Time for NATO to Close Its Door
The Alliance Is Too Big—and Too Provocative—for Its Own Good
Since 1974 - when the oil crisis hit and the world recession began - a number of developing countries that do not export oil (non-oil LDCs) have been borrowing heavily in the international credit markets. Surprisingly, private banks, not international agencies like the International Monetary Fund (IMF) and the World Bank, have underwritten most of this massive debt buildup, as the table on page 734 shows. And this shift from traditional to non-traditional sources of financing, along with the sheer magnitude of the borrowing, has led to criticism of private banks for their handling of the situation.
Banks, it is said, have overextended themselves in lending to LDCs. They have made bad loan decisions and are now being forced to absorb still more LDC debt in the hope of salvaging earlier credits. Major U.S. and European banks will therefore experience large losses on LDC loans, which would lead to a general financial crisis. To forestall that danger, official agencies should intervene to help reschedule loans. Henceforth, private bank lending abroad should be subject to closer official supervision and, in many cases, public lending should supplant private lending.
These criticisms and proposals raise two basic questions. The first is whether banks have acted prudently in lending to developing countries, evaluating risks reasonably and taking steps to protect shareholders and depositors from loss. The second question is what caused the explosive rise in demand for foreign loans by LDCs during the last three years. If the causes are persistent and LDCs continue to incur debt as rapidly as in the last three years, many of them may eventually have trouble servicing their debts. But if the causes are temporary, borrowing by LDCs will decline from present levels and most LDCs should be able to meet their obligations without difficulty.
This essay sketches the broad outlines of the debate on these issues and gives our reasons for believing that the extraordinary surge of demand by LDCs for foreign loans is a temporary phenomenon and that private banks have behaved rationally and prudently in responding to it.
Evaluating risk is the business of banking and a good banker is one who judges well and truly the degree of risk in a particular loan or class of loans. Since 1974, bankers seem to have sensed a substantial increase in the "country risk" of lending to developing countries - the probability that an LDC may be unable to service its foreign debt at some time in the future - and they have altered their strategy accordingly.
Country risk may be measured in several ways. We prefer a comprehensive indicator such as the ratio of outstanding foreign debt to national income or GNP, which focuses attention on the size of the debt relative to the total flow of resources out of which it must be serviced. Of course, a country's creditworthiness cannot be judged simply by reference to this or any other statistical ratio. A low debt-to-GNP or debt-service-to-exports ratio might mean high rather than low risk, if the low ratio reflected the country's past inability to borrow. Creditworthiness depends more fundamentally on intangible factors such as the government's political stability and its will and administrative ability to divert resources from current consumption and investment to meet its foreign obligations on schedule. The latter is principally a matter of the government's general economic management and more particularly of its fiscal, monetary and exchange-rate policies. An overly expansive domestic monetary policy, due perhaps to an overly ambitious development program, accompanied by a fixed exchange rate, will cause a country to lose reserves and its credit rating to fall, even if its existing debt burden is low. Per contra, a country with a high foreign debt burden, objectively measured, will be a "good risk" if lenders expect the economic authorities consistently to give a high priority to keeping the country's overall payments in equilibrium or surplus.
On balance, however, it is clear that the perceived risk of lending to LDCs has increased. As we see in the accompanying table, in the last three years, the net foreign indebtedness of the 16 non-oil LDCs that have done most of the borrowing grew at an annual rate of 15 percent, in real terms (i.e., allowing for inflation). In the same period, these same countries' combined real GNP grew about 4.5 percent per annum. Thus their external debt has been growing faster than their capacity to carry it, which raises the question of what private international lenders have done to protect themselves.
TOTAL LONG-TERM PUBLIC AND PUBLICLY GUARANTEED EXTERNAL DEBT OF SELECTED DEVELOPING COUNTRIES*
(billions of U.S. dollars)
End - 1973 End - 1974 End - 1975 End - 1976
Total Owed to Total Owed to Total Owed to Total
Debt Banks Debt Banks Debt Banks Debt
Argentina 2.89 .66 3.35 .63 3.16 .52 4.20
Brazil 6.92 2.60 9.30 4.40 11.50 5.80 15.00
Chile 3.04 .51 3.73 .55 4.00 .50 4.20
Colombia 1.94 .22 2.12 .26 2.36 .37 2.70
Mexico 5.42 2.36 8.08 4.24 11.25 6.66 14.00
Peru 1.44 .56 2.07 .95 2.67 1.34 3.30
China, Republic .96 .05 1.16 .14 1.69 .42 1.90
Korea 3.20 .31 3.98 .64 5.23 1.00 6.20
India 10.40 .03 11.24 .01 11.88 .02 12.20
Pakistan 4.30 .06 4.52 .08 4.89 .11 5.50
Philippines .81 .22 1.03 .28 1.28 .32 1.80
Thailand .44 .01 .51 .02 .62 .05 .80
Egypt 1.73 .19 3.89 1.08 6.31 1.21 7.30
Tanzania .46 .02 .61 .02 .79 .01 .96
Zaïre .89 .41 1.31 .66 1.68 .89 1.90
Zambia .57 .23 .68 .21 .95 .35 1.20
Total of above countries 45.41 8.44 57.58 14.17 70.26 19.57 83.66
Others 11.89 1.21 15.82 3.03 20.24 4.53 26.34
Total, 71 non-oil 57.30 9.65 73.40 17.20 90.50 24.10 110.00
SOURCE: World Bank and Citibank estimates.
* Debt contracted by the public sector of the borrowing country or by a private borrower with the guarantee of the public sector with an original or extended maturity of more than one year, repayable in foreign currency, goods or services. Such debt encompasses the bulk of a country's external debt, but not all of it. By definition, it excludes all short-term credits and all long-term non-guaranteed private borrowings. Adding these credits to long-term external public debt brings the total debt of 71 non-oil LDCs outstanding at the end of 1976 to somewhere around $160 billion. Data on public and publicly guaranteed debt are collected and published by the World Bank. Unfortunately, the data on the other types of debt are neither as reliable nor as comprehensive as those of the World Bank and are therefore of only limited value.
Banks respond in two ways to a change in the perceived risk of lending to a particular class of borrowers. On the one hand, they vary their interest charges on loans in accordance with differences in risk, as judged by the market. On the other hand, they diversify their portfolios so as to avoid the possibility that the failure of an important group of borrowers - those located in one LDC, for example - to meet payments on schedule would involve a substantial loss relative to the lending bank's capital. To this end, banks establish "country limits" - limits on the total amount of credit they will allow to be outstanding to borrowers in a particular country.
Commercial banks like other financial intermediaries exist for the purpose of assuming credit risks that their depositors would be unwilling to assume individually. To carry on the function of intermediation and risk-bearing, banks must charge borrowers a markup over the cost of funds, which includes a risk premium. The premium is analogous to the "pure insurance" element in an insurance premium. Its economic function is to provide an additional stream of earnings which a bank may draw on to compensate losses.
The credit judgments that underlie the rates at which banks are willing to lend are market judgments. Each lender makes his own decision whether and how much to lend, but it is the collective judgment of the market that determines the rate the borrower pays. Lending officers do not normally vary the rates they charge on the basis of individual borrowers' risk characteristics; nor do they make a loan they consider risky simply because a would-be borrower may be willing to pay more than the going rate for a borrower in his class. If the risk appears higher than normal, the bank will cut back the amount or term of the loan or refuse the loan entirely. The rates charged the various classes of borrowers, including the risk premia they contain, are determined by competition and arbitrage in credit markets - a process which is neither consciously controlled nor directly observable by lenders and borrowers. The underlying judgment the market makes is that the total amount of premium which a bank with a normal loan portfolio will receive will be sufficient over time, in the circumstances expected to prevail, to compensate for earnings the bank foregoes due to deferred payments of interest and principal.
From these considerations it seems to follow that an international bank with a loan portfolio that is relatively normal in its distribution among classes of borrowers will probably be adequately cushioned against the loan losses it will occasionally incur. This is true at any rate in the absence of a general economic catastrophe, such as a prolonged world depression, which falsifies fundamentally the expectations that underlie the market's assessment of risks.
The bulk of bank lending to non-oil LDCs during the last three years has consisted of Eurodollar or other Eurocurrency term loans provided by syndicates of U.S. and European banks. Such loans are customarily floating-rate notes priced by charging the borrower the cost of Eurocurrency funds to the lending banks - that is, the London interbank offering rate (LIBOR) at the beginning of each payment period - plus an additional percentage, known as the "spread over LIBOR." The accompanying chart covers syndicated Eurodollar loans extended to 12 important LDCs (Zaïre, Chile, Zambia, Korea, Egypt, Mexico, Taiwan, Colombia, the Philippines, Brazil, Thailand, Argentina) in 1974, 1975 and 1976. For each country and each year, it plots the size of the additional percentage, or "spread" against the country's current debt-to-GNP ratio. As may be seen, the spread varied more or less systematically with the country's debt-to-GNP ratio. While factors other than credit risk also cause spreads to vary among countries, the chart implies that loans to countries with higher debt burdens were seen by the market as presenting a somewhat greater risk of loss.
The chart also shows that the banks' evaluation of risks of lending to this group of LDCs increased over time; in other words, abstracting from the effect of rising debt-to-GNP ratios, spreads were generally higher in 1975-76 than in 1974. Evidently, the deepening recession caused lenders to modify earlier credit judgments and to require more insurance against risk of default. These findings suggest that the credit markets indeed responded rationally as the burden of LDC debt rose and world economic conditions worsened.
While the size of the risk premium contained in the rates charged LDC borrowers is often only a fraction of a percentage point, when it is multiplied by the principal amount of the loans involved, it yields a substantial addition to a bank's earnings stream which serves to insure the bank's capital and its depositors against loan losses.
Banks have also protected their capital from the risks of lending to non-oil LDCs by avoiding undue concentration of their total lending in any one country and in the non-oil LDCs as a group, by means of country and regional limits. In addition, they have obtained special guarantees on LDC loans from private firms or public agencies located outside the borrowing country.
Based on data published by a subcommittee of the Senate Committee on Foreign Relations (Church subcommittee), 21 larger U.S. banks that account for nearly all the foreign lending had each invested over one-third of their consolidated assets in foreign countries at the end of 1975. More than one-quarter of these credits went to non-oil LDCs, or roughly 10 percent of the 21 banks' total assets. No single LDC accounted for more than 2.5 percent of the total. The major part - some 75 percent - of the loans to non-oil LDCs were either covered by the guarantee of a U.S. government agency or a large U.S. firm or were short-term credits (less than one year) or credits in local currencies. In other words, only one-quarter of the loans to non-oil LDCs were term loans in U.S. dollars or other strong currencies, without special guarantees. At year-end 1975, such unguaranteed term loans to non-oil LDCs totaled $10 billion, or 2.5 percent of the 21 banks' total assets. Comparable figures for 1976 are not available for a large group of U.S banks. However, at the end of 1976, Citibank, which accounts for about one-fifth of total foreign lending by U.S. banks, had $3.5 billion of term loans payable in strong currencies, without special guarantees outstanding to non-oil LDCs - a figure equal to 5.7 percent of the bank's total assets.
EUROCURRENCY LOANS TO 12 NON-OIL LDC'S
In judging the adequacy of the measures banks have taken to protect their capital and deposits from the risks involved in lending to LDCs, it is important to observe that in recent years loans to LDCs have normally been made to, or guaranteed by, the government of the borrowing country. Governments that expect to continue to borrow for development purposes do not repudiate their obligations. A default, if it occurs, is likely to involve at worst a temporary suspension of payments on interest and principal, until a rescheduling of principal payments can be arranged. More typically, no technical default occurs (no payments are missed). The borrowing government simply requests the lenders to spread repayments of principal over a longer period.
In consequence, default or rescheduling does not mean that the total value of the loan is forfeit. The loss is only the opportunity cost of having to wait for interest and for a fraction of the principal until a settlement is reached and payments are resumed. And if there is no actual default but only a rescheduling, there is no out-of-pocket loss to the lender. His detriment is only having to extend a doubtful credit for a longer period.
However successful banks have been in meeting LDCs' credit needs during the last three years, the future remains uncertain. This is why the key question in the LDC debt debate is not how banks have behaved so far but what conditions they will encounter in the future. To answer this question we need to know why the debt buildup occurred and whether the causes will continue unabated in the future.
It is generally agreed that the rapid rise of LDC debt was due to a combination of three factors: the oil-price hike, the recession in the industrial countries, and overly expansive domestic policies followed by some LDCs. But there is no agreement on the relative importance of these factors. Some people believe that the main cause was the increase in the price of oil decreed by the Organization of Petroleum Exporting Countries (OPEC). Since the real price of oil is unlikely to fall, it is argued that many non-oil LDCs will continue to borrow heavily until their debt burden reaches the point where they are unable to service it. Another and more sanguine view is that the buildup of LDC debt can be traced mainly to causes that are temporary and self-limiting. According to this view, the primary cause was the severe cyclical recession and slow recovery in the industrial countries. The oil-price hike was a less important factor and most of its effect on LDCs' foreign borrowing was also temporary. Still another view is that the debt problem reflects a failure of many LDCs to "adjust" their domestic policies so as to reduce their payments deficits.
To resolve the debate, an analysis of the LDC debt problem must attempt to determine the relative quantitative importance of the oil-price rise, the recession, and domestic policies in LDCs as causes of LDC borrowing, both in the past and in the future. Moreover, the effect of these abnormal factors must be distinguished from normal development financing. We have tried to answer these questions with a simple statistical model, whose results are given below. First, however, we should explain more precisely how each of the factors contributes to developing countries' balance-of-payments deficits and foreign borrowing needs.
With regard to development finance, developing countries normally import more goods and services than they export - i.e., they have deficits in their payments balances on what is known as the current account - and finance the difference by importing capital from industrial countries in the form of aid, direct investments and loans. The abundance of savings relative to investment opportunities in advanced industrial countries and the opposite condition in developing countries - a relative scarcity of savings - causes this normal flow of capital from developed to developing countries.
As a result of this flow of development finance, foreign claims on developing countries are constantly increasing, and the debt-service and investment-service burdens of developing countries are constantly rising. In normal circumstances, however, this does not cause a balance-of-payments problem, nor does it impair the LDCs' international credit. For the inflow of foreign capital augments domestic savings and accordingly makes possible a higher rate of domestic investment and a more rapidly rising stock of productive capital. In other words, foreign borrowing and investment increase a developing country's total income and create additional resources out of which foreign owners and creditors can be compensated for the use of their money. Particular loans or investments may fail to yield the expected return out of which owners can be compensated or lenders paid. Occasionally, a borrowing country's entire development effort may be so poorly conceived or managed as to yield returns too small to provide the margin needed to service foreign debt obligations. But in the normal course of events, this does not occur.
Economic conditions in the last three years have been far from normal, however. Developing countries have been whipsawed by the 1972-73 boom, by the fourfold increase in oil prices in the winter of 1973-74, and by the recession. The boom temporarily strengthened LDCs' international payments. But, by temporarily accelerating their economic growth and raising their imports, it set them up for the one-two punch that followed: the oil-price rise and the deep and prolonged recession in the industrial countries.
The oil-price rise was like the imposition of a heavy tax on the income of oil-consuming countries, for the benefit of oil-producing countries. It transferred some $100 billion of real income in 1974 and somewhat smaller amounts in 1975 and 1976 from oil-consuming countries to the members of OPEC. The LDCs' share of this income loss amounted to about $10 billion in 1974, or 2.5 percent of their combined GNPS. The proportions in 1975 and 1976 were similar.
This income loss was sudden and unexpected. Users of oil in LDCs - private firms, households and public enterprises - had suddenly to increase their spending for oil, but they were unable or unwilling to reduce immediately other spending for consumption and investment to make room, so to speak, for higher oil bills. Thus, their first response was to draw upon existing cash balances and other liquid assets in order to cushion the impact of higher expenditures for oil on their living standards and investment programs. They also tried to restore their impaired liquidity by borrowing. The macroeconomic consequences of these microeconomic decisions were a sharp increase in the value of LDCs' oil imports, a corresponding increase in their current account deficits, a temporary deterioration in their overall balance of payments (i.e., a loss of foreign exchange reserves), a sharp rise in the demand for foreign credit and more borrowing abroad. The borrowing was mainly from U.S. and European banks whose funds had been augmented by OPEC deposits of "petrodollars."
The reader should note that these initial consequences of the oil-price hike were independent of any changes which governments in developing countries may have made in their general economic policies in response to the oil shock. They reflected the spontaneous reactions of economic units, public and private, to the impact of high oil prices.
As time passed, initial responses to the oil-price rise were modified. Economic units in LDCs gradually cut back other spending to make room for higher oil payments. They did so in order to rebuild cash balances and reduce their abnormal demand for credit; they could not postpone indefinitely a cutback in spending to match real incomes that had been lowered by the high price of oil. As this adjustment occurred, cash balances were rebuilt, the abnormal demand for credit receded and aggregate spending tended to slow down. In consequence, LDCs' total imports slowed, their current account deficits narrowed, their overall payments balances moved back toward equilibrium and foreign borrowing - so far as it was attributable to the price of oil - fell off. This process of adjustment to the oil shock appears to have taken about a year to 18 months.
A homely analogy may help to make this process clear. An oil-importing country's situation in 1974 may be compared to that of a family whose income has been curtailed unexpectedly and permanently when the breadwinner's wages are reduced. The family's response will not be to cut back spending immediately to equal its reduced income. Rather, savings will be liquidated and loans negotiated to ease the transition to a lower standard of living.
But cash reserves are limited and emergency borrowing is a temporary expedient. As the family's debt rises, the utility of further postponing the inevitable reduction in living standards is in time outweighed by the disutility of taking on a larger burden of debt. Sooner or later, the family - or its banker - will call the borrowing to a halt, and spending will be cut back. When the transition to a lower living standard is complete, the family's net worth will be lower by the amount of savings spent and additional debt incurred. This loss of wealth is the price paid to ease the transition.
So in the case of the LDCs that import oil, it has been advantageous to postpone the impact of higher oil prices on domestic consumption and investment, by spending foreign exchange reserves and borrowing abroad. But borrowing for this purpose is, by its nature, temporary. As the burden of debt service rises relative to a borrowing country's income, lenders perceive a higher risk in extending more credit to borrowers in the country, and would-be borrowers find it increasingly expensive to obtain funds abroad. The alternative of curtailing domestic consumption and investment becomes increasingly attractive. Sooner or later, the emergency borrowing ceases and domestic spending is held down to a level commensurate with income.
The completion of this short-term adjustment is not the end of the oil-price story, however. As long as some OPEC members lend or invest a part of their oil revenues abroad, oil-importing countries will have larger current account deficits than they would have had if the price of oil had not risen. Collectively, this deficit will be matched by - and financed by - the OPEC members' capital outflow. Individual countries may tap this flow directly, or indirectly via the Western money markets and international banks.
This consequence of the rise in the real price of OPEC oil has led some commentators to say that the foreign debt caused by the oil deficit must eventually exceed the ability of oil-importing countries to service it. But this proposition implies that the additional borrowing by oil-importing countries that is the counterpart of OPEC's surplus oil revenues makes no contribution to the borrowing countries' ability to carry foreign debt. It implies that such borrowing goes mainly to enlarge consumption in the borrowing country rather than for investment, thereby adding little to the country's capital stock and GNP. But why would borrowers go on borrowing indefinitely or lenders go on lending, knowing that the loans were making little contribution to the borrowers' ability to carry debt? Initially, as we have seen, a large part of the borrowing due to the oil shock went to cushion the impact of higher oil prices on consumption as well as on investment in the borrowing countries. But neither lenders nor borrowers could have allowed this sort of borrowing to continue for long. To do so would have undermined the borrowers' ability to borrow.
These points may be put in another way. As noted, the oil-price hike brought about a substantial and lasting transfer or redistribution of income from oil-consuming to oil-producing countries. Assuming that this redistribution of income has not greatly altered the proportions of income saved and consumed in the oil-consuming and oil-producing countries taken together, the transfer does not entail much change in the total world supply of savings. But it does mean that more of the world's saving is now done in oil-exporting countries and less in oil-consuming countries. Furthermore, we may assume that the shift in income from oil-consuming to oil-producing countries would not have a substantial effect on investment opportunities (that is, on the marginal productivity of capital) in the two groups of countries. On these reasonable assumptions, there would be a lasting tendency for capital to flow from oil-exporting to oil-importing countries, as a consequence of the higher oil price.
This residual, long-term effect of high oil prices on LDCs' current account balances and their need for foreign capital may be thought of as a moderate addition to their normal development deficits. It reflects a change in one of the structural factors determining international capital flows: namely, the distribution of income between oil-producing and oil-consuming countries.
Booms and recessions in the industrial world lead to cyclical fluctuations in the trade and current account balances of developing countries. In a boom, when output in the industrial countries rises faster than its trend rate of growth, LDCs' normal current account deficits diminish. Likewise in a recession, when the growth of output in industrial countries falls below trend, LDCs' current account deficits grow larger.
The explanation of this cyclical pattern is that developing countries' exports go mostly to industrial countries and consist in large part of tropical foodstuffs and industrial raw materials, for which industrial countries' demand is highly responsive to changes in income. On the other hand, LDCs' demand for imports, which come mostly from industrial countries, is a good deal less responsive to changes in LDCs' income. Thus when world income declines, LDCs' exports fall more than their imports, and their current account deficits widen. This result is intensified when income in LDCs falls less than and later than income in the industrial countries. In the 1974-75 recession, the growth of income in the industrial countries slowed down about a year earlier than in the developing countries. The recession in the industrial countries was also considerably deeper than in the LDCs. As a result, the cyclical deterioration in LDCs' current account balances was unusually large and prolonged.
The 1974-75 recession was caused by a turn to restrictive monetary policy in the developed world during 1974 and the early part of 1975. The marked slowdown in money growth in the face of rapidly accelerating inflation (due in part to the oil-price hike) caused monetary conditions to tighten sharply. This led to a very large surplus not only in the trade and current balances, but also in the overall balance of payments of the industrial countries vis-à-vis the LDCs. With LDC exchange rates mostly fixed or supported, their foreign exchange reserves declined, slowing the growth of their domestic money supplies. In this way, tight money in developed countries was transmitted to developing countries.
Tighter money in LDCs increased their demand for foreign credit. This led in turn to a sharp rise in foreign borrowing, as the recession and easing inflation in the industrial countries reduced domestic demands for credit and enlarged the supply of funds available for lending to LDCs. The supply of credit available for lending to LDCs increased still more in the course of 1975 and in 1976 as monetary policies in industrial countries eased to stimulate recovery. Thus monetary policy and the deep recession in the developed world led to a steep rise in LDCs' borrowing from industrial countries.
It follows from this analysis that, as recovery in the industrial countries progresses, this complex process will go into reverse. LDCs' current account deficits will diminish, their foreign exchange reserves will rise and their demand for foreign loans - insofar as it was caused by the recession - will shrink. This appears to be happening in 1977. The non-oil LDCs' reserves are now rising strongly and their current account deficits are falling. Bank lending to LDCs is also declining, to judge from U.S. bank statistics and the volume of private Eurocurrency credits in the first months of 1977. However, these welcome developments have come slowly because the recession was unusually deep and the recovery relatively slow. Some foreign borrowing by LDCs, due to the recession, will continue until recovery in the industrial world is complete.
In sum, our analysis suggests that the preponderant cause of the non-oil LDCs' large current account deficits and heavy foreign borrowing during the last three years was the recession, while the price of oil also made an important contribution at the beginning of the period. It must also be recognized that LDCs' current account deficits and borrowing requirements in this period were being steadily enlarged by world inflation, which averaged 8.9 percent a year in 1974-76 (in terms of U.S. dollar prices), and by a rising inflow of capital for normal development purposes.
These conclusions about the causes and probable duration of LDCs' current account deficits and foreign borrowing are based in part on a statistical model.1 But since statistical methods are hard to apply to LDCs - primarily because the data are so poor - model results must be interpreted with more than usual care. The major difficulty was that we could find only ten non-oil LDCs - Argentina, Brazil, Chile, Colombia, Costa Rica, Jamaica, Korea, the Philippines, Taiwan and Thailand - where relevant data covered the whole period 1965-74. Since any shorter time period would have greatly reduced the reliability of the results, we decided to limit the study to those ten countries. We cannot be absolutely sure, then, that our generalizations would apply to LDCs other than the ten chosen. However, at the end of 1976, the countries in our sample accounted for more than 56 percent of the outstanding foreign debt of 71 non-oil LDCs, and for a still higher proportion of LDC debt owed to private banks.
For these ten countries we found, not surprisingly, that borrowing for normal development purposes accounted for only 10 to 15 percent of the deterioration in their current account balances from 1973 to 1975. The increase in the price of oil contributed about 30 percent of the deterioration between 1973 and 1974 but only 10 to 15 percent of the deterioration between 1973 and 1975. Thus, the recession in the industrial countries bears the brunt of the explanation. The recession explains over 55 percent of the deterioration from 1973 to 1974 and nearly 75 percent of the deterioration from 1973 to 1975.
The model was also used to make a projection of the current account deficit (and therefore of the need for foreign financing) in the year 1980. This projection was based on the assumptions that GNP in the industrial countries (defined to include the members of the OECD) will grow six percent per year in real terms between now and 1980, that the price of oil remains constant in real terms and that inflation, measured in terms of U.S. dollar prices, will be six percent per year. The result was a projected current account deficit for the ten countries, in 1980, of about $15 billion, most of which is accounted for by normal development needs. And if anticipated future inflation is allowed for, the projected 1980 deficit expressed in 1975 dollars would be reduced to $10 billion. Such a deficit in the sample countries may be compared with the peak deficits of $10.8 billion and $12.2 billion, respectively, in 1974 and 1975.
Thus, the model appears to support the conclusion that LDCs' current deficits and foreign borrowing needs will decline in real terms from now on. Since developing countries will continue to grow, it follows that the LDCs' outstanding foreign debt will tend to decline relative to their GNPS. If so, worries about the LDC debt problem in an aggregate or overall sense should soon fade.
So far in this discussion we have attributed none of the LDCs' foreign borrowing to their own economic policies. Indeed, our model suggests that for the "representative LDC," this factor does not have much weight. The model includes a variable that measures the effect of domestic economic policies in LDCs on their current account balances, but this policy variable had no statistically significant effect on the current account balances.
This finding is of considerable interest. It contradicts a widespread opinion to the contrary, which we also shared before making the analysis. It probably reflects the fact that many LDCs are of small or medium size, with open economies and domestic markets rather highly integrated with markets in the rest of the world. Such countries have little to gain and much to lose from following monetary policies that are more expansive than those of the industrial countries. If they do so, holding their exchange rates fixed, the resulting losses in foreign exchange reserves will quickly offset the effect of the expansive policy on the domestic money supply. As a result, the expansive policy will have little effect on domestic income and employment.2
On the other hand, a large developing country, by adopting an expansive policy, can cause its domestic money supply to grow more rapidly even though the exchange rate is fixed. The country will lose reserves, but because the economy is less open, the loss of reserves will only partly offset the effect of the expansive monetary policy on the domestic money supply, income and employment. In consequence, during a world recession, a large developing country will be able to maintain a higher rate of real growth, and suffer less unemployment, by following an expansive monetary policy, if it is willing to sacrifice foreign exchange reserves.
As an alternative to spending foreign exchange reserves, the government of a large country can support the currency by official borrowing abroad - or by offering special incentives to domestic banks and enterprises to do so. Such compensatory or balance-of-payments borrowing, as it is called, can have the same effect on the exchange rate as direct support by the central bank. An advantage of such borrowing is that it serves the purpose - partly political or cosmetic - of keeping the gross revenues of the central bank from declining.
Among the non-oil LDCs, Mexico, Brazil, the Philippines and Korea appear to have followed highly expansive monetary policies along with fixed or supported exchange rates at one time or another during the last few years. They did so for various reasons. In Mexico's case, this policy began as early as 1971 and was a consequence of President Echeverría's ambitious development and reform goals. In Brazil, Korea and the Philippines, an expansive policy was mainly a response to the external shocks of the last three years. It was an understandable effort to cushion the impact of world recession and the price of oil on domestic economic growth and employment. Of these four countries, only Mexico undertook official or officially induced borrowing on a large scale, to avoid a decline in official reserves. The Korean authorities also did some balance-of-payments borrowing during 1974. In Brazil and the Philippines, the main consequence of expansive monetary policy and support of the exchange rate was to reduce official reserves - and perhaps to increase the perceived risk of lending to these countries - rather than to increase foreign borrowing.
Be that as it may, the pressures exerted by the international markets for bank credit and foreign exchange now appear to have persuaded these four governments to reduce domestic money growth and to allow their exchange rates to respond more fully to the forces of the market. In Korea's case, the policy shift came more than two years ago and resulted in a large gain in reserves in 1975 and 1976. In the other countries, the authorities are moving, or have already moved, to slow domestic monetary expansion, to reduce support of the exchange rate and, in Mexico's case, to scale down official or officially induced borrowing.
Discussions of the LDC debt problem often end with the conclusion that there is need for LDCs to hasten the process of balance-of-payments adjustment, by altering their macroeconomic and exchange-rate policies. Yet this prescription has little relevance to the main causes of the LDC debt buildup. The borrowing that was caused by the industrial countries' recession and the price of oil was the result of external forces and the responses of economic units in LDCs to them, rather than of LDCs' internal policies. Borrowing increased because economic units - households, private firms and public enterprises - in developing countries responded to the adverse impact on their liquidity of higher oil costs and declining export revenues by borrowing abroad. Borrowing of this kind cannot be attributed to any failure on the part of LDCs to adjust domestic policies to eliminate balance-of-payments deficits. The relevance of "balance-of-payments adjustment" to the LDC debt problem is confined to the cases where monetary and exchange-rate policies led for a time to large reserve losses or official borrowing.3
Diagnosis is prescription. People who believe that LDCs' indebtedness will continue indefinitely to rise faster than the debtor countries' capacity to service debt advocate a general restructuring of existing public and private debt, under international auspices. Those who believe that the debt buildup was the result of banks' eagerness to expand foreign loans at a time of slack loan demand propose more official surveillance of bank lending abroad. Those who attribute a large part of the debt to a failure of "balance-of-payments adjustment" on the part of LDCs want the International Monetary Fund to intervene more actively in LDCs' policymaking. Some also recommend a slowdown of economic growth as a remedy for LDCs' payments problems. And there are those whose misgivings about the rationality of private lending decisions lead them to the extreme conclusion that private banks should withdraw from longer term lending to LDCs, leaving that function to public agencies such as the World Bank.
The preceding analysis lends little support to such proposals. We have shown that as recovery in the industrial world continues, aggregate borrowing by developing countries will decline in real terms. Thereafter, LDC borrowing will consist for the most part of normal development finance augmented to some extent by the long-run consequences of higher oil prices. Proposals for general rescheduling or refinancing of LDC debts, which assume that the debt-service burden must inevitably reach unmanageable proportions, are founded more on fear than on sound analysis.
The burden of LDC debt, measured as a proportion of GNP, has been ratcheted up by the traumas of the early 1970s. But even in the case of the 15 or 20 countries that have done most of the borrowing, the burden so measured appears to be bearable except in the case of a few smaller countries. It is noteworthy, too, that the great bulk of the LDC debt is owed by the larger, more advanced countries, which are better able than smaller, less-developed countries to hold back domestic consumption and investment to make room for debt-service payments. Nor does it appear that banks generally have been imprudently eager to expand their foreign loans. The evidence cited above suggests that as the perceived risk of lending to LDCs increased, banks moved to protect their capital and deposits by raising the risk premia in their lending rates and tightening country limits.
In a more positive vein, if U.S. and European banks had not responded when LDCs' demand for credit rose so swiftly in 1974 and 1975, those countries would have suffered far more severely than they did. Foreign borrowing allowed many countries to absorb the oil and recession shocks with relatively small losses of output and employment. This experience hardly suggests that private banks should be shut out of the term lending to LDCs. Rather it suggests that private banks and public agencies such as the IMF have complementary roles.
The IMF enters the picture when a country has trouble servicing its foreign debt and difficulty in borrowing privately. When the problem is traceable to the debtor country's own macroeconomic policies, the IMF may be able to persuade the government to adopt a stabilization program in return for a standby credit and the IMF's seal of approval. This in turn helps create conditions in which the country's private creditors will be willing to reschedule their claims, in order to reduce its current debt service burden. However, the IMF is rarely able to intervene on its own motion in the economic policymaking of a member country, on the ground that Fund officials or other Fund members consider these policies likely to lead to balance-of-payments trouble. The agency must be invited to come in, and this is unlikely unless the country's international credit has already been affected. Until that point is reached, the only external discipline on a country's economic management is the discipline exerted by the markets for international credit and foreign exchange. In recent years, the effectiveness of that discipline has been demonstrated in the case of several developing countries.
A developing country should not be expected to slow its growth in order to damp demand for imports and thereby improve its overall balance of payments. A persistent overall balance-of-payments deficit usually reflects not a high growth rate but an overly expansive domestic monetary policy and a high rate of inflation, in combination with a fixed or supported exchange rate. The trouble can be cured in either of two ways. On the one hand, the country can adopt a severely restrictive monetary policy while maintaining the same exchange-rate policy. In that event, there will be - temporarily, at least - an adverse impact on economic growth and employment, until domestic inflation slows down. On the other hand, the country can allow its exchange rate to depreciate more or less freely, while at the same time reining in domestic monetary expansion gradually rather than abruptly. The falling exchange rate will immediately eliminate the overall payments deficit. Domestic inflation will accelerate at first as the exchange rate falls, but will then slow down, reflecting the gradual cutback in money growth. Meanwhile, there will be little sacrifice of real growth and employment. In our opinion, a gradual stabilization policy of this kind is the right remedy in those cases where an LDC's own policies are the cause of its debt problem.
It is true that developing countries' imports rise faster when their incomes grow faster. But the recent changes in LDCs' current account balances are mainly a consequence of cyclical changes in economic conditions in industrial countries. In present circumstances, this might suggest the conclusion that industrial countries should adopt highly expansionary fiscal and monetary policies for the LDCs' good as well as their own. If this could be done without serious risk of higher inflation, the prescription would be sound. Yet that is hardly possible. More money means more inflation and - sooner or later - another round of monetary restraint and another recession. LDCs and their foreign creditors would surely be better off with a more gradual decline in the need for foreign loans than with a quick decline now, followed by another round of recession-induced borrowing two or three years hence.
Right now, what the world economy needs most is a little peace and quiet. We need a long period of steady expansion of world output and trade, not another inflationary boom followed by another slump that would bring the LDC debt problem back to frighten the world again.
1 The statistical model is fully developed in the authors' unpublished memorandum entitled "Developing Countries' External Debt and the Private Banks" which is available on request. Briefly, the model relates fluctuations in the ten countries' current account balances, measured as a proportion of their nominal gross national products, to fluctuations in four variables: the value of their oil imports; the industrial countries' output relative to its trend; the trend growth rate of the ten LDCs themselves, and their own internal economic policies. Current account imbalances were measured relative to GNP in order to scale down the contribution of the larger countries in the sample; otherwise, these countries would have dominated the results. Current account balances were defined net of aid and other official "unilateral transfers." The data for the ten countries were combined to form one sample, which is the basis for the statistical estimates. The model's results, therefore, may be said to describe a "representative LDC" having the average characteristics of countries included in the sample.
2 In principle, a small country could avoid reserve losses by allowing its exchange rate to float. However, a floating rate can be disruptive in an open economy where the ratio of foreign trade to domestic output is high. Smaller countries are therefore inclined to maintain relatively fixed exchange rates, and to avoid highly expansive domestic monetary policies in order to safeguard their reserves.
3 At the present time there is probably more need for balance-of-payments adjustment, in the sense of a change in domestic economic and exchange-rate policies, in some of the OECD countries than among the non-oil LDCs. This is the case, for example, in Italy and some smaller OECD members, which have followed overly expansive monetary policies and supported their exchange rates by "balance-of-payments borrowing."