The balance and terms of trade and economic growth.
1. This paper attempts to clarify the relations between three approaches to the problems of International Equilibrium between developing economies , which have not always been kept sufficiently distinct in earlier treatments of the problems involved . These are a consideration of the forces governing changes in the terms of trade, as analysed by a simple «real » model, an analysis of some of the forces governing the development of the balance of trade, most simply treated in a «Marshallian» manner, and an assessment of some of the types of development most likely to occur in practice.
2. Economic Expansion and the Terms of Trade.
The «real » model used to analyse changes in the terms of trade is that put forward in «The Manchester School » (May 1955) by Professor H. G. Johnson. There are assumed to be two countries, in each of which the same two goods are produced, in each case at increasing opportunity costs in terms of the other good. This model can be used to analyse both the conditions of static equilibrium, and the movement of the equilibrium trading and production positions in the course of the growth of the two countries. This is true whether the causes of growth are increasing supplies of factors of production, increasing productivity, or both.
3. Figure 1 is used to show the amounts of trade one such country would wish to conduct at any given terms of trade, at different stages of its development. The axes represent amounts of the two commodities. AB represents the production possibility function, TT the terms of trade, II the highest available community indifference curve, and P and C the combinations of goods produced and consumed at time 1, so that PC is proportional to the amount of trade desired at time 1. Similar magnitudes at time 2 are shown by primes (A'B', etc.). Such diagrams can be constructed for both economies ; if PC represents the actual trade conducted at time 1, then the terms of trade must turn against the country whose amount of trade desired at the given terms increased more (i. e. T'T' being parallel to TT, the terms of trade turn against the country whose P'C' is greater).
4. In discussing the effects of different types of development it is helpful to use a few special terms to describe the changes in the production or consumption which would take place if each economy could do the amount of trade it chose at given terms of trade at any time. Thus we describe as «unbiased » growth which leaves production of exportables as a constant fraction of total production, and similarly for consumption. Changes which would increase the proportion of total production taking the form of exportables, or decrease their proportional share of total consumption, tend to increase the relation of international trade to national income, and are therefore called «trade-biased» (or «export-biased y>). Changes which decrease the proportion of total production taking the form of exportables, or increase their proportional share in total consumption, tend to bias development in the direction of autarky, i. e. diminish international trade as a proportion of national income, and are therefore called «auto-biased » (or «import-biased»). If either form of bias is so extreme that the absolute share of either commodity in production or consumption is diminished as the economy grows, we get cases of «ultra-bias».
«Total bias » is the ratio of the marginal to the average proportion of international trade to national income ; this can be derived from the separate biases of production and consumption by a simple algebraic method, given in the text, or by a geometrical construction, as in figure 1 (where both production and consumption are auto-biased, and total bias is ultra-auto-biased, since P'C' is less than PC. All possible results are shown in table 1 (where U = unbiased, andfsignifies alternatives). It will be noted that for each type of bias in production or consumption, the greater the trade-bias of the other, the greater the trade-bias of their total effect. The growth of a country’s desire to trade, on fixed terms, is governed by the total bias of its development, and by its rale of growth of total produclion.
5. If a country had a uniform rate of growth of productivity in all industries, its production growth would be unbiased ; in consumption unbiased growth would result from unit income elasticities of demand for both goods. If productivity increased faster in the production of exportables than in importables-produdion, trade bias would result, since not only would the output per unit of factors employed in making exportables rise tnore than in importables, but, since if product prices were constant real wages in exportables production would increase faster, there would be an incentive for factors to move from importables into exportables-industries. The extent of the resulting bias would depend both on factor mobility and the shape of the production functions of the two industries.
6. A Simple Marshallian Model.
The effects of growth of various types on the balance of trade are most conveniently analysed by a Marshallian model, which can take account of the effects of changes in relative prices, as well as the effects of changing techniques and tastes, to which the «real» model is confined. Each country in such a model is assumed to produce and consume «home-goods», which are not traded internationally, and also to export to the other goods of a type which it does not consume, an export supply curve connecting the supply of «export-goods» with their price relative to home-goods ; and to import from the other goods which it does not produce for itself, an import demand curve relating the demand for «import goods» to their price relative to home-goods. The import and export markets are assumed to be independent, and each country is assumed to utilise a growing supply of factors of production at some fixed ratio to full employment level ; thus the multiplier effects of changes in the balance of trade are assumed to be offset.
7. If the prices of home-goods in each country are assumed to be constant, the supply and demand curves for imports and exports can be combined, as in figures 2 and 3, to give the balance of trade. The four curves will shift over time at rates depending on the growth of total productive capacity and on the types of bias in each economy. The types of bias are, as before, classified according to whether the demand for imports or supply of exports (at any given price relative to home-goods ) would increase proportionately faster or slower than total productive capacity (yielding trade and auto-bias respectively). If either curve shifts to the left as total productive capacity increases we have a case of ultra-auto-bias, and if its shift to the right takes up all the increase in total production, we have ultra-trade-bias. For either country, the greater its trade-bias in import demand, the more adverse its balance of trade tends to become ; while the greater the trade-bias of its export supply, the more its balance of trade tends to improve provided the foreign price-elasticity of demand is over unity (or to deteriorate if the foreign price-elasticity of demand is less than unity).
8. Thus the Marshallian model can be used to find the effects of «struc¬ tural » changes in the two economies on the balance of trade given the relative price of home goods in the two countries. It can also be used to find the effects of changes in the relative price of home goods, whether through changes in their prices in own-currencies, or via alterations in the exchange rate. The effects of such changes can be shown by a bodily shift up or down of country Ts curves relative to those of country 2, in figures 2 and 3. The results can be found from the figures, or by the use of Mrs J. Robinson’s celebrated formula (in «The Foreign Exchanges y>), and these can be combined with the results of structural changes occurring at the same time. Unbiased expansion by one country, while the other is static, tends to worsen the balance of trade of the expanding country under any plausible set of supply and demand elasticities ; thus cel. par. the faster growing of two countries is likely to suffer from worsening balances unless this tendency is offset by appropriate structural bias or price-trends.
9. Even the Marshallian model cannot satisfactorily be used to analyse the effects of short-run fluctuations in employment, which demand special treatment for which space is not here available. This is because both price elasticities and biases are likely to be different, in the course of short-run fluctuations, from their long -run values.
10. Some Probable Biases in Economic Development.
The following sections discuss some of the types of bias likely to appear in the course of economic development, and the interaction between «real » aspects of growth, and relative prices (though the treatment is necessarily far from exhaustive).
The influence of innovations, and the resulting development of new and improved products, on the biases, can best be dealt with via the case of quality improvements. These will cause changes in supply and demand to be inter-related. Improved export goods tend to increase foreign demand by attracting expenditure from other objects : at the same time they make it possible to satisfy a given range of wants with less expenditure, but assuming that the former influence is stronger, innovations in export goods will tend to make for trade-bias in the growth of foreign demand, which will tend to improve the exporter’s single-fadoral terms of trade, providing the improved commodity terms of trade are achieved without corresponding increases in factor-inputs. The same features as those which attract foreign customers are likely to divert home consumers’ purchases towards the improved exportables, with similar beneficial effects on the terms of trade (using our ureal» model).
In the Marshallian model, if we again assume that improvements in the quality of a good tend to attract more purchasing power towards it, innova¬ tions in export-goods will tend to produce trade-bias in foreign demand, and innovations in home-goods competitive with imports will tend to produce auto-bias in home demand, both being good for the balances of the innovating country.
11. Productivity and Price Trends.
These can be formally treated as if they were independent, but in practice they may be related. Thus if development took forms involving worsened terms of trade for one country, it might be argued that the resistance to real-wage reductions would cause inflation in costs and prices in the country concerned, with adverse effects on its balance of trade. Productivity increases may counteract this inflationary tendency by preventing real wages from falling ; but the rate of increase of real wages consequent on a certain rate of productivity increase (with constant terms of trade) may itself come to be regarded as the minimum acceptable, in which case any worsening of the terms of trade would again lead to inflationary pressure.
12. Both this relation, and a possible relation between the pressure for increased real wages and the degree of unemployment, are expressed in algebraic form in the text.
13. Where the rate of real wages (or of their increase), which is regarded as the minimum acceptable in an economy, is impossible given the terms of trade resulting from the course of a development, a country may be forced to choose between an adverse balance of trade due to wage-cost inflation, and counteracting this by demand deflation ; (the third alternative, that of direct controls on trade or wages, is not discussed). Yet, though demand-deflation is likely to decrease the pressure for higher wages in the short run, it is also > likely to involve decreases in investment , which in the longer run accentuate the disparity between the possible and the «acceptable » rates of growth of real wages, so that a vicious circle may be set up. The influence of changes in the terms of trade on real incomes may in the long run be far greater than the direct effects, since these are magnified as the result of attempts to avoid them.
14. The Time Pattern of Growth.
This section discusses some of the effects of lags in the adaption of an economy to changing conditions of long run equilibrium (in the, models used above this adaption was assumed to be immediate). For example , during a period of adaption of production to a greater degree of trade-bias, a country’s producers of exportables would enjoy real incomes above their long-run normal, and producers of importables would suffer from lower ones, the difference being needed to induce people to move ; and the terms of trade would for the time being be more favourable than when full adaption had taken place.
Allowance must also be made for the tendency of innovations made by exporters in one country to spread to importing countries, either with the passage of lime or the growth of the market, thus giving a tendency for trade-bias in consumption due to technical inferiority to be succeeded by auto-bias.
15. This paragraph discusses briefly the possibility of a tendency to auto-bias, or even ultra-auto-bias, in the development of advanced countries, due both to economies in the use of raw materials, and a tendency for the proportion of incomes spent on services to rise.
16. Income-Elasticity of Demand and the Direction of Innovations.
While the quantity of investment and the «amount » of innovations may
fairly be treated as autonomous, the direction in which they are applied may itself be related to the bias of the growth of demand ; on the whole, investment and innovation will pay best if applied to expanding markets, i. e. those with high income elasticities of demand, whether at home or abroad. Thus high income elasticity of demand for importables may tend to make the development of production auto-biased, while high income elasticity of demand for exportables may, by inducing quality improvements, tend to make foreign demand trade-biased in its growth (this latter would appear to be approximately the state of affairs with regard to United States exports of durable consumer goods).
17. Adaptability and the Terms of Trade.
This section discusses the widely held belief that some special disadvantage with respect to the terms of trade attaches to underdeveloped countries. Such countries obviously suffer directly from low levels and rates of growth of productivity, and they may also suffer from worse double-factor at terms of trade (i. e. relative real wages compared with advanced countries) than they would have had if the greater enterprise, adaptability and factor-mobility of developed countries did not tend to leave the backward countries with the secularly less prosperous industries (including primary production). Also cyclical instability of demand may affect adversely the terms of trade of less adaptable countries (since others are enabled by their greater adaptability to share the booms and leave them to bear the slumps). But neither of these factors would account for a worsening trend in the terms of trade of underdeveloped countries, except insofar as low incomes at any one lime hold down investment and education, and thus tend to perpetuate poor adaptability.
When increased production in primary producing countries is concentrated on a narrow range of commodities with low income and price elasticities of demand abroad, the commodity terms of trade are bound to worsen. If production rises through extensive increases of cultivation with unchanged levels of productivity, single facioral terms of trade (i. e. the amount of imports received per unit of factors devoted to export production ) will worsen also ; but if increased production is due to improved productivity, single factoral terms of trade will worsen only if foreign demand is price-inelastic.
18. Conclusions.
The above discussion includes both an exposition in sections 1 to 9 of two of the models most readily applied to the problems of international equilibrium between growing economies, and a consideration of some of the forces affecting the values which the parameters in such models are likely lo assume in the course of actual economic development.