of the markets for imports, but export effects have figured just as prominently in most discussions of accession. In part concern over exports per se reflects the still widespread mercantilist fallacy - that the direct determinants of economic welfare are producing and selling, rather than buying and consuming. In part, however, it represents a genuine concern. Trade creation and diversion in partner countries stimulate exports from the acceding country, and any rents accruing from exporting (surpluses of export prices over costs including normal profits) raise economic welfare. Export effects are obviously also important in the determination of necessary accommodating policy changes. Not only will exports to partners rise, but those to non-partners may fall, depending on whether sales lost at home from trade creation offset the increased sales to new partner countries.

To calculate the effects of accession to the European Communities (EC) on international trade flows and economic welfare requires measures trade both under accession and also under identical circumstances but without accession. In ex-post analyses of the type described here, the 'with-accession' data are the observed trade flows and are easily calculated. The 'no-accession' figures, however, require estimation. The hypothetical 'no-accession' scenario is referred to as the anti-monde; it is derived from a model of trade flows and is the crucial element of any numerical assessment of accession. Behind any statement about the quantitative effects of Britain joining the EC lies some hypothesized anti-monde, even if it is merely a crude one of 'no change'.

Although studies which estimate the effects of accession differ principally in their assumptions concerning the anti-monde, one further distinction should be drawn. having specified the basic model of trade one can approach the calculation of integration effects in two ways. First, one could estimate the anti-monde by setting all variables affected by integration to their 'no-accession' values and solving the trade model. The result will be an estimate of what trade would have been without accession, and subtracting this from the actual trade data provides a measure of the effects of accession: This method, known as residual Imputation' attributes all of the difference between the anti-monde and observed trade flows to accession. Conceptually, although not necessarily numerically, this overestimates the effects of accession, for it includes both the random shocks experienced by the economy and any factors which were incorrectly excluded from the model of trade. The second approach is to model integration more explicitly. This is possible only in the most stylized fashion; a popular and practical method is to include 'dummy' or shift variables in the trade model to represent the difference between pre- and post-accession observations.

British Imports of Manufactures 1962-84

Note: All figures, including GDP, refer to the manufacturing sector only. Figures from 1970/2 onwards are three-year averages.

Source: Department of Trade and Industry, Evidence to the House of Lords Select Committee on the European Communities, HL41, Session 1983/4; and Winters' own calculations, based on the CSO Pink and Blue Books and Overseas Trade Statistics.

Before attempting to estimate the effects of accession, it is clearly sensible to search for prima facie evidence that such an effect exists. Table 1 reports the sources of UK manufactured imports and the imports/GDP ratio for manufacturing. The import shares by source show a clear trend towards the six original members of the European Community. This trend was in evidence before accession, based perhaps on 'natural' European economic integration and the rapidly rising productive capabilities of the Six, but it clearly accelerated after 1973. Interestingly the other principal European countries, Ireland and the members of the  European Free Trade Association (EFTA}, were expanding their share of UK imports of manufactures even more rapidly than the EC Six before 1973, but lost ground subsequently. This pattern presumably reflects the establishment of EFTA and the Anglo-Irish Free Trade Association during the 1960s, followed by the Six catching up after 1973 when they achieved comparable treatment in British markets. The final column of Table 1 shows the imports/GDP ratio in  manufactures. It reflects the post-war tendency to- wards greater worldwide integration of manufacturing industries, but again the rate of change of this ratio appeared to increase significantly just after 1973.

Similar patterns are observable for manufactured exports: a noticeable drift towards Europe -especially EFTA -over the 1960s, followed by an obvious switch since accession towards the Six and away from EFTA, Ireland and the rest of the world. The behaviour of UK exports also demonstrates the increasing openness of  manufacturing, especially after 1973, but this is not as marked as in the case of imports, for which the initial trade/GDP ratio was much lower.

UK Shares of Selected Countries' Imports of Manufactures (%)

Note: All figures, including GDP, refer to the manufacturing sector only. Figures from 1970/2 onwards are are three-year averages except 1982/3.

Source: Monthly Review of External Trade Statistics Annual Supplement 1985.

Table 2 looks at UK shares of other countries' imports of manufactures. Prior to 1973 the decline in the UK's shares of almost every market was apparently inexorable, but subsequently, while the decline continued elsewhere, it was at least temporarily reversed for every one of the original Six members of the EC. The decline in the UK's share of Danish and Irish imports accelerated after 1973, as might be expected given the dilution of Britain's preferential treatment in these markets.

Table 3 presents my own estimates of the effects of accession on UK trade with its principal partners, based on recent research described more fully in CEPR Discussion Paper No.11 0. There, I argue that it is not appropriate to model only the relative shares of EC versus other countries' suppliers in UK imports.

The correct approach is to model the allocation of UK demand for manufactures across all suppliers, both foreign and domestic, and to allow for substitution by UK consumers among these suppliers. In studying the effects of accession I have adapted techniques commonly used in econometric estimation of complete systems of consumer demand, which can accommodate general patterns of substitution among goods. In such systems, the consumer's expenditure on a particular good (as a share of total expenditure) depends upon the good's own price, the prices of all other goods available to the consumer and total real expenditure. We can model the demand for manufactures in a particular country in exactly the same way. A supplier's share of total expenditure in this market will depend on the supplier's own price and the prices charged by other suppliers. The supplier's share also depends on total expenditure on manufactures in that market, which I assume is determined prior to its allocation across suppliers by a higher stage budgeting process.

Consumer demand theory can be applied to study accession

This approach can be applied to data on UK purchases of manufactures, and will allow us to model the UK allocation of purchases across domestic, EC and other countries' suppliers. The approach can also be applied to data on other countries' imports of manufactures, and the share of UK suppliers in these markets will provide us with a model of UK exports.

This approach has three important Implications. First, the model allows suppliers shares to depend not just on Income levels (through the total, expenditure variable). As UK consumers become richer, their expenditure on BMWs may rise, even if relative prices are unchanged. Second, relative prices are important: it is not legitimate to model UK consumers' allocation of a given level of imports over suppliers without taking into account the price of, and the demand for, home goods, that is, the share of home suppliers. For example, the allocation of British demand for imported cars between Germany (BMW) and, Italy (Fiat) is not independent of the, demand for British cars (Jaguars). The econometric estimates described in the Discussion Paper demonstrate this. The equations explaining the detailed allocation of imports among suppliers have considerably greater explanatory power once domestic prices are included. Home sales must be Included In the analysis of accession effects, not only in order to separate trade creation from diversion, but also to model correctly the switches in import shares caused by accession.

This approach also underlines the importance of modelling the allocation of expenditure on manufactures over sources in a rigorous fashion which is consistent with consumer demand theory. Common sense alone suggests the most obvious constraint: that shares of total expenditure should sum to 100%. Economic theory suggests that allocation models should satisfy other less obvious constraints. Equi-proportionate changes in all prices and total expenditure will leave relative prices and real expenditure unchanged, and hence should have no effect on trade shares. Theory also suggests that the effect of good i's price on the demand for good j should equal that of good is price on the demand for good j's for good i. These constraints must be imposed in order to estimate the trade model efficiently. A model which violates these theoretical constraints lacks credibility, and its predictions must be treated with caution.

In principle, the tariff changes which resulted from accession could be incorporated into the trade model quite simply by altering those explanatory variables which are affected by accession, namely the relative prices. In my study, however, I used dummy variables to represent accession, both because It involved a number of non-price effects and because data on the average increase in competitiveness of each suppliers bundle of exports were not available. The substantial size of the unadjusted accession effects is evident from Table 3. For every member of the Six, over half of actual, UK imports in 1979 can be attributed to accession, with imports from Germany alone rising by  £3.75 billion. For the non-EC countries smaller, but still sizeable, increases in imports are reported. This is plausible because accession reduced UK tariffs on third-country imports. The most striking result is the massive decline in UK sales in its own market, by £12 billion, which suggests very strong trade creation -the displacement of home by partner sources of supply.

The effects of accession on exports are modelled in a novel fashion in my research. The figures in Table 3 are generated by applying the shares model separately to each of the five major export markets for UK manufactures. They show the additional exports to the new partner countries in 1979 that can be attributed to accession. For Germany and Italy this amounts to over half of actual exports in that year. Interestingly, they also illustrate that UK exports to the United States and on relative prices,but also, Japan fell relative to what would have been expected given relative prices and these countries' incomes.

Note: Percentage figures refer to percentages of actual trade in 1979.

Source: Winters' own calculations

The figures in Table 3 refer only to a subset of UK trade flows. They can be generalized to cover all flows and to allow for the possible over-estimation of the effects of accession on imports. I would estimate very roughly that, as a result of accession, UK imports of manufactures from her new partners increased by £8 billion, home sales fell by £8 billion, exports to partners increased by £4.5 billion and exports to non-partners fell by £1.5 billion.

This analysis of British export markets also sheds some light on other aspects of British accession for example, accession induced some trade creation in Europe. Some previous trade diversion was reversed, as Britain captured some EC markets. from other EC suppliers. There was also export diversion, as German performance In the United States declined, but very little new net trade diversion. The results also suggest strongly that the observed changes in trade patterns were not due to adjustments to higher oil prices, a concern voiced by some investigators. Although Britain's increasing concentration on trade with Europe could have arisen from the sudden increase in transport costs that the oil price rise entailed, the growth of US and Japanese exports to Europe and of French exports to the United States casts doubt on this possibility.

While there are obvious uncertainties about the precise magnitudes of the estimated trade effects, one cannot avoid the conclusion that British accession to the EC worsened her trade balance in and reduced her output of manufactures quite substantially, by at least 1.5% of GDP according to my estimates.

following the analysis discussed above it is natural to wonder whether or not. Britain has benefited from accession to Europe. This is a very complex question, even assuming that one has modelled trade flows correctly.

Assessment of welfare effects is complex

Assessment of welfare effects is sometimes based on the behaviour of macroeconomic aggregates such as the balance of payments or unemployment. One school of thought, for example, argues that trade deficits are harmful. By reducing demand for domestic output, deficits reduce domestic income and investment and retard technical progress, and oblige macroeconomic policy to be deflationary in order to restore external balance. This analysis is typically conducted  within a simple Keynesian macroeconomic framework, in which unemployment can and does occur. This framework has much to teach us about the contemporary economy, for it is clear that unemployment imposes very significant costs. For several reasons, however, these models are not adequate for analysing customs unions.

First, the .macroeconomic framework typically involves only a single good, output, and is therefore quite unable to take account of the potential gain from specialization in production when relative prices change. Second, the overall current account deficit is largely determined by macroeconomic factors, not by trade policy. Thus to the extent that the exchange rate floated freely in the mid-1970s, the deficit in manufactures resulting from accession was offset by surpluses (lower deficits) in other components of trade and had no overall effect on output. To the extent that the exchange rate was fixed by government action, on the other hand, the current account deficit could indeed have been increased by accession, but then it could equally well have been reduced by choosing a more appropriate exchange rate target. Third, the emphasis on output and production in this analysis is essentially mercantilist, and it entirely ignores the increase in real income which individuals experience as a result of the greater availability of foreign goods. By stressing production and sales above purchase and consumption, mercantilism turns economics on its head and misses the most important factor in trade policy.

However economic welfare is measured, the growth of unemployment over the 1970s has obviously had a crucial impact. It is possible that unemployment may have increased as a result of the interaction of accession with more fundamental fac!ors such as labour market rigidities. It seems misguided to argue, however, either that accession was a major cause of unemployment, or that unemployment is the principal channel through which accession has affected economic welfare.

An alternative and more traditional approach to the measurement of the welfare effects of changes in trade is through the calculation of changes in consumer and producer surplus. Consumer surplus is the excess of utility that people gain from a good over and above what they pay for it. Producer surplus is the excess of revenue over the (total) costs of producing a good, that is, roughly speaking, profit in the economic rather than the accounting sense. Estimates of the change in these surpluses as a result of British accession are typically very small -around £5 per head per year. They are, however, subject to a number of well-known objections, especially under circumstances, like accession, in which many prices change. Compensating and equivalent variations have been suggested as more appropriate measures of the effects of relative price changes on consumer welfare. 'Compensating variation' is the increase in income that a consumer requires after a change in prices to allow him to remain at his former level of welfare; 'equivalent variation' is the increase in income he would need before the change in prices to allow him to reach the welfare level he actually achieves after the change.

Unfortunately the calculation of compensating and equivalent variations is quite complex, and it has not been applied to the analysis of accession to date. A preliminary attempt to calculate these measures from the import figures in Table 3 suggests that accession brought huge benefits to consumers -around £45 billion per year! This attempt is somewhat unsatisfactory from ~ technical point of view, however, and the welfare estimate should not be taken seriously. Nevertheless the approach is valid in principle, and the rigorous estimation of the welfare effects of UK accession to the EC should be a priority for future research. On the basis of existing studies, the most that can be said is that the changes in manufactures trade were probably beneficial to Britain, because they  reduced the costs of consuming manufactured goods.

The research outlined here could usefully be extended in a number of directions. CEPR Programme Director Alasdair Smith and Research Fellow Tony Venables are exploring the consequences for industry and consumer welfare of the increased variety of goods that accession has made available in the United Kingdom. The effect of accession on agricultural trade has been enormous and without question harmful. There is a need to examine the commodity- or industry-specific consequences of accession and also their regional impact. Moreover, accession influences the effectiveness of other policy measures - for example, industrial policy or import restrictions may have no impact on an industry if there is free intra-community trade and investment. Anglo-French Conference

L Alan Winters is a Lecturer in Economics at the University of Bristol and, from August 1986, Professor of Economics at the University College of North Wales, Bangor. He is a Research Fellow in the Centre's International Trade programme. The research described in this article is discussed at greater length in CEPR Discussion Paper No.110, 'Britain in Europe: A Survey of Quantitative Trade Studies'. Alan Winters discussed non-tariff barriers to trade at a March lunchtime meeting, reported in Bulletin No.14. Other CEPR research on trade issues can be found in Discussion Paper Nos. 38, 41, 42, and 74.