Donal J. Donovan
Some empirical evidence
The recent increase in the frequency and severity of external debt-servicing difficulties experienced by many developing countries has prompted numerous assessments of the nature and causes of the problems. Several economic factors, assigned varying degrees of importance, have frequently been held responsible; these include, among others, a deterioration in the external terms of trade, the recession in industrial countries and an accompanying rise in world interest rates, inappropriate economic and debt management policies on the part of debtors, and unforeseen and abrupt changes in lenders’ behavior.
These assessments tend to remain couched in rather general qualitative terms and, in most cases, do not appear to have been based upon explicit quantitative investigation. An empirical approach could be useful, however, in clarifying the bases for these views. Why, for example, have some developing countries clearly experienced considerably more severe debt-servicing difficulties than others? And to what extent have the sources of the debt problems differed among developing country subgroups that exhibit relatively more homogeneous economic characteristics?
An analysis of the evolution of major economic variables for the group of 22 non-oil developing countries that rescheduled their debts on a multilateral basis during 1981–82 sought to provide an empirical basis for assessing the role of the economic factors frequently cited as contributing to the emergence of the debt-servicing problems. Thus the study examined, in turn, the role of certain exogenous factors (namely, terms of trade changes, the world recession, and changes in concessional aid flows); aggregate macroeconomic variables (the external current account deficit, the volume of exports and imports, the rate of economic growth, and the domestic rate of inflation); major economic policy indicators (the growth of domestic credit and the money supply, some partial indicators of fiscal trends, and movements in the real effective exchange rate); and finally, variables relating to debt management policy (which included the rate of growth and maturity structure of outstanding debt, the composition of external borrowing, and—incorporating the effects of changes in world interest rates—trends in debt-service payments). The evolution of these variables was assessed during the five-year period ending with the year before the rescheduling occurred.
The analysis dealt only with the aggregative outcome experienced on average by rescheduling countries and thus masks differences in individual country behavior that may be quite significant. However, two subgroups of rescheduling countries were distinguished: those that had not relied significantly on external borrowing on commercial terms (Group I) and those that had (Group II).
Group I consisted largely of low-income sub-Saharan African countries that had relied very substantially on external financing from official sources. It included the Central African Republic, Liberia, Madagascar, Malawi, Senegal, Sudan, Togo, Uganda, and Zaïre, as well as Guyana and Pakistan. At the end of 1981 each of these countries had less than 25 percent of its disbursed external medium- and long-term public debt contracted at variable interest rates. On average, the proportion of the group’s debt incurred at variable interest rates was 9 percent.
By contrast, the 11 remaining countries that rescheduled debts—Group II—had much greater recourse to international financial markets. Apart from Nicaragua, the proportion of their debt incurred at variable interest rates averaged 52 percent. With the exception of Romania and Yugoslavia, all were in Latin America or the Caribbean (Argentina, Bolivia, Brazil, Chile, Costa Rica, Ecuador, Jamaica, Mexico, and Nicaragua). Compared with Group I, they had higher real income levels, a more diversified export base, and generally fewer restrictions on external current and capital transactions. (An additional country, Poland, rescheduled its debt during this period but was not included in this study because the relevant data were not readily available.)
To assess the factors that might help explain the relative incidence of debt difficulties, the data for the rescheduling country groups were contrasted for the corresponding time period with those of a comparator group—those non-oil developing countries that did not reschedule their debt in 1981–82. For each variable, a separate average was calculated for rescheduling and nonrescheduling groups. The median, rather than the mean, was employed in order to reduce the distorting effect that outlying observations might have had.
While not necessarily implying a causal link, the analysis nevertheless yielded useful evidence on those common elements that may have been present among the group of countries that encountered particularly severe difficulties. However, several qualifications ought to be noted. First, the general inferences drawn for a group cannot be applied to individual countries. Second, the analysis dealt only with major macroeconomic variables susceptible to quantitative analysis; thus, the roles played by microeconomic policy variables not readily quantifiable (for instance, inappropriate pricing or investment decisions) or noneconomic variables (such as political factors or regional “contagion effects”—that is, a situation where problems in one country affect lenders’ perceptions of the entire region) were not addressed directly. Third, a clear-cut conceptual distinction between countries that did and those that did not formally reschedule their debts during a given period did not necessarily exist; for example, informal “unilateral” rescheduling (most commonly reflected in the accumulation of arrears) may occur. Fourth, several countries not included in the 1981-82 rescheduling group subsequently did reschedule their debts; thus, any conclusions are relevant only to the incidence of rescheduling that actually occurred in the period under review. Finally, the analysis generally concentrated on measurable aspects of debtor country behavior, although the conclusions take into account some important aspects affecting the perceptions of lenders.
These qualifications notwithstanding, the study did shed some light on the extent to which certain factors may have been associated with the debt-servicing problems of the developing countries (considered as a group) that entered into a rescheduling. The main empirical findings were as follows:
• Analysis of exogenous factors suggested that so far as changes in the terms of trade, export demand, and net concessional aid flows were concerned, rescheduling countries on average were not affected any more adversely by international economic disturbances than were the countries that did not reschedule. Indeed, to the extent that the data indicated differences, Group II rescheduling countries, in contrast to the general tendency for countries in the comparator group, appeared to have experienced on average some slight improvement in their external terms of trade in the periods prior to rescheduling. With regard to concessional assistance, Group I reschedulers also appeared to fare somewhat better than those that did not reschedule.
• Macroeconomic performance in general, however, differed considerably between the two rescheduling groups. While the lower-income countries (Group I) had overall a somewhat larger external current account deficit to GDP ratio than average (Chart 1), a striking feature of their external performance was stagnant or declining real export growth accompanied by sharp cutbacks in imports (Chart 2). In turn, this was associated with very low overall rates of economic growth and relatively high inflation (even on the basis of official price indices).
Chart 1External current account deficit relative to GDP1
Source: IMF, data base underlying the World Economic Outlook exercise.
1Data shown are medians of individual country data for each group. In a small number of instances, ratios to GNP have been used instead.
Chart 2External trade indicators1
Source: Data base underlying the World Economic Outlook exercise.
1Data shown are derived from medians of individual country data for each group.
By contrast, the ratio of the external current account deficit to GDP for Group II rescheduling countries tended to be somewhat less than that of comparator countries. Over the entire five-year period prior to rescheduling, the real growth rates of both their exports and imports were somewhat lower than the average. In the year immediately prior to the rescheduling, these countries experienced on average a noticeably faster export growth, a significant reduction in imports, and a slowdown in economic growth. The ratio of their current account deficit to GDP, however, remained roughly unchanged. Rescheduling countries in this group also had consistently higher inflation rates than the average—a difference that increased significantly over time (see table).
(Change in percent, annual average)1
|Years prior to rescheduling|
|Money and quasi-money|
• So far as other macroeconomic policy indicators were concerned, both rescheduling country groups experienced more rapid rates of expansion in total credit, net credit to government, and the money supply than their comparators (see table). Available partial indicators also suggested a somewhat more expansionary fiscal stance than average, especially in the case of the lower-income group (Group I). Group II rescheduling countries also experienced a sizable appreciation in their real effective exchange rates in the years prior to the one in which the rescheduling occurred (Chart 3). While a considerably smaller recorded real appreciation was noted for Group I, this may well have understated the appreciation that actually took place, due to an underestimation of inflation by official consumer price indices. In the case of Group II rescheduling countries, indirect evidence also suggested that a substantial amount of unrecorded net capital outflows occurred—a phenomenon, moreover, that increased over time.
Chart 3Effective exchange rate indicators1
Source: Fund staff estimates.
1Data shown are medians of individual country data for each group.
2Calculated on the basis of trade weights for major trading partners. An increase in the index indicates a relative real appreciation of the currency.
3Refers to end June of the year of the rescheduling.
• On debt management policies, two findings were particularly striking. First, while the evolution of total medium-and long-term external public debt was not very dissimilar between rescheduling and non-rescheduling countries, there was for the former group a marked deterioration in the maturity structure of external debt owed to commercial banks. This deterioration was most noticeable when the evolution of short-term debt was viewed against indicators such as available international reserves (Chart 4), exports, or imports. It was especially marked for Group II countries that rescheduled their debts during 1982, for whom short-term commercial bank debt was of particular quantitative significance. Such an excessive recourse to short-term debt—which in many cases reflected a postponement of needed adjustment in the face of reduced access to medium- and long-term financing—meant that these countries were particularly vulnerable to unforeseen changes in lending behavior. Second, rescheduling countries exhibited higher than average (and rising) reliance on debt owed to private creditors and/or contracted at variable interest rates (Chart 5); they were therefore prone to be affected more adversely by the worldwide increase in interest rates.
Chart 4Short-term commercial bank debt relative to gross official reserves plus unused bank credit commitments1
Sources: Bank for International Settlements, The Maturity Distribution of International Bank Lending; and IMF, International Financial Statistics and data base underlying the World Economic Outlook exercise.
1 Debt with a remaining maturity of one year or less owed to foreign commercial banks.
2Data shown are medians of individual country data for each group.
The findings of the study thus suggested, albeit tentatively, that some distinctive features did increase the incidence of debt-servicing difficulties for particular country groups. For the lower-income rescheduling countries (Group I), the adverse impact of exogenous factors appeared, on average, to have been a little less than is sometimes suggested. Rather, a major underlying source of their severe debt-servicing problems appeared to have been their weaker-than-average economic performance in terms of real export growth, inflation, and economic growth. These trends, besides contributing directly to foreign exchange pressures, undoubtedly served to erode the confidence of lenders and donors alike. Moreover, to the extent that possible significant reasons for the above shortcomings could be identified, inadequate demand management and exchange rate policies were likely to have been contributing factors in many cases.
The overall assessment of the possible factors underlying the difficulties experienced by the higher-income rescheduling countries (Group II) was somewhat more complex. So far as the macroeconomic performance and policies of the group on average were concerned, over the entire period reviewed, the trends in the external current account (in relation to GDP), real export growth, and economic growth performance did not appear to have been strikingly different from those experienced by the comparator country group. Indeed, in the year immediately preceding that of the rescheduling, rescheduling countries appeared to have had, on average, a turnaround in some aspects of their external position, namely, trends in real exports and imports. At the same time, some slowdown in the rate of domestic economic activity occurred.
Despite these changes, however, the external current account deficit as well as net medium- and long-term external borrowing—both measured in relation to GDP—actually rose slightly on average for the group in the year prior to the rescheduling. An important contributing factor to this outcome was probably the adverse impact of the rise in worldwide interest rates on the servicing of the existing stock of debt. This reflected, in turn, the relatively greater reliance (compared to nonrescheduling countries) that rescheduling countries had on variable interest borrowing. In the year prior to rescheduling, two important economic variables, namely, the rate of domestic inflation and the real effective exchange rate, moved in a clearly unfavorable direction, while at the same time, substantial evidence indicated a marked increase in capital flight.
Nevertheless, since the Group II rescheduling countries were faced with foreign borrowing needs (measured in relation to their GDP) not on average very different from those of other countries, why were they unable to obtain the amounts of foreign financing required, so that rescheduling was precipitated? Unmeasurable political and regional contagion effects may have played some role (as well as the fact that the average outcome concealed a range of diverse individual experiences), but several additional aspects involving the perceptions of commercial bank lenders during this period need to be considered.
First, from the perspective of creditors, the absolute size of borrowing needs rather than financing requirements in relation to the debtor country’s GDP is likely to have been of crucial importance. Thus, although other smaller nonrescheduling countries had higher deficit/GDP ratios, it was probably inevitable that the prospect of having to continue to provide several countries in the rescheduling group with unprecedentedly large amounts—not only in the immediate period but also for several years ahead—would in any event have soon begun to run counter to lenders’ portfolio balancing considerations. This aspect is particularly relevant in view of the prior sharp rise in international bank lending relative to domestic lending and the concentration of bank claims in a number of important borrowing countries.
Second, concern over the absolute size of borrowing needs was probably greatly reinforced by the sharp fall in world inflation and in inflationary expectations in the early 1980s. Thus, even had the deficits of some of the largest debtors hitherto been judged marginally sustainable, the fall in world inflation and the associated shift to positive real interest rates were likely to have caused lenders to reappraise what constituted a sustainable deficit from the viewpoint of the borrowers’ medium-term debt-servicing capacities. This consideration was probably accompanied, moreover, by a realization on the part of lenders that their loanable resources (i.e., deposits) would grow far more slowly in an environment of reduced inflation.
Third, all of these developments coincided with some particularly unfavorable (and well publicized) aspects of debtor country performance, including, for example, accelerating domestic inflation and a significant real appreciation of the exchange rate. These factors were likely to have contributed directly to the phenomenon of capital flight, while at the same time undermining lenders’ confidence in the future ability of the borrowers to contain external imbalances even within their present levels (thus further affecting adversely their judgment as to what constituted a sustainable path of external borrowing).
Finally, it is also clear that the debt management policies of the Group II borrowing countries were an important additional element that greatly aggravated the difficulties. A distinguishing feature of the experience of these countries was the fact that large-scale recourse to short-term commercial bank borrowing had caused a very sharp deterioration in the maturity profile of their external debt. To a considerable extent, this increased reliance on short-term debt was symptomatic of their underlying problems, reflecting, in many cases, a reduction in access to medium- and long-term financing sources. Nevertheless, once the reassessment of lending attitudes on the part of commercial banks became widespread, these countries were left highly vulnerable to a sudden withdrawal of funds or even simply to a slowdown in the rate of growth of new lending. Thus, while the fact that the underlying situation was unsustainable would probably have become apparent in any event, there can be little doubt that inadequate debt management policies of borrowing countries greatly contributed to the abruptness and severity with which the 1981–82 crisis unfolded.
Finance & Development is available on microfilm from University Microfilms, P.O. Box 1346, Ann Arbor, MI 48106, USA, and on microfiche (English only) from Microphoto Division, Bell and Howell Company, Old Mansfield Road, Wooster, OH 44691, USA