Global Macroeconomics
North-South interactions

Recently there has been renewed interest in the macroeconomic linkages between the developed and less developed countries in the world economy. This concern with North-South macroeconomic interactions is in part a response to the recent LDC debt crisis and its consequences for Northern financial markets and Southern growth prospects. In addition, the increased interest in international interdependence and policy coordination in the context of the Northern economies has highlighted a 'missing link' in these analyses: the Southern economies, which are not modelled in any detail in existing theoretical and econometric models of North-North interaction. Analyses of policy coordination are incomplete and possibly misleading if they do not take account of the LDCs.

These considerations inspired a recent conference on 'Macroeconomic Interactions Between North and South' held at the University of Sussex on 18/20 September. The conference was a joint venture by CEPR and the International Economics Study Group (IESG). It was organized by David Currie (Queen Mary College, London), Co-Director of the Centre's International Macroeconomics programme, and David Vines (Glasgow University and CEPR). Anton Muscatelli acted as rapporteur. Financial support was provided by the Ford Foundation, the Alfred P Sloan Foundation and the IESG itself.

The papers presented at the conference approached the subject of North-South interdependence from a variety of perspectives. One difficulty which the conference highlighted relates to the precise classification of 'North' and 'South'. The difficulties involved in treating as a single bloc countries which have different institutional, political and economic characteristics recurred in much of the research presented. Second, the precise approach adopted in constructing models of North-South interaction itself became an important source of controversy. The modelling of some particular North-South linkages, especially the interdependence caused by international financial and commodity markets, also proved contentious.

Three of the papers presented shared a common objective: to provide a comprehensive picture of North-South interdependence through the use of complete theoretical or empirical macroeconomic models. In the first paper of the conference, entitled 'An Econometric Investigation of North-South Interdependence', Michael Beenstock (Hebrew University of Jerusalem and CEPR) argued forcefully for an 'econometric' approach to modelling North-South interdependence. The construction of an empirical model, he contended, is preferable to the alternative 'oracular' approach, whereby a (usually complex) theoretical model is constructed and its parameters given a priori values.

Beenstock described the preliminary results of estimating a model of North-South interdependence. In contrast to models designed to study North-North interdependence, this did not attempt to distinguish regions or categories within the broad groups 'North' and 'South'. Although he conceded the possibility of bias in the estimates, as a result of inappropriate aggregation and data problems, Beenstock nevertheless advanced a number of tentative conclusions. The estimates suggested that the strongest influences in the model ran from North to South rather than vice- versa. In particular, the South appeared to benefit from an expansion of aggregated demand in the North, in part due to the commodity price increases caused by the higher demand for Southern commodities as Northern production increases. Spillover effects in the opposite direction, from higher Southern aggregate demand and imports onto Northern economic activity, are small in magnitude. Financial markets provide another channel through which the North affects the South. Capital and aid flows from the North to the South, for example, may bring the 'Dutch Disease' to the South through an appreciation of their exchange rates and a resulting squeeze on the tradable goods sector.

In his comments Michael Dooley (IMF) echoed Beenstock's concerns about the appropriate degree of aggregation in the model: there were a variety of criteria by which one could classify developing countries, such as by the types of commodities they produce or their degree of indebtedness. Beenstock conceded that a fuller specification of the South's economic structure and a distinction between manufactures and non-manufactures in Southern output would be useful. Detailed treatment of individual countries and regions would be a demanding exercise, however, and might not yield any deeper insights into North-South interactions. Jorge de Macedo (University of Lisbon and CEPR) emphasized that the appropriate classification of Southern countries may vary over time, as the structures of developing economies change. Other participants defended the use of 'oracular' models, which facilitated analysis of the sensitivity of policy measures to the underlying economic structure of the model.

In contrast to Beenstock's empirical model, the 'oracular' approach was presented in a paper by David Currie (Queen Mary College, London, and CEPR), David Vines (University of Glasgow and CEPR), Thomas Moutos (University of Glasgow), Anton Muscatelli (University of Glasgow), and Nic Vidalis (Queen Mary College, London). Their paper, entitled 'North-South Interactions: Cooperation and Non-Cooperation', developed a prototype simulation model which attempted to capture the main channels of North-South interdependence. The model was used to examine the spillover effects of exogenous shocks originating in each of the two blocs. Knowledge of these spillover effects allowed the gains from cooperation between North and South in setting macroeconomic policies to be evaluated using the game theory techniques common in the literature on policy coordination.

The model incorporated a variety of North-South linkages. First, commodity prices are forward-looking, and in addition to flow demands for commodities for consumption and production, speculative stocks of commodities are held by Northern investors. Spillover effects occur through the trade balance, and financial markets also link the two blocs: Northern interest rates affect Southern debt payments. Simulations of the model suggest that an expansion of aggregate demand in the North (with the Northern money stock unchanged) may not be welcomed in the South. Northern expansion may raise interest rates and hence Southern debt service payments. In addition, higher interest rates today and expectations of higher rates in the future may discourage speculative holdings of commodity stocks. This will depress commodity prices, even though increased Northern demand for commodities would otherwise raise their price. Similar transitory effects are caused by monetary contractions in the North. On the other hand, positive aggregate supply shocks in either region have beneficial effects shared by both North and South.

Although the authors recognize that the results of North-South cooperation may to some extent depend on the choice of parameters in the simulation model, the conclusions are nevertheless similar to those arrived at in analyses of policy coordination among the Northern economies. In particular, North-South cooperation in setting policies appeared to be beneficial only if both Southern and Northern governments had established reputations with respect to the private sector. In the absence of reputation, the prospect of more inflationary policies under policy coordination lessened its appeal.

In his comments Paul Armington (World Bank) argued that the results obtained by the authors could well be sensitive to their choices of parameter values and to the welfare functions used to evaluate policies. The robustness of the results clearly needed further investigation. Armington argued that in the 1970s and 1980s governments gained reputations through combating inflation, and hence cooperative solutions may yield welfare gains through the movements away from a deflationary equilibrium. Armington also suggested a number of possible extensions to the model, including finer disaggregation of the Southern economies and the introduction of creditworthiness into financial market interactions.

Other participants argued that any strategic analysis of policy- making must take into account the behaviour of different interest groups in the North as well as issues such as income distribution, which may concern Southern countries but were not analysed in the paper. Jorge de Macedo also noted that the use of techniques pioneered in the analysis of North-North policy coordination may not be readily applicable in the North-South context, given the distinctive political and institutional features of Southern policy-making. Michael Beenstock also felt that more emphasis should be placed on model-building than on complex simulations of policy coordination. The models used were only rudimentary, and their underlying structure may strongly influence the simulation results.

The paper by Andrew Hughes Hallett (Newcastle University and CEPR) provoked similar responses, even though he based his analysis on the econometric rather than the oracular approach. In 'Commodities, Debt and North-South Policy Interactions', Hughes Hallett examined the benefits of cooperative policies using an existing empirical model, the Marquez-Pauly econometric version of Lance Taylor's 'structuralist' model. Taylor's model is designed to highlight the relationship between growth rates in the North and South, based on asymmetries in the economics of the two regions. The North is characterized by rigidities, surplus capacity and labour market slackness; the South has a single commodity export whose price adjusts flexibly in world markets. Hughes Hallett used the model to examine whether cooperation over commodity policy, would improve the debt and growth prospects for developing economies. He argued that commodity policy and debt management may be considered alternative forms of policy coordination, which may provide welfare gains for both North and South.

Hughes Hallett's analysis was based on an existing empirical model which incorporates economic growth, in contrast to the simulation model presented by Currie et al. But the scope for spillover effects between North and South in Hughes Hallett's model was limited. Although the model includes spillover effects through the North's demand for commodities, there was no forward- looking behaviour in commodity and financial markets. In addition, the 'commodity policy' analysed was limited to cooperative agreements over the level of oil prices: the analysis needed to be extended to other commodities which are relevant to non-oil-producing LDCs. Furthermore, the scope for analysis of debt management in Hughes Hallett's paper was hampered by the absence of financial markets in the Marquez-Pauly model. Thus, while the model allowed the impact of debt management on Southern growth to be assessed, the consequences of such a policy for the North (through its repercussions in financial markets) could not be considered.

Ralph Bryant argued that the analysis of policy coordination presented by Hughes Hallett and by Currie and his co-authors may be somewhat premature in the absence of empirical models which incorporate the important macroeconomic linkages between North and South. Debt management and economic growth require proper modelling of financial markets and the determinants of North- South capital flows. In particular, one cannot treat net capital flows to the South as exogenous, Bryant argued. Michael Dooley and David Vines agreed, but argued that it is not sufficient merely to model the supply and demand for international capital flows: any attempt to model the debt problem adequately must also take account of the use to which the capital flows are put in the South, i.e. the South's consumption and investment decisions.

Other papers at the conference focused on specific aspects of North-South interdependence rather than simulations of more comprehensive models. Two dealt with the commodity market linkage. Christopher Gilbert (Institute of Economics and Statistics, Oxford) and Andrew Powell (St Catherine's College, Oxford) presented a paper entitled 'The Use of Commodity Contracts for the Management of Developing Country Commodity Risks'. Many LDCs depend on a single commodity for their export revenues; volatile commodity prices can have a severe effect on the stability of export revenues and hence on these countries' ability to finance essential imports.
Gilbert and Powell noted the recent dissatisfaction with intergovernmental agreements on commodity price stabilization, due not only to the technical difficulties but also the political controversies involved in operating such agreements. The authors contrasted such agreements with 'market' solutions to the risk- management problem faced by LDCs. In particular, they focused on the possibility of 'hedging' export earnings using futures markets and commodity options. One simple options hedge available to a commodity producer is to purchase a put option. If the price of the option is equal to the current price of the commodity the producer is ensuring that he will obtain not less than the current price at or before maturity. Put options at prices other than the current price will be significantly less expensive, but give a lesser degree of price insurance. Gilbert and Powell argued that an appropriate hedging strategy for credit- constrained LDC governments involves a combination of 'put' purchases and 'call' sales. This is preferable to hedging in futures markets, which involve higher costs because of margin calls.

The authors used a hypothetical simulation, to illustrate how the use of market instruments such as put and call options could have benefited Zambian copper export revenues over the 1980-4 period. The simulation revealed that the use of sufficiently long-dated options could offer substantial benefits to LDCs, but the costs to LDCs in terms of lower average earnings may be substantial unless some agency (such as the IMF or World Bank) were to underwrite the performance risk of such options.

Existing LDC debt consists almost entirely of claims in which debtor governments are committed to interest and principal payments whatever their current circumstances. Gilbert and Powell therefore considered ways in which the nature of debt instruments could be altered to alleviate the debt management problems which LDCs have faced in recent years. There may be advantages to both borrowers and lenders in adopting 'commodity- contingent' debt instruments, in which repayments are tied to commodity prices. Such instruments, the authors argued, could offer lower default risk and a more appropriate sharing of risk between borrowers and lenders.

In his discussion of Gilbert and Powell's proposals, Jorge de Macedo suggested they had placed too much reliance on the options markets to stabilize revenues. Few such options markets or commodity-contingent instruments have emerged spontaneously to date, he argued, and the authors had presented a convincing argument in favour of public sector intervention to encourage the formation of such markets. Michael Dooley observed that options cannot entirely eliminate risks as long as the producer's output is uncertain. Chris Gilbert agreed but argued that output variability is more significant for agricultural products than for other commodities.

Hossein Samiei (University of Essex) also considered the problem of commodity price variability in his paper, 'Financial Constraints in Trade Between the Oil-Exporting and the Industrial Countries: A Switching Regression Analysis'. Samiei presented an econometric analysis of how oil price movements have caused financial constraints on trade between the North and the oil- producing South. Samiei's regression results indicated that in the 1970s oil price shocks seemed to impose financial constraints on the North, while the 1980s created financial constraints on trade at a global level. Although the paper was presented as an econometric analysis of historical events, the disequilibrium mode of analysis developed could have applications to other studies of North-South interactions.

In his comments Youngil Lim (UNIDO, Vienna) observed that although Samiei's study was valuable as a piece of historical investigation, its policy implications were unclear. David Currie agreed that the regression results were suggestive but questioned their robustness, particularly given Samiei's assumption that the terms of trade variable which entered the regressions was exogenous.

More historical evidence on both commodity and income terms of trade was presented by D T Nguyen (University of Lancaster), in his paper on 'Movements in the Terms of Trade of Commodities Versus Manufactures: Effects on LDCs'. Nguyen presented evidence supporting the view that movements in the income terms of trade have adverse effects on the trade balance and growth prospects of LDCs. In discussion John Spraos (University College London) and John Black (University of Exeter) argued that Nguyen's use of the terms of trade to explain economic performance may be inappropriate: factors such as technical innovation and the encouragement of entrepreneurship, which boost growth, may also in turn affect the terms of trade. Thus a simultaneous model may be more appropriate to study these questions.

The importance of international financial markets in the linkages between North and South was a recurrent theme of the conference. In a paper on 'Global Savings, Investments and Adjustment: On Micro- and Macroeconomic Foundations of North-South Interdependence', Rob Vos (Institute of Social Studies, The Hague) examined issues involving the terms and availability of external finance to the South. Vos illustrated the possible effects on the South of Northern fiscal and monetary policies using flow-of-funds identities. Credit rationing at a microeconomic level as well as the macroeconomic effects of fiscal policy meant that Northern macroeconomic policies could have a large potential impact on interest rates and credit availability in the South. If the availability of credit is severely constrained, the South may face adjustments in its absorption in the short run. If capital is readily available, large capital flows into the Southern economies may bring about real exchange rate appreciation and the symptoms of the 'Dutch disease'.

In discussion Alasdair MacBean (University of Lancaster) and Val Koromzay (OECD) both pointed out that any analysis of North-South financial linkages from which reliable policy conclusions are to be drawn should take account of particular conditions in individual LDCs. Jorge de Macedo and Michael Dooley observed that credit rationing models should be treated with care, as it is not necessarily a disequilibrium phenomenon. Furthermore, in the presence of many different lenders and borrowers, credit rationing may not be a problem, unless all countries face constraints at the same time.

The effects of the availability of external finance on LDC growth were analysed for the case of Brazil by Winston Fritsch (Catholic University, Rio de Janeiro) in 'Brazil's Growth Prospects: Domestic Savings, External Finance and OECD Performance Interactions'. Fritsch used simulations from a 'two-gap' model, popular in the development literature, which indicated that Brazil's growth prospects were adversely affected in the early 1980s by the reluctance of foreign banks to lend to Brazil and the fall in the share of GDP devoted to capital formation. The simulations also suggest that fluctuations in world interest rates have a larger short-term impact on Brazil's growth prospects than changes in OECD growth. A reduction in US interest rates (through an appropriate reformulation of US fiscal and monetary policy) may therefore do much to alleviate Brazil's debt problems.

In the subsequent discussion, Michael Dooley pointed out that Fritsch's model did not include a treatment of the incentives to invest. Similarly, David Vines pointed out that the analysis lacked a proper treatment of the split between consumption and saving (which took into account the effects of permanent income and wealth) or of movements in the terms of trade. The behaviour of aggregate demand is essential to understanding whether the economy grows along the 'capacity-constrained' or the 'externally constrained' path.

David Vines concluded the conference by noting that the papers provided a useful agenda for further research into North-South interactions. In particular, the aggregation issue merited further consideration. It was important to consider in both theoretical and empirical models whether it was appropriate to subdivide the economies of the South into different regions or categories and, if so, what criteria should be used. More detailed research was needed into each of the wide range of North-South linkages, which operate through primary commodity prices, balance of payments flows, financial markets and the terms of trade.

A number of the conference participants also travelled to the Paris headquarters of the OECD on 21 September for a one-day workshop. The meeting discussed the implications of the issues raised at the CEPR/IESG conference for model-based approaches to the study of global interdependence. It also considered the agenda for future research on North-South issues, and in particular questions such as the appropriate method of aggregating the Southern economies and the data problems involved in modelling and identifying the channels of influence between the two regions. Researchers from the World Bank and the IMF described their comparative advantages in model-building: ready access to data and country-specific knowledge. Hopes were expressed that there would be increased cooperation between the OECD and other international agencies, as well as academic researchers and CEPR, in modelling the interactions between North and South.

The papers and proceedings of this conference, edited by David Currie and David Vines, will be published by Cambridge University Press in 1988.