Abstract
This paper argues that many developing countries may find it difficult to buttress disinflation programs purely through the adoption of traditional credibility-enhancing devices (such as monetary anchors and central bank independence), owing to “technical problems” (for example, high instability of money demand, increased capital mobility) and an insufficient endowment of credibility in the political institutions. In these cases, borrowing credibility from an outside agency like the International Monetary Fund may be the most effective solution. The paper discusses the different options that would allow the Fund to support programs aimed not at external adjustment—the Fund’s traditional role—but at disinflation.
Well-designed institutions can improve economic performance, but some institutions may be feasible or effective only at specific levels of economic development. […] Precommitment must represent more than a resolute inner stance - the precommitting agent must act through the external environment.
Many features of central bank laws in developing countries have been copied from the central bank charters of developed countries. However, due to different degrees of compliance with the law between the two groups of countries, legal independence is not as effective in insuring price stability in developing countries as it is in developed countries.
I. Introduction
1. Ever since its establishment, the IMF has made its lending conditional on the existence of a “balance of payments need” (henceforth called simply “BOP need”). That is, IMF programs have typically been geared to restoring a viable payments position, defined as “a current account deficit not larger than that which could be financed by net capital inflows on a sustainable basis, on terms compatible with the development and growth prospects of the country” (Polak, 1991, p. 17).
2. The IMF Articles of Agreement were drafted at a time when extensive capital controls, and the consequent relative immobility of capital, were taken for granted and member countries were committed to pegging their exchange rate. In such an environment, external and internal imbalances were highly correlated. A country running expansionary policies would likely face both an increase in inflation and a widening of the external current account. The latter would rapidly result in a loss of reserves, calling for policy adjustment.
3. In the present world of freely mobile capital and floating exchange rates, however, the correlation between external and domestic disequilibria is likely to be lower (see Section II), particularly in the short run. Copious private sector finance makes it possible to finance current account deficits of a size and duration unheard of since the Gold Standard era. Moreover, loose domestic policies need not result in an external current account imbalance, if the nominal exchange rate adjusts accordingly.
4. An internal imbalance that has received considerable attention in recent years is a high and sustained rate of inflation. Indeed, by now a broad consensus exists among both economists and policymakers that inflation is detrimental to growth, and that successful disinflation (that is disinflation that involves low output losses) requires “credibility”. The fact that inflation is bad for growth explains why the Fund—an institution whose ultimate goal is to promote sustainable growth—should regard the fight against inflation as one of its goals. The fact that successful disinflation requires credibility explains why the Fund could play a key role in backing up disinflation programs in developing countries: in these countries, in which acquiring credibility through the mere announcement of traditional inflation anchors (the money supply, the exchange rate) is comparatively more difficult, the countries’ interests might be served best by borrowing credibility from the Fund.
5. Of course, in the conduct of its daily business, the Fund has already partially adjusted to the new reality, as shown for example by the increasing attention paid in publications by the Fund staff to the relationship between inflation and growth performance (see, for example, IMF, 1996; Sarel, 1996; Christoffersen and Doyle, 1998). It seems to us, however, that the role that the IMF can play in sustaining disinflation programs has not yet been fully explored. The present paper intends to help fill this gap by developing the arguments above and discussing their institutional implications, particularly as regards the practice of conditionality and the Articles of Agreement regulating the Fund’s activities.
II. Three features of inflationary processes
6. The view that inflation is detrimental to growth—the first feature of inflationary processes that is relevant for our argument—has been convincingly supported by several empirical studies, some of which find that a negative relation between inflation and growth holds also at relatively low inflation rates (see Feldstein (1997), Sarel (1996), and the references therein). That view has also been endorsed by the major international economic organizations (see, for example, International Monetary Fund, 1996, Chapter VI; and Bank for International Settlements, 1997, Chapters VIII and IX) and by policymakers: the “Declaration on Cooperation to Strengthen the Global Expansion” adopted at the conclusion of the Forty-Third Meeting of the Interim Committee of the Board of Governors of the IMF held in Madrid on October 2, 1994 (reproduced in International Monetary Fund, 1994) regarded the attainment of price stability “as a condition for sustaining medium-term growth.”
7. The recognition that inflation hampers growth provides the rationale for the Fund to be involved in disinflation programs, as the Fund is institutionally interested in supporting policies conducive to sustainable growth (see Section VI).
8. A second feature of inflationary processes—namely that in the current world of flexible exchange rates high inflation and external imbalances are only loosely related—is relevant to understand why Fund’s involvement in programs aimed primarily at disinflation would constitute an innovation. If inflation were always, or at least typically, associated with external disequilibria, the Fund’s involvement in disinflation programs could take place within its traditional field of action. This was indeed the case in the Bretton Woods world of fixed exchange rates: in that world, high inflation would likely lead to an unsustainable real appreciation and to an external imbalance. However, in a world of increasingly flexible exchange rates, the relation between inflation and external accounts has become more blurred. Figure 1 illustrates this development: the cross-country correlation between inflation and external current account position, which was high during the 1960s, has become low after the breakdown of the Bretton Woods system in 1971. The correlation is low not only between current (annual) inflation and current external position, but also between inflation and the external position after one or two years: high inflation does not seem to lead systematically to immediate or future external disequilibria.2
Developing Countries: Coefficient of Correlation Between External Current Account Balance in Percent of GDP and Inflation, 1966-96 1/
Citation: IMF Policy Discussion Papers 1998, 012; 10.5089/9781451969337.003.A001
Sources: International Financial Statistics.1/ The coefficient of correlation has been computed using all the information available in each year in the IMF’s publication, International Financial Statistics. Therefore, the size of the country samples tends to increase in lime, as new countries start providing information on external accounts and inflation.9. A third relevant feature of inflationary processes—a feature that explains why the Fund’s action can raise the chances of successful disinflation in developing countries—is that bringing inflation down without incurring significant output costs requires credibility. The credibility conundrum—which has dominated the debate on monetary policy over the last twenty years—relates to the fact that, under the present monetary standard based on intrinsically useless currency, the supply of money can be expanded to whatever level by fiat. “Credibility” involves persuading private agents that the flexibility inherent in this particular type of money will not be exploited either to achieve short-run, i.e. transitory, output gains, or to serve the private interests, such as electoral gains, of those in charge of monetary policy. In other words, to attain a low-inflation equilibrium under a fiat standard, people must be made to believe that whenever the monetary authority announces a target, every action thereafter will be undertaken with the purpose of achieving the target. If a disinflation program is credible, inflation expectations will adjust rapidly, thus minimizing the risk that disinflation is initially associated with rising real wages and a loss of competitiveness and output. It stands to reason that, if policy-makers perceive that disinflation is likely to involve limited output costs, they will be more likely to undertake it.
10. How can a government increase its credibility in fighting inflation? This is no easy task, as the government is the ultimate enforcer of whatever rule, or institution has been set in place to buttress the credibility of monetary announcements. Credibility could be achieved through a sizable up-front investment in reputation on the part of monetary authorities aimed at establishing a strong track record. However, this option has the obvious drawback of requiring a sunk cost potentially so large as to be politically unpalatable.
11. In the rest of the paper we argue that alternative means to achieve credibility—traditional monetary anchors or institutional devices such as central bank independence—are likely to be less effective in developing countries (Section III), and that, if this proves to be the case, borrowing credibility from abroad, which for our purposes basically means the Fund, can be the next-best alternative (Section IV).
III. The announcement of Traditional Inflation Anchors is Likely to be Less Credible in Developing Countries
12. Disinflation strategies in industrial countries during the last quarter of a century have often relied on the announcement of inflation anchors (such as monetary or exchange rate targets) or on institutional devices to raise the credibility of purpose of the authorities (central bank independence, inflation targeting frameworks). These devices are likely to be less effective in developing countries than in industrial countries for a number of reasons.
13. Consider first the use of monetary targets. Monetary targeting frameworks require a degree of stability (or predictability) of real money demand that is hard to find in practice. Although shared by industrial countries, especially as a consequence of financial innovation, the problem has been particularly pronounced in developing countries. In fact, Figure 2 (top panel) shows that real money balances have been persistently more volatile in developing countries than in industrial countries over the last 25 years. This higher overall volatility is due to a higher volatility of both real output growth (central panel) and money velocity (lower panel). The latter probably reflects two factors: the role as store of value that cash and bank deposits play in developing countries; and, the continuous process of financial innovation that characterizes financial systems that are still catching up with those of industrial economies. Thus, the two key parameters of Fisher’s price equation are more volatile in developing countries, making the use of monetary targets less effective.3
Volatility of Real Money, Output, and Money Velocity, 1970-96 1/
Citation: IMF Policy Discussion Papers 1998, 012; 10.5089/9781451969337.003.A001
Sources: International Financial Statistics; and staff calculations.1/ The figures are the standard definition of real broad money, real GDP, and broad money velocity from ten-year rolling trends ending in the reference year. The country sample includes all countries for which information is available in International Financial Statistics, published by the IMF.14. As an alternative to monetary targeting, developing countries could rely on institutional arrangements such as an independent, and relatively “conservative,” central bank, or inflation targeting.4 Such devices—together with legal constraints limiting central bank credit to the government—are certainly useful. However, they are subject, when applied to developing countries, to a number of limitations that may severely hamper their effectiveness. First of all, Masson, Savastano and Sharma (1997) have recently argued that freedom of monetary policy from fiscal dominance and absence of commitment to other nominal anchors such as wages are necessary prerequisites for achieving credibility through domestic institutional reform. In such countries seignorage is often an important source of financing for the government, while a vast consensus among the population on low inflation as the overriding objective of monetary policy may be difficult to achieve. As it happens, both prerequisites are hard to meet in many developing countries. Furthermore, both inflation targeting and central bank independence build on a resource that is scarce in many developing countries, namely an inherent confidence in institutions—a kind of capital that needs time to be accumulated. In countries in which political and economic institutions have been subject to drastic changes in the recent past, it is also more difficult for the authorities to achieve credibility through institutional devices. A case in point would be the countries in transition from communism. Another example is represented by countries whose democratic institutions were shattered by military coups.
15. Yet another alternative would be to rely on external anchors, such as an exchange-rate peg or a currency board. It is not clear, however, why reliance on such anchors should be per se more credible, if the institutional environment is defective, than the adoption of purely domestic anchors. Moreover, developments in the 1990s, including the recent foreign-exchange crises, have shown that pegging the exchange rate involves more significant risks than in the past owing to increased capital mobility. It is therefore not surprising that pegging the exchange rate has become over time less and less frequent, including as part of stabilization programs (Cottarelli and Giannini, 1997; Quirk et al., 1995). As to currency boards, they are ordinarily even more demanding than ordinary pegs. As such, their effectiveness is predicated on the existence of additional sources of credibility. Is it not telling that most existing currency boards have been introduced as part of IMF programs?
IV. The IMF as Provider of Credibility
16. Economic theory has pointed to the role that delegation can play in acquiring credibility. As noted, however, delegating monetary policy to a domestic entity may not be very effective in developing countries. An alternative way of increasing the credibility of adjustment policies is to submit those policies to the approval of a supranational organization, such as the Fund.
17. In ordinary circumstances the Fund is playing more and more the role of provider of credibility, rather than lender of resources, to governments involved in external adjustment programs (Cottarelli and Giannini, 1997; Dhonte, 1997).5 This view is supported by three pieces of evidence. First, the share of net IMF credit over total net external financing (including foreign direct investment) to developing countries dropped from 4¼ percent during the 1980s to less than 1 percent during 1990-96 (based on data published in the World Economic Outlook of the Fund),6 Second, the number of so-called precautionary IMF programs has rapidly increased in recent years: at end-1996 about one-third of the stand-by and EFF arrangements were precautionary, that is, they had been undertaken by the authorities without any intention of drawing.7 Third, the ratio between actual and potential borrowing in all outstanding IMF arrangements—excluding those off-track—has declined since the early 1980s, while the number of countries with programs has risen to historical, peaks (Figure 3). Thus, although the interest in Fund programs has increased over time, the relative importance of Fund lending has declined in parallel. In sum, the Fund’s stamp of approval has been used by countries to enhance the authorities’ credibility in association with adjustment programs predicated on the availability of private international savings.
Transactions of the Fund, 1955-97
(In percent)
Citation: IMF Policy Discussion Papers 1998, 012; 10.5089/9781451969337.003.A001
Sources: IMF Annual Reports, 1955–97; and Transactions of the Fund, 19971/ The figures refer to the outstanding amount borrowed from the Fund at the end of the program over the maximum amount that could have been borrowed, for all programs that started in the reference year and ended by December 31, 1997. The drop registered in 1996 is likely to overstate the actual decline as it was computed over the more limited number of programs that started in 1996 and were completed by 1997.18. Could the Fund’s stamp of approval also strengthen the authorities’ credibility in adjustment programs aimed at disinflating the economy? This question raises two issues. The first relates to the mechanism through which credibility is enhanced: in programs aimed at external adjustment the system of incentives is such that there are costs, for both the borrower country and the Fund, if the program is not successful; this raises the credibility of the program. Will similar mechanisms be at play in disinflation programs? The second issue is whether the existing Articles of Agreement regulating the activity of the Fund would allow it to support programs that are aimed purely at disinflation. These issues are taken up in the following two Sections.
V. What Makes IMF-Supported Adjustment Programs Credible?
19. There seem to be four different mechanisms ensuring that failure to implement the conditions imposed by a Find-supported program will be associated with a loss for the parties involved (the Fund and the country’s authorities), and thereby that the program will be undertaken only if both parties believe it is going to work.
20. The first mechanism has to do with the direct financial loss the country would incur if the program were interrupted. The notion of conditionality—according to which a country can borrow from the Fund only under certain conditions (meeting certain intermediate policy targets and receiving the IMF Board’s approval on the completion of the program’s periodic reviews; see Guitián, 1995)—was indeed developed to strengthen the likelihood that adjustment would take place through the introduction of an explicit penalty for noncompliance.
21. The second mechanism operates because the program’s failure may be interpreted by the market as a signal of the country’s unwillingness or inability to meet external obligations. This mechanism, which may be expected to be more effective the greater the catalytic role of Fund’s finance, operates mainly through risk premia, but may also have an impact on the quantify of finance a country can raise in international financial markets.
22. The third mechanism has to do with the fact that Fund programs, being highly visible and carrying the signature of an independent and multilateral agency, raise the political costs of reneging. By signing a letter of intent with the Fund, a government raises the stakes both for itself and for the country at large (as opposed to the announcement of an adjustment program not supported by the Fund), by making any infringement of the agreed course of action visible, verifiable, and reprimandable.
23. The fourth mechanism works on the Fund’s side. As demonstrated by the oft-heard criticism of excessive conservativeness leveled against the Fund, this institution has over the last fifty years acquired a reputation for extreme cautiousness in granting approval of adjustment policies. This reputation represents precious capital that the Fund is unwilling to jeopardize by approving programs that are likely to fail. The private sector recognizes this, which adds credibility to the programs approved by the Fund. Moreover, the Fund in most cases buttresses its reputation by putting its own resources at stake. Lending clearly acts as a powerful incentive for effective monitoring (Rodrik, 1995). However, nothing suggests that it is a logical necessity. Rodrik himself, for example, remarks that the insistence of multilateral lenders that their claims be senior to private claims undercuts the signaling value of their exposure. It may be worth recalling in passing that the balance-sheet concerns of central banks had to be considerably attenuated, through the removal of the profit motive or outright nationalization, before they could be regarded as dependable guardians of monetary stability (Goodhart, 1988). Today, nobody would seriously argue that the credibility of a central bank as monetary authority or lender of last resort rests on its putting its own financial resources at stake. If anything, the opposite would be closer to the truth. Thus, we prefer to join ranks with Guitián (1992) in regarding Fund lending as an additional safeguard of credibility rather than the hinge on which programs revolve, since reputation forces may work quite effectively even absent explicit pecuniary incentives.
24. How would these mechanisms operate in the case of programs aimed at disinflation? Clearly, the first two mechanisms would be much weaker, as the program would not be undertaken with the primary purpose of obtaining external financing (either directly from the Fund or from private international markets). However, direct economic costs would still be sustained by noncomplying countries. Unlike the case of countries where monetary authorities are highly credible—in these countries news of unexpectedly high inflation may lead to a strengthening of the exchange rate—in developing countries the information that a Fund-supported disinflation program is off track is likely to result in downward pressures on the exchange rate and the country’s risk premium.
25. Moreover, reputational forces constraining country authorities (the third mechanism above) would still be at work. But would the Fund feel equally constrained? After all, the Fund’s reputation has been built on programs aimed at external adjustment. As Schadler et al. (1995) argue, the inflation record in countries with Fund-supported programs is “at best mixed.” This, however, cannot per se be taken as evidence that the Fund cannot work as an disinflation anchor, nor that the reputation capital built in external adjustment programs cannot be spent for disinflation programs. Past programs were primarily addressed at external adjustment. Moreover, nothing in the Articles of Agreement suggested that the institution would be held accountable for the inflation performance of members relying on its finance, a point we shall return to in the next Section. With inflation playing at best a role of secondary target, in case of conflict between external equilibrium and inflation control the latter was necessarily expected to give way. Perhaps more importantly, the reputation the Fund has acquired as strict enforcer of its traditional programs could be spent on the inflation front, once it is recognized that the “code of conduct” enshrined in the Fund’s charter includes price stability as a constituent element. Thus, it is not clear that making disinflation the explicit goal of some programs would entail a need to build a new stock of reputation capital.
26. A final point relates to the interaction between disinflation programs and external adjustment programs. Failure to deliver in the first type of programs would likely be regarded by the Fund as indicative of the country authorities’ inability to stand by their commitments (unless this was due to force majeure). As such it will increase the likelihood that, should the need for an external adjustment program arise, Fund support might not be forthcoming. The possibility of pre-screening countries for access to more rapid financing in case of external crisis has been discussed in the aftermath of the Mexico crisis. Clearly, the failure of a disinflation program—even if without immediate financial consequences—could greatly impair a country’s likelihood of receiving financial support in case of need.
VI. Consistency with the Existing Articles of Agreement
27. Several questions must be answered in order to assess whether Fund-supported programs aimed at disinflation would be consistent with the current Articles of Agreement.
28. The first question is whether supporting disinflation programs falls within the Fund’s overall mandate. Assessing this issue carefully is crucial not only to provide a legal basis for the Fund’s involvement, but also to avoid altering radically the specialized nature of this institution—a nature that represents one of its key assets. The goal of price stability at present does not figure prominently in the Articles.8 However, the Fund’s involvement in disinflation programs can be justified on several grounds. As discussed above, inflation is ultimately bad for growth, and Article I(ii) of the Articles of Agreement states that one of the purposes of the Fund is to contribute to the “promotion and maintenance of high levels of employment and real income and to the development of the productive resources of all member as primary objectives of economic policy.” This article refers to the achievement of the above goals through “the expansion of balanced growth of international trade.” However, Article IV—as amended following the 1976 Jamaica meetings—recognizes that “the essential purpose of the international monetary system is to provide a framework … that sustains sound economic growth.” As a result, over the last two decades it has become common practice to regard one of the Fund’s roles “to help its member countries establish a macroeconomic environment, in their own economies and in the global economy, that will foster such [i.e., “private-sector-led, sustainable] growth” (Camdessus, 1995, p. 2). Of course, all this does not imply that the Fund’s mandate involves supporting any policy program aimed at raising growth, as this would alter fundamentally the scope of action of the Fund. The Fund’s involvement can be justified only insofar as it is consistent with the specialized nature of the Fund as a monetary institution. Disinflation programs appear to meet this criterion. Indeed, Fund’s involvement can also easily be justified by Article I(iii), which indicates promoting exchange stability among the purposes of the organization. While exchange stability could in theory be consistent with inflation rates that are equally high across countries, in practice exchange stability is more likely to be achieved when inflation is low across countries (one reason being the strong empirical correlation between inflation and its volatility).
29. The second question relates to the form in which the Fund could provide its endorsement of disinflation programs. There are three alternatives, differing in terms of degree of Fund commitment, and hence in the degree of credibility associated with them.
30. As a first option, disinflation programs could be formulated under one of the existing forms of collaboration between the Fund and its membership that do not involve a formal endorsement by the Fund Board of a country’s policies—such as the so called staff-monitored programs, shadow programs, and programs formulated under the procedure for enhanced surveillance.9 Lack of formal endorsement by the Board and lack of commitment of Fund resources would reduce their significance in terms of provision of credibility. Indeed, these programs are relatively infrequent and do not attract much attention from the media.10
31. A second alternative would be to set up a procedure that would involve a formal periodic endorsement of the program by the Fund Board (quarterly or six-month reviews), but without opening access to Fund resources: the outcome of the program’s review would not be allowing the purchase of a new tranche, but rather the confirmation that the Fund continues to regard the authorities’ policies as consistent with their disinflation goals. Such an arrangement would clearly be more effective than the previous option, and would have the advantage of not requiring changes in the Articles of Agreement (see below). Moreover, as noted above, even in the absence of pecuniary incentives, reputational forces would be at work to buttress the commitment of the parties involved.
32. The third option would be to have a full-fledged program, giving right of access to Fund resources. This would be, of course, the strongest option as a signaling device. It would also be more expedient, in light of the changes that would have to be introduced in the present procedures in order to make the IMF effective as a source of disinflation credibility. However, if Fund resources are to be involved, a revision in the Articles of Agreement is, in our view, needed. The reason is that Article V, Section 3(b)(ii) constrains the use of Fund resources to the existence of a BOP need. Note that a BOP need is not a condition for setting up a program; it is only a condition for borrowing, as the possibility of precautionary arrangements illustrates. However, a prospective or potential BOP need is regarded as a condition for a program.
33. One could argue that high inflation—possibly a symptom of underlying real disequilibria—might eventually result in an external crisis, and therefore gives always rise to a potential BOP need. But this argument is not supported by empirical evidence. Indeed, as highlighted in Section II, in a world of flexible exchange rates, the link between inflation and external imbalances is blurred. And, while there may have been cases in which the existence of a potential BOP need has been used to justify Fund-supported programs only loosely related to external adjustment, the argument cannot be abused. Stretching the interpretation of the Articles of Agreement would eventually challenge the transparency of the Fund’s operations and the foundations of its mandate.
34. Amending the Articles of Agreement is no easy matter, and rightly so. This is a strong argument in favor of less ambitious options. However, explicit recognition of the goal of price stability in the IMF charter would greatly strengthen the credibility of disinflation programs, as the IMF would as a consequence be held responsible for the inflation performance of members. The argument here is essentially the same that has led in recent years to emphasize this goal in the statutes of national central banks. It seems indeed rather odd, for a monetary institution such as the IMF is often purported to be, that price stability should be the one macroeconomic objective not explicitly mentioned in the “code of conduct” laid out in Article IV. The omission is not without consequence, if we are to judge on the basis of the less than satisfactory inflation record of traditional IMF programs. Moreover, it appears to be due less to a conscious decision than to the low weight generally attached to price stability at the time of the last major revision of the Articles.
35. There is an additional reason why programs aimed at lowering inflation would benefit from the introduction of an ad hoc facility; a country characterized by high inflation is likely to be discouraged from asking for Fund support if such support were to be automatically identified by the market as involving a BOP need. The last thing that a country in such a position would want to signal to the market is that it faces not only inflation problems, but also balance of payments problems.
36. One further issue concerns the amount of the resources that the Fund should make available. In the absence of a BOP need, which principles should be followed in determining the appropriate size of the resources to be made available to a country? This question is somewhat misleading. Currently, the potential gap in the balance of payments is only one of the determinants of the amount of resources made available to a country. What is more relevant is the signal given to markets through the Fund’s approval of a country’s policies. One recognized principle that the Fund uses in assessing the amount of support is that “the stronger the policies, the stronger the access to Fund resources.” The same principle could be the main guideline in determining the amount of resources to be made available for anti-inflationary programs.
37. Endorsement by the IMF of a disinflation program need not imply espousing inflation targeting. Indeed, one could easily envisage instances where adoption of a currency board, or even of an exchange-rate peg, fits better the member country’s needs and conditions than inflation targeting, because of greater visibility or lesser technical complexity. A case-by-case assessment will therefore have to be made.
References
Camdessus, Michel, 1995, “The IMF and the Challenges of Globalization—Three Addresses by Michel Camdessus,” Washington, October.
Cottarelli, Carlo, and Curzio Giannini, 1997, “Credibility Without Rules? Monetary Frameworks in the Post-Bretton Woods Era,” Occasional Paper No. 154, International Monetary Fund, Washington, December.
Christoffersen, Peter, and Peter Doyle, 1998, “From Inflation to Growth: Eight Years in Transition”, IMF Working Paper 90/100 (Washington: International Monetary Fund (August)).
Cukierman, Alex, 1992, “Central Bank Strategy, Credibility, and Independence. Theory and Evidence,” MIT Press, Cambridge, Massachusetts.
Dhonte, Pierre, 1997, “Conditionality as an Instrument of Borrower Credibility,” IMF Paper on Policy Analysis and Assessment, PPAA/97/2, International Monetary Fund, Washington, February.
Elster, Jon, 1995, “The Impact of Constitutions on Economic Performance,” Proceedings of the 1994 World Bank Annual Conference on Development Economics, Washington.
Feldstein, Martin, 1996, “The Costs and Benefits of Going from Low Inflation to Price Stability,” NBER Working Paper Series No. 5469, February.
Goodhart, Charles, 1988, “The Evolution of Central Banks”, MIT Press, Cambridge, Massachusetts.
Guitián, Manuel, 1992, “The Unique Nature of the Responsibilities of the International Monetary Fund”, International Monetary Fund, Washington.
Guitián, Manuel, 1995, “Conditionality: Past, Present, Future,” Staff Papers, International Monetary Fund, December.
International Monetary Fund, 1996, “World Economic Outlook,” Washington, October.
Killick, Tony, 1995, “IMF Programmes in Developing Countries,” Routledge, London and New York.
Masson, Paul R, Miguel A. Savastano and Sunil Sharma, 1997, “The Scope for Inflation Targeting in Developing Countries,” IMF Working Paper 97/130, International Monetary Fund, Washington, October.
Milesi-Ferretti, Gian Maria, and Assaf Razin, 1996, “Current-Account Sustainability,” Princeton Studies in International Finance, No. 81, October.
Obstfeld, Maurice, 1986, “Rational and Self-fulfilling Balance of Payments Crises,” American Economic Review No. 76, 72–81.
Polak, Jacques J., 1991, “The Changing Nature of IMF Conditionality,” Essays in International Finance No. 184, September.
Quirk, Peter J., et al., 1995, “Issues in International Exchange and Payments Systems,” International Monetary Fund, Washington, April.
Rodrik, Dani, 1995, “Why is There Multilateral Lending?,” NBER Working Paper No. 5160, June.
Rogoff, Kenneth, 1985, “The Optimal Degree of Commitment to an Intermediate Monetary Target,” Quarterly Journal of Economics No. 100, 1169–90.
Sarel, Michael, 1996, “Nonlinear Effects of Inflation on Economic Growth”, IMF Staff Papers, Vol. 43, No. 1 (March).
Schadler, Susan, Adam Bennett, Maria Carkovic, Louis Dicks-Mireaux, Mauro Mecagni, James H. J. Morsink, and Miguel A. Savastano, 1995, “IMF Conditionality: Experience Under Stand-By and External Arrangements, Part I: Key Issues and Findings,” International Monetary Fund, Occasional Paper No, 128, Washington, September.
C. Cottarelli is with the European I Department, International Monetary Fund, while C. Giannini works in the Research Department of the Banca d’ltalia. The paper was completed while the latter was visiting the Monetary and Exchange Affairs Department of the International Monetary Fund. The authors wish to thank Mark Allen, Tomas J. T. Baliño, Gérard Bélanger, Robert Feldman, Edward Frydl, Joshua Green, Javier Hamann, Esteban Jadresic, Christopher Lane, Eduardo Levy Yeyati, and Sunil Sharma for their helpful comments on a previous draft.
This summary evidence is confirmed by well-known cases of countries which have experienced prolonged periods of inflation in the absence of major external crises. For example, during 1988–97 the average current account deficit in Turkey and Uruguay were as low as ½ of a percent of GDP, with peaks of 3½ percent and 2½ percent of GDP, respectively, hardly unsustainable levels for developing countries, against annual inflation rates of 76 percent and 61 percent, respectively.
The evidence above refers to broad money. Similar findings apply to narrow money (Cottarelli and Giannini, 1997).
We regard inflation targeting as an institutional device as, in our view, its key feature relates to the introduction of institutional procedures that facilitate the monitoring of monetary policy actions.
The role of lender of resources is still very important, though, under extraordinary circumstances, as events in both East Asia and Russia have shown over the recent past.
The decline in the role of Fund financing is clear even excluding foreign direct investment, which has risen rapidly in recent years; based on data published in the 1997 issue of the Global Development and Finance of the World Bank, the share of outstanding Fund credit over total credit to developing countries fell from 3¾ percent in 1986 to 2¾ percent in 1996.
From a legal standpoint, there is no difference between precautionary arrangements and other arrangements: in either cases, countries can borrow from the Fund as long that the conditionality conditions are met. However, in precautionary arrangements the authorities’ intention not to borrow is formally stated. The precautionary status of a program is taken into account in computing the Fund’s short-term liquidity needs.
The original text did not mention price stability at all. In the context of the first amendment, in 1969, a reference was added to avoiding deflation and inflation among the considerations that should guide the allocation of SDRs (Article XVIII), With the Jamaica amendment, a reference to “reasonable price stability” appeared in the context of members’ obligations with respect to their exchange arrangements (Article IV).
Staff-monitored programs are programs monitored by the Fund staff on the basis of a quantitative framework agreed with country authorities. These programs are undertaken outside formal procedures approved by the Fund Board. Explicit Board-approved procedures exist instead for enhanced-surveillance programs. In these cases, the Board is involved in the monitoring, although, like in staff-monitored programs, it does not endorse the program.
In most of the cases, these programs are undertaken to establish a “track record” leading eventually to a formal program. When this is not the case, staff-monitored programs or programs undertaken as part of the enhanced surveillance procedure may be seen as a signal that policies are not sufficiently strong to deserve the commitment of Fund resources.
