Stabilization Programs and External Enforcement
Experience From the 1920's
  • 1 0000000404811396 Monetary Fund

Credibility and financing problems are important reasons why countries may seek to involve external institutions in the design and implementation of stabilization programs. In particular, governments may rely on external institutions to ‘enforce’ programs that would otherwise lack credibility. This paper analyzes six European currency stabilizations sponsored by the League of Nations in the 1920s. It emphasizes the means by which the League provided a ‘commitment technology’ and enforced compliance, thereby helping to ensure successful stabilizations. Empirical evidence indicates that countries with greater credibility problems relied more heavily on external enforcement to stabilize their currencies.


Credibility and financing problems are important reasons why countries may seek to involve external institutions in the design and implementation of stabilization programs. In particular, governments may rely on external institutions to ‘enforce’ programs that would otherwise lack credibility. This paper analyzes six European currency stabilizations sponsored by the League of Nations in the 1920s. It emphasizes the means by which the League provided a ‘commitment technology’ and enforced compliance, thereby helping to ensure successful stabilizations. Empirical evidence indicates that countries with greater credibility problems relied more heavily on external enforcement to stabilize their currencies.

I. Introduction

Since the advent of the international debt crisis in the early 1980s, many developing countries have been striving to stabilize and adjust their economies. More recently, Eastern European countries and republics of the former Soviet Union have been facing an equally challenging task: to achieve domestic and external stability while simultaneously transforming their economies from central planning to market-oriented systems. While some countries have undertaken adjustment and reform on their own, many others have embraced stabilization and structural programs endorsed by the International Monetary Fund, the World Bank, and other institutions. As a result, there has been a very substantial increase in the number of countries implementing stabilization and reform programs under Fund arrangements during the last decade, as well as in the level of financial resources committed by international organizations and bilateral donors. The widespread recourse of developing countries to these programs is unlikely to abate in the near future. While considerable progress has been achieved in resolving the debt crisis of the middle-income countries, many low-income developing countries, as well as formerly socialist economies, are now initiating adjustment and seeking financial assistance. It is, therefore, a propitious time to examine the role that external institutions can play in assisting an adjusting country, and to ascertain the general factors that can influence the success or failure of such programs.

For the adjusting nations and for the international institutions that have been involved in the process, there are many interesting lessons that can be drawn from past experience. One episode that is particularly rich in insights for today’s adjustment efforts is the currency stabilizations that were implemented in a number of European countries during the 1920s. Most of the features of these European post World War I stabilizations have already been extensively discussed in the literature. 2/ However, a central aspect of these episodes which has received much less discussion than the stabilization per se, is the extent of foreign assistance and external enforcement that were involved in these early but comprehensive macroeconomic adjustment programs. Yet this specific aspect is of considerable importance nowadays in the context of the stabilization and reform efforts being implemented by a number of countries in the 1990s. During the 1920s six European countries undertook macroeconomic adjustment programs under the auspices of the League of Nations: Austria, Hungary, Greece, Bulgaria, Estonia and Danzig. 3/ Nonetheless, there has as yet been little analysis of the reasons why these countries resorted to the League for external enforcement of the programs they implemented to stabilize their currencies while others, in fact the majority of the European countries, stabilized their currencies de facto without recourse to external loans or the discipline of the League. This paper addresses some issues that are raised by this experience. Our central theme is the role that an external institution can play in the adjustment process. In particular, we examine the effectiveness of the League of Nations in enforcing and making credible the stabilization programs that it endorsed during the 1920s.

In order to tackle these issues, we first review the theoretical justification for intervention by external agents in a macroeconomic adjustment program. The paper discusses a number of arguments that go beyond the standard explanation, which focuses almost exclusively on the technical advice and financial support that external institutions can provide. We find that credibility and financing problems play an important role in explaining reliance on external institutions. A major reason why countries may seek external intervention is because they lack the credibility or ‘reputation’ needed to ensure the success of their stabilization efforts. We show how an external agent may be able to provide a ‘commitment technology’ that enhances the credibility of the program. We illustrate the credibility problem and the role of external enforcement with a simple model, and provide some insights into the circumstances that make external involvement in a stabilization program more likely to occur. In particular, we argue that the extent of nominal instability is likely to be an important determinant that leads to external participation in a stabilization program.

Once the theoretical framework has been outlined, we analyze in detail the experience with the stabilization programs that were endorsed by the League of Nations during the 1920s. We emphasize the comprehensiveness of the economic reforms contained in programs implemented by the League. A brief analysis of the general macroeconomic accomplishments of the programs is also undertaken. Furthermore, because the degree of credibility of the commitment to stabilize the economy appears to have been a crucial element in a program’s success, we place special emphasis on the instruments of control that the League of Nations employed to enforce its programs. If a program is expected not to be enforced then it will not be credible; in turn, a non-credible stabilization program is less likely to be successful. An empirical assessment of the effectiveness of the enforcement by the League is undertaken based on the outcomes of several of its programs.

By contrasting the European currency stabilizations of the 1920s that had recourse to external enforcement with those that did not, we attempt to test the theoretical explanations that have been offered for the reliance of some adjusting countries on external institutions. We present empirical evidence in support of the view that the objectives of enhancing credibility and to resolving financing problems were important determinants of a national government’s decision to rely on an external agent to carry out its stabilization.

Policy implications can be drawn from this experience for both countries and external agents. Of course, the importance of fiscal adjustment and monetary discipline are again confirmed by a careful study of these episodes. But, in addition, analysis of the experience of the 1920s also highlights the necessity of undertaking fundamental institutional reform—a ‘regime change’—in order to establish the foundation for macroeconomic stability. Our analysis suggests that the League of Nations enforced such regime changes with remarkable success.

The plan of the remainder of the paper is as follows. Section II reviews the standard economic rationale for the involvement of external institutions in stabilization episodes. Section III analyzes the League of Nations’ programs that were implemented in Europe after World War I. It also compares those currency stabilizations that did rely on external enforcement with those that did not and endeavors to investigate empirically the factors that determined the recourse to external enforcement. Section IV concludes.

II. Theoretical Arguments for External Enforcement of Stabilization Programs

This section analyzes the reasons why external institutions may become involved in a country’s macroeconomic stabilization and reform program and the specific role that they can play in its implementation and enforcement. These issues have not been dealt with in detail by the theoretical and empirical literature on macroeconomic stabilization. For the most part the literature has been concerned with the causes of instability of nominal variables, as well as the proper way to carry out the stabilization and adjustment. 4/ Beyond the traditional explanation which stresses the technical advice that external institutions can give, two conceptually distinct but not mutually exclusive arguments have been advanced as an explanation of the reasons for a country’s recourse to external assistance in designing and implementing a stabilization program. The first argument stresses the support an external agent can bring to solve national authorities’ credibility problems; the second acknowledges that an external institution may provide financial resources that can ease the external position of a country experiencing external adjustment problems.

1. Credibility, external enforcement and financing

Beginning with the seminal papers by Kydland and Prescott (1977) and Calvo (1978), the theoretical literature on stabilization policies has emphasized the time inconsistency problem; that is, the conflict between the ex ante and ex post optimality of economic policies in rational expectations models. In the case of monetary policy, this problem occurs because the government has an ex post incentive to renege on an announced policy that was optimal ex ante. 5/ The time inconsistency problem undermines the credibility of any announcements that are made about future policy because rational private economic agents anticipate that the authorities may later have an incentive to deviate from the announced course of policies. Time inconsistency can impart an inflationary bias to monetary policy, thus making stabilization of the aggregate price level more difficult. However, the time inconsistency problem will only emerge if the authorities have the opportunity to reoptimize later, after announcing their policy stance; that is, only if they operate under a discretionary regime. On the contrary, if the government finds itself with a binding commitment that prevents the reoptimization process in the future, then the rational expectations policy that was optimal originally will remain optimal in subsequent periods. Thus under a commitment regime there are no surprises and the inflationary bias of monetary policy can be avoided. 6/ This kind of regime can be established, in principle, by imposing various institutional checks on the government’s conduct of monetary policy. 7/

The literature on repeated games provides some solutions to the time inconsistency problem. 8/ In this framework it is possible for a government to build a reputation as a low-inflation policymaker. Reputation is built up through repeatedly playing a low inflation strategy over a number of periods. Unfortunately, the reputation-building process that is found in the game-theoretic literature is not generally available to the authorities of a country that is suffering from a macroeconomic disequilibrium involving nominal instability. The existence of strong inflationary pressures in itself reflects the government’s lack of a reputation as a tough inflation fighter. When a government begins with a bad reputation, the task of building credibility will take so long that many governments, given the extent of economic disarray they face, are unable to survive until the reputation can be established. Under these circumstances, a government that wants to stabilize the economy in the context of a monetary regime characterized de facto by discretion may have to choose the alternative of implementing a change of regime; specifically a shift to a commitment regime that embodies consistent low-inflation policies. A change of regime requires a fundamental redirection of the policy strategy or the rules that determine the whole sequence of monetary and fiscal policies, and not simply an isolated shift in policy measures (Sargent, 1986). In order to change a monetary regime it is necessary to modify not only the nature of monetary policy itself, but also the expectations held by economic agents about that policy. 9/ In order for this alternative solution to be successful, it is necessary that both the announcement of and the commitment to a regime change are credible to economic agents. Only then will they adjust their inflationary expectations in conformity with the new monetary rules.

A stabilization program is said to “lack credibility” if it is not perceived as a definite and permanent change in the way economic policy is determined. If the credibility of a stabilization program is in doubt the attempt to stabilize can entail heavy costs in terms of higher unemployment, lower levels of economic activity, capital flight or higher interest rates. In these circumstances, the government may have to support the announcement of a regime change with some signaling behavior. This action is needed in order to influence expectations favorably by convincing economic agents that the ‘rules of the game’ have indeed been modified. If economic agents believe in the attempt, stabilization can be achieved at lower cost. It is precisely at this point that the presence of an external agent in the stabilization effort is justified: one way for the monetary authority to “signal” that it is really altering the policy regime is by tying its hands to a stabilization program that is endorsed by some external or international agency, especially some agency that has an established reputation as a firm enforcer of stabilization. Sachs (1989) and Edwards (1989) suggest that the role played by the external agent is to provide the government with a commitment technology, and hence to improve the chances of resolving its credibility problems. 10/

Of course, it is not necessary for the ‘commitment technology’ to be supplied by an external agent. For example, domestic institutions could be altered by setting up institutional checks that established a commitment regime that gave the monetary authority the credibility it required. In many countries, however, the combination of economic, social and political disorder, as well as institutional weaknesses is such that credibility problems plague not only the economic authorities themselves but also the economic policymaking institutions. The widespread existence of credibility problems confronting national governments suggests that achieving the needed commitment technology solely from domestic sources will often be difficult or infeasible. In such cases, reliance on an external institution may be a viable option for restoring credibility.

In most analyses, the commitment technology is simply assumed exogenously. It is possible, however, to formulate a weaker version in which the credibility is not assumed at the outset, but instead is derived endogenously in a (subgame perfect) equilibrium. This weaker version can be used to explain the involvement of an external agent or institution in a country’s stabilization program, provided the external agent is credible in the sense that it possesses all the necessary means to enforce adjustment and guarantee full compliance with the stabilization scheme. In order to deter the government from violating its commitments, the external agent must provide compliance incentives. In general, an optimal combination of these would include a combination of ‘carrots and sticks’ that constitute a credible threat to the government that fails to comply.

The intervention of an external agent can also entail some costs for national policymakers. To the extent that an external agent imposes simple rules for the behavior of national authorities rather than optimal rules that are contingent on the ‘state of nature’, these rules lack a discretionary regime’s flexibility to react to unanticipated shocks. Moreover, intervention by an external institution in a country’s domestic affairs is likely to generate a political burden on the government that requests the intervention. Accordingly, only in cases of extreme crisis will an external institution be sought; these cases are exactly the ones where credibility problems are most severe.

A second reason for recourse to an external agent to carry out a stabilization program is that the agent can provide the resources to finance an external imbalance, as well as serve a catalytic role in mobilizing other credits. Of course the credibility and financing arguments for seeking external enforcement, though different in nature, are not mutually exclusive. In order to obtain external resources it may first be necessary for the government to resolve its credibility problem so that external creditors are more likely to be willing to extend credits. 11/ The latter reasoning implies that the government’s credibility problem is multiple: it has a domestic aspect—the need for a government to establish credibility with its own citizens and economic agents—as well as an external aspect—the need for credibility in the eyes of foreign creditors and the international community. 12/ In practice, of course, the existence of credibility problems and the need for financial assistance are so intertwined that in the empirical section we will not try to test for them separately.

Most economists agree that the urgency of needed macroeconomic adjustment is greater for economies facing high nominal instability and large macroeconomic disequilibria than for economies with milder inflations and smaller imbalances. When macroeconomic disequilibria are large, a stabilization strategy that relies on reputation-building strategies alone will be more difficult to implement than it would be if the imbalances were small. It follows that the greater the degree of nominal instability, the more likely it is that the authorities will be inclined to stabilize the economy through a regime change, rather than through a gradual build-up of reputation. In particular, if the authorities are truly concerned about stopping inflation they will try to switch to a new regime that embodies low inflation rules or commitments; this, in turn, may cause them to seek enforcement of the adjustment program by an external institution. The greater the costs imposed by the nominal instability, the more likely it is that they outweigh the costs of intervention by an external agent.

In summary, the general explanation why some countries resort to external agents to stabilize their economies is that the latter may help to ease the adjustment in two ways. First, an external agent can restore credibility to a stabilizing authority by providing a commitment technology. Second, the external institution may lend new resources to finance the stabilization effort, thereby making the adjustment less drastic than it would otherwise have been. Both credibility and financing problems seem to be more pressing in high inflation countries; therefore, these countries are more prone to appeal to an external agent to stabilize their economies. It will be shown that both arguments seem to have played a role during the European stabilizations of the 1920s, helping to explain the reliance of some countries on external enforcement of their adjustment programs.

2. A model of credibility problems and external enforcement

A simple model is provided here to illustrate the credibility argument associated with external enforcement of a stabilization program. The model here will contrast the different outcomes in terms of monetary expansion and rates of inflation for two different regimes: commitment and discretion. The stabilization will be understood as a change of regime from discretion to rules. The framework stresses the role of inflation in financing a fiscal deficit, and is based on a simplified version of the model in Barro (1983). In this model there is a natural interpretation of the role of external enforcement. We use standard notation and the convention that upper-case letters denote levels of a variable (or parameter) while lowercase letters denote its logarithm.

In this model the government wants to find the rate of growth of the money supply μtmt-mt-1 that minimizes its loss function. In particular, the problem of the government is given by


where πt = pt-pt-1 is actual rate of inflation and πt+1e=E(pt+1|It)pt1 is the expected rate of inflation in period t+1 conditional on the available information at time t. The first two terms of the loss function (1) are intended to capture two different kinds of costs associated with inflation. The term in πt+1e represents the usual distortionary costs of inflation, while the term in πt reflects other costs, such as the so-called menu costs of changing nominal prices or the costs derived from the fiscal lag (Tanzi effect). The last term in (1) represents the benefit the government obtains from inflation; it is proportional to the revenue from inflation Rt defined as


The demand for real money balances is of the Cagan type, and given by


Taking first differences in equation (3), the inflation rate that clears the money market is given by


The rate of money growth μt can be found from equation (4) and substituted into the loss function (1) to solve the optimization problem faced by the government; alternatively, we can solve for πt and find the rate of money growth implied by (4). The assumption that θΓ > k1 + k2 implies that the equilibria are characterized by positive rates of inflation.

The solution for the discretionary regime is characterized by the fact that the government cannot make binding commitments about its future behavior; hence it will be solving (1) in every period. Under these circumstances the government will not take into account the effect of its current behavior on the future expectations of private agents, and it will take πt+1e as given. In other words, the government will be playing Nash. After substituting equations (2) and (3) into (1) and using the fact that ∂πt/∂μt = 1 from (4) and that πt+1e/μt=0 (because πt+1e is taken as given), the first order condition is


Private agents understand the government’s behavior. Therefore, to obtain the expected inflation rate the mathematical expectation operator E(| It) is used in (5) conditional on {Mt, Pt} ∈ It. The expression for πte can then be substituted back into the logarithm of (5) to find the inflation rate that will prevail under discretion. In this equilibrium the full solution is


On the other hand, in the rules regime the government is able to make binding commitments about its future behavior. Thus it will take into account the effect of its current decision on the future expectations of private agents when choosing the (constant) rate of monetary expansion that will be in place forever. In other words, the government will be playing like a Stackelberg leader. To find the rate of inflation under this regime it is easier to impose the equilibrium condition πt=πte=μt in the loss function (1) and then take the first order condition. To simplify the algebra, we use the approximation that log [(1+α) exp(-π)-α] ≈ -(1+α)π around zero, and then the full equilibrium for the regime under rules is


Using superscripts D for discretion and R for rules to denote the regime in place, it follows that


the rates of inflation and monetary expansion are lower under rules than under discretion, since we have assumed strictly positive parameters. The reason is simply that in the regime under commitments the government internalizes the consequences of its own actions on the behavior of the private agents. In particular, the government understands that by committing to a lower rate of monetary expansion it can alter the inflationary expectations of the private sector and hence achieve a lower rate of inflation.

After some manipulations it is easy to show that the following inequality prevails for the different equilibria:


the rules regime is Pareto-superior to the outcome under the discretion regime. The government is worse off with the equilibrium inflation rate obtained in the discretion regime than that under the regime with commitments.

In this model there is a straightforward role for external enforcement of a stabilization program. It is clear that a government that finds itself under a discretionary regime will want to switch to a rules regime. However, it can only do so by being able to make a credible commitment to follow a binding rule forever. If the external agent is capable of guaranteeing compliance with this commitment, then external enforcement restores the government’s credibility and the involvement of the external agent becomes an alternative to effecting a regime change domestically: the external agent provides the commitment technology. To sharpen the analysis further, assume that the presence of an external agent imposes a fixed and positive cost C on the government. This cost may be a fee paid to the external institution, a political penalty involved in accepting foreign intervention in domestic affairs, or the loss to the government from the diminution of its “sovereignty”. Then it is obvious that the government will seek external enforcement of the stabilization if


Note that, as argued above, the higher the inflation under discretion relative to the commitment regime, the higher will be the likelihood of an external agency’s involvement.

III. Stabilization Programs Implemented Under the Auspices of the League of Nations During the 1920s

The stabilizations in a number of European countries following World War I can be better understood in light of the previous section’s analysis of the reasons why a government might choose to rely on an external institution to enforce stabilization of its economy. In particular, it is important to determine whether the two arguments mentioned earlier can throw some light on the involvement of the League of Nations. During the 1920s, the League endorsed six macroeconomic adjustment programs, or “reconstruction schemes” as they were then called, supported with foreign loans and credits. In the cases of Austria (1922) and Hungary (1924) the programs tackled the financial reconstruction of two countries that had been devastated by war, dissolution, and hyperinflation. In the case of Greece (1923) and Bulgaria (1926) the initial involvement of the League was in refugee settlement schemes, and later it also took on a financial character (in 1927 and 1928, respectively). Other programs under the auspices of the League were those for the Mortgage Loan to Danzig (1925) and the Currency Reform Loan to Estonia (1927). In all six cases, the government of the country in need voluntarily appealed to the Council of the League of Nations for assistance to tackle a given macroeconomic problem. After an agreement was reached, the League endorsed the reconstruction scheme.

In order to assess the effectiveness of programs under the auspices of the League of Nations in resolving credibility problems, one needs first to understand the policy contents of these schemes as well as the means of enforcement that the League possessed. Together, the contents of each scheme and the enforcement of the program, determined the force of the commitment technology. The procedures and workings of the League’s adjustment programs are explained in detail in League of Nations (1930), but a summary of the relevant elements is provided below. 13/

1. The policies recommended by the League of Nations

In each financial reconstruction scheme the Financial Committee of the League of Nations followed closely the main guiding principles of the International Financial Conference that had been convened by the Council of the League in Brussels in 1920. The basic premise was simply that in order to stabilize a currency it was of paramount importance to achieve fiscal equilibrium. If the budget was balanced there would be no need for the central bank to extend credits to the government or to increase the note issue beyond the amount that was demanded at a stable price level. Thus the centerpiece of financial reconstruction was a set of fiscal and administrative reforms intended to eliminate the budget deficit. The reforms included a number of measures to increase tax and non-tax revenues, eliminate subsidies, reduce the size of the bureaucracy and cut other expenditures. In principle, the League only set targets for aggregate tax collections, leaving government officials free to decide what particular taxes were to be raised.

The administrative reform also comprised the establishment of domestic institutions and practices that permitted better control and monitoring of the public finances by a central authority. Among the institutions established or reformed were the central bank and other accounting or auditing offices. Through these reforms, policies to achieve fiscal balance were complemented with monetary discipline. The creation of money through note issue was to be checked, especially via advances to the public sector. In order to enforce this discipline the League’s schemes contemplated that the central bank of issue was to be constituted with private capital and independent of the government. To achieve the needed degree of monetary discipline, it was recommended that the central bank should try to maintain the following guidelines: (1) central bank independence; (2) monopoly of note issue; (3) prudence in lending and discounting operations; (4) limits on new advances to the government, (sometimes even an overall reduction in the stock of public debt held by the central bank); (5) centralization of the monetary transactions of all public sector entities in the central bank; and (6) proper cover and reserve backing for the note issue. 14/

In most cases the financial reconstruction included a currency reform whose main purpose was to consolidate the stabilization of the currency de jure by implementing a gold exchange standard. As is evident from this brief description, programs undertaken with the support of the League were very comprehensive in that the attack on inflation and currency depreciation was undertaken simultaneously on various fronts: fiscal, monetary, currency, financial, and administrative reforms were incorporated. Special emphasis was given to establishing institutional structures that would give the government both the needed incentives to stabilize and enough policy instruments to maintain sound public finances. In order to implement such comprehensive schemes governments were asked to appeal to their parliaments to grant special emergency powers to the executive branch.

The League of Nations stressed that the reconstruction schemes implemented under its auspices were not the same for all countries that solicited its assistance. But it acknowledged that when circumstances were similar, adjustment schemes were also similar. These similarities were evident for the financial reconstruction schemes followed by Austria and Hungary on the one hand, and for the refugee settlement programs and later financial restructurings of Greece and Bulgaria on the other. Other programs of the League, such as those for Danzig and Estonia, differed in both scope and content.

2. Structure of the League and its instruments of control

The ultimate responsibility for a financial reconstruction scheme under the sponsorship of the League of Nations fell on the Council of the League. However, the Council instructed either the Secretariat or the Financial Committee to carry out the detailed work. The duties of these organs included preliminary investigations, negotiations with the governments and institutions involved (both debtors and potential creditors), the design and formulation of the actual stabilization program, preparation of the draft resolutions, and supervision of the execution of the scheme. The Council also set up special committees to address certain international political problems that needed to be cleared up before the adjustment agreement—the Protocols—could be signed.

Of course, the League recognized that the execution of the program was as crucial as the design of a sound reconstruction scheme in restoring confidence in the adjustment effort. Moreover, the League was also keenly aware of the repercussions that sponsoring a stabilization program would have on its own reputation. 15/ For this reason, it took care to arrange various means to guarantee the enforcement of its program, as well as safeguard the interests of foreign creditors. Basically, there were three direct aspects of control included in the Protocols or other official documents of the League’s financial reconstruction schemes.

The first was the appointment by the Council of a League Commissioner to the country undertaking the League-sponsored stabilization. The Commissioner, whose tenure was to last until the reconstruction was regarded as completed, controlled two very important instruments: the external loan account and the security revenues account. The external loan account was extended to the debtor government by the foreign creditors to cover fiscal deficits during the transition period when the necessary fiscal reforms and adjustments were being implemented. 16/ The authorities of the country were not allowed to draw on this account without the consent of the Commissioner, who in turn would only authorize disbursements when the funds were to be spent to meet the agreed objective. The second account under the control of this League official was the one in which were deposited all the proceeds from the specified revenue sources that were set as security for the foreign loan. Security revenues usually included customs receipts or income from state monopolies. The Commissioner would set aside enough funds to service the external obligations, plus any precautionary margin, and refund the rest to the government only if the program was being carried out in conformity with the agreement. Furthermore, the Commissioner usually had the power to veto any government action that would endanger the value of the revenues assigned as security for the loan, or even to call for additional receipts to serve as collateral. The Commissioner was to be assisted in his duties by the domestic government and provided with all relevant information. This official had enough power to control the public finances and to influence the behavior of the government. In principle, he could be reappointed if the Council decided to reinstall control over the authorities when there was a risk that the achievements of the reconstruction scheme were being compromised. In modern game theoretic parlance, the presence of the Commissioner represented a credible threat to the government.

The second instrument of control by the League was the nomination of an Advisor to the central bank. Although this official was a bank subordinate and not part of the Financial Committee staff, in practice he maintained close contact with League representatives. His duties were established in the statutes of the central bank that were drawn up by the Financial Committee. They included supervising the sound administration of the central bank by the Boards of Management and Directors, in particular ensuring that there was not an unjustifiable expansion of the domestic component of the monetary base. To enforce this discipline, the adviser was sometimes entrusted with the power of veto in the Board of Directors.

The final aspect of control by the League of Nations was the trustees of the external loan, who represented the interests of the foreign bondholders. Their usual duties were minor, such as the delivery of interest payments due or control of the revenues that were directed at servicing the external debt once the Commissioner’s tenure had ceased. Their major role was in the case of default, a contingency that was foreseen in the official documents of the reconstruction scheme. If default occurred, the trustees could make the payment good out of the entire assigned revenues.

Other support functionaries with different control capabilities were sometimes appointed, such as experts to supervise the conditions of the foreign loan subscriptions, or special committees to address certain issues or inquiries into specific issues during the preparation and execution stages. As with the design of the financial reconstruction schemes, not all the programs had the same instruments of control. Again, variations were made depending on the gravity of the economic situation and institutional structure; not all the programs had a Commissioner and not all schemes entrusted the central bank adviser with the same powers. 17/

3. Effectiveness of external enforcement under the League’s stabilization programs

As discussed in Section II, the degree of credibility attached to a stabilization program depends on firm enforcement of the commitment to the new rules. Thus it is important to evaluate the League’s effectiveness in inducing an acceptable level of compliance with its stabilization programs. Although it is difficult to make a precise assessment of how effective this enforcement was, there are signs that the League was a tough enforcer of its programs. For example, there was no recorded instance where an appointed Commissioner had to withhold any portion of a loan in order to exercise the League’s prerogatives. Similarly, the legal veto powers of central bank advisers were never brought into play. Nor was it deemed necessary to apply sanctions on any government. 18/ In itself, the absence of such sanctions does not necessarily imply that the enforcement was fully credible. However, as we show below, the economic outcomes of the programs were so remarkable that one can safely dismiss the case of a non-credible enforcement.

Another clue to the credibility of the League’s enforcement is provided by one incident in the implementation of the Austrian program. In 1924 the Austrian government requested a release from the reconstruction loan for purposes other than that of meeting its budget deficit. The government wished to use some of the loan proceeds for productive investment in order to stimulate the Austrian economy. But the Financial Committee and the Commissioner denied such a release of funds, arguing that it was inconsistent with the reconstruction program and with ensuring the security of the bondholders’ investment. 19/

The degree of control exercised by the League was so strict that the number of countries willing to undertake its programs remained small. In 1928, for example, Portugal refused to sign such an agreement even though all the preliminary investigations and recommendations had taken place. The loss of sovereignty and discretion in the management of economic policy was unacceptable to the Portuguese government. 20/ The cases of Poland and Romania appear to be similar. The Polish authorities tried by all means to avoid the intervention of the League in the issuance of their stabilization loan, perhaps because they did not want to be identified with losing parties in World War I that already had programs with the League. 21/

The enforcement, monitoring, and compliance with of the original programs was, of course, far from perfect. In the case of Austria, for example, in 1924 the League considered that the reduction of expenditures and dismissal of government employees did not proceed at the expected pace. 22/ This delay in reform caused further program adjustments and an extension of the period of League’s control beyond the original schedule foreseen in the Protocols. Another indication of less-than-full implementation of the stabilization programs was the fact that the League was not fully satisfied with the ultimate performance of the Austrian and Hungarian central banks because the proportion of short-term paper in the assets of the banks was thought to be excessive.

The only major enforcement difficulty in the reconstruction programs came to light not during the 1920s but years later when, in 1931, Hungary, Bulgaria and Greece defaulted on their international reconstruction loans. However, by that time most of the financial reconstructions were regarded as completed and the League was not in a position to intervene effectively.

4. Macroeconomic performance under the programs of the League of Nations

This section provides a general evaluation of the macroeconomic outcomes of programs endorsed by the League of Nations during the 1920s. Clearly, the overall assessment of the program is favorable: they were able to stop the Austrian and Hungarian hyperinflations; resettle more than 750 thousand refugees in Greece and Bulgaria; establish or reform central banks in all six countries; and stabilize, de facto, six national currencies. A more detailed account of the performance of the economies under these programs is given in Figures 1 to 6 and Tables 1 to 6. Although only a small number of indicators are shown, they give a general sense of the macroeconomic performance of the countries that undertook the programs sanctioned by the League of Nations.

Figure 1.
Figure 1.

Austria: Exchange Rate and Price Level, 1919-1926.

(Logarithmic scale)

Citation: IMF Working Papers 1993, 003; 10.5089/9781451841824.001.A001

Note: Exchange Rate as quoted in New York; the Price level refers to retail prices.Source: International Statistical Yearbook, League of Nations, and Young (1925).
Figure 2.
Figure 2.

Bulgaria: Exchange Rate and Price Level, 1922-1929

Citation: IMF Working Papers 1993, 003; 10.5089/9781451841824.001.A001

Note: Exchange Rate as quoted in New York; the Price level refers to retail prices.Source: International Statistical Yearbook, League of Nations.
Figure 3.
Figure 3.

Danzig: Exchange Rate and Price Level, 1924-1928

Citation: IMF Working Papers 1993, 003; 10.5089/9781451841824.001.A001

Note: Exchange Rate as quoted in New York; the Price level refers to wholesale prices.Source: International Statistical Yearbook, League of Nations.
Figure 4.
Figure 4.

Estonia: Exchange Rate and Price Level, 1921-1929.

Citation: IMF Working Papers 1993, 003; 10.5089/9781451841824.001.A001

Note: Exchange Rate as quoted in New York; the Price level refers to wholesale prices.Source: International Statistical Yearbook, League of Nations.
Figure 5.
Figure 5.

Greece: Exchange Rate and Price Level, 1921-1929.

Citation: IMF Working Papers 1993, 003; 10.5089/9781451841824.001.A001

Note: Exchange Rate as quoted in New York; the Price level refers to retail prices.Source: International Statistical Yearbook, League of Nations.
Figure 6.
Figure 6.

Hungary: Exchange Rate and Price Level, 1921-1927.

(Logarithmic scale)

Citation: IMF Working Papers 1993, 003; 10.5089/9781451841824.001.A001

Note: Exchange Rate as quoted in New York; the Price level refers to retail prices for 1921-1923 and wholesale prices for the rest of the period.Source: International Statistical Yearbook, League of Nations and Young (1925).
Table 1.

Austria: Monetary and Fiscal Indicators, 1920-1927

article image

The figure for 1920 was estimated using fiscal years 1919/1920 and 1920/1921.

Source: Estimates based on data from International Statistical Yearbook, League of Nations; Memorandum on Public Finance; Memorandum on Currency and Central Banks, League of Nations (1926a) and Young (1925).
Table 2.

Bulgaria: Monetary and Fiscal Indicators, 1923-1930

article image

Fiscal year beginning that calendar year. The fiscal balance includes expenditure incurred in as a consequence of the war during 1923-25, and paid by advances from the National Bank. For 1928-1930 it includes special funds incorporated in the budget.

Source: Estimates based on data from International Statistical Yearbook, League of Nations; Memorandum on Public Finance; and Memorandum on Currency and Central Banks.
Table 3.

Danzig: Monetary and Fiscal Indicators, 1924-1930

article image
Source: Estimates based on data from International Statistical Yearbook, League of Nations; Memorandum on Public Finance; and Memorandum on Currency and Central Banks.
Table 4.

Estonia: Monetary and Fiscal Indicators, 1923-1930

article image
Source: Estimates based on data from International Statistical Yearbook, League of Nations; Memorandum on Public Finance; Memorandum on Currency and Central Banks, and The Estonian Yearbook, Pullerits (ed.).
Table 5.

Greece: Monetary and Fiscal Indicators, 1922-1929

article image

Fiscal year ending that calendar year. Proceeds and utilization of various loans which were not formerly entered into the budget were accounted for in 1929.

Source: Estimates based on data from International Statistical Yearbook, League of Nations; Memorandum on Public Finance; Memorandum on Currency and Central Banks, and Mazower (1991).
Table 6.

Hungary: Monetary and Fiscal Indicators, 1921-1927

article image

Fiscal year ending that calendar year.

Source: Estimates based on data from International Statistical Yearbook, League of Nations; Memorandum on Public Finance; and Memorandum on Currency and Central Banks. League of Nations (1926b) and Young (1925).

The first and clearest characteristic that is apparent from the figures is that the programs were all able to stabilize the exchange rate. Under the programs, all six currencies were brought onto a gold exchange standard. In the cases of Estonia and Bulgaria stability was achieved in 1924 (Figures 2 and 4), before the intervention of the League in financial matters, but was finally consolidated and made de facto under the auspices of the League. The behavior of the price level also reveals that the programs endorsed by the League were disinflationary. The general price levels in Austria and Hungary were stabilized abruptly in 1922 and 1924 respectively, putting an end to their hyperinflations (Figures 1 and 6). The cases of Greece and Bulgaria in Figures 2 and 5 also show how rates of price inflation were reduced; the price level actually fell in Bulgaria. Interestingly enough, some of the disinflation in the latter two cases began with the League’s intervention in the settlement of refugees (1923 for Greece and 1926 for Bulgaria). Although these schemes were not yet dealing strictly with a macroeconomic problem, the League insisted that governments should make every effort to balance their budgets. Price stability in Greece seems to have been reaffirmed with the subsequent financial reconstruction in 1928: inflation, which had exceeded 80 percent in 1923, was reduced from an average of 15 percent during 1925-1926 to around 7 percent during 1927-1928. In the cases of Estonia and Danzig no dramatic price instability was evident prior to the League’s intervention, and the programs preserved this stability (Figures 3 and 4).

One important feature of the League’s programs was the emphasis accorded to reestablishing sound fiscal and monetary policies. In order to make a currency convertible it was necessary to reshape the balance sheet of the monetary authority to a more conservative and orthodox composition. Tables 1-6 show how, as the central bank was being established or reformed in each country, the program involved a portfolio reshuffling away from domestic credit granted by the central bank towards an asset composition that entailed increases in holdings of gold and other external or metallic assets. In some cases the change in the balance sheet of the monetary authority was quite spectacular: the Austrian central bank’s ratio of gold and foreign reserves to note circulation went from almost nil to more than 75 percent by the end of the program (Table 1); conversely, the ratio of discounts by the central bank went from 100 percent of the note circulation in 1921 to only 34 percent by 1926.

With respect to rates of monetary expansion there are some interesting features. First, in cases where there was initially hyperinflation, money growth was substantially reduced but was never eliminated. In fact, the money supply grew 81 percent in Austria during 1923 (Table 1) and 57 percent in Hungary during 1925 (Table 6). The monetary expansion immediately after the stabilization of the exchange rate and the price level permitted a replenishment of real money balances. Second, in the cases where there was no initial hyperinflation the rates of monetary expansion were kept under control for most of the stabilization period, except for small increases in Bulgaria, Estonia and Greece during the reform years. The money supply increased 29 percent in Bulgaria during 1928 (Table 2), and 27 percent in Estonia during 1927 (Table 4), while the circulation of notes increased almost 15 percent in Greece during 1928 (Table 5). Given the relatively stable behavior of prices, these rates of money growth may also be explained by upward shifts in the demand for real money balances induced by the new regime, owing to lower anticipated inflation. Finally, it is also worth mentioning that the rate of monetary expansion was substantially reduced in Greece well before the financial intervention of the League began in that country. In fact, the reduction in the rate of expansion of the note circulation, from 49 to 4 percent after 1924 (Table 5), coincided with implementation of the refugee program that was also under the control of the League of Nations.

The fiscal data that are presented in the tables tend to confirm the movement towards more conservative macroeconomic policies, including balanced budgets. This is particularly striking in the cases of Austria and Hungary (Tables 1 and 6) where budget balances were substantially improved. In other cases the movement was not so drastic. Nevertheless in some cases, such as Greece in 1929 (Table 5), accounting practices distort the fiscal data so much that a precise assessment is very difficult.

IV. Determinants of External Enforcement: Empirical Evidence

The theoretical discussion of Section II predicts that countries with more severe credibility problems and more pressing financial needs should be more prone to rely on an external agent to help stabilize the economy. Moreover, it also argues that both kinds of problems should be more acute in economies with high nominal instability. This section provides some empirical evidence on the characteristics of countries that relied on external agents during the European currency stabilizations of the 1920s. The theoretical arguments are confronted with the empirical evidence in two stages. The first stage involves casual observation of the relation between the degree of nominal instability, the issuance of external loans, and the external enforcement of stabilization. The second stage is more rigorous and attempts to test the credibility-financial problems hypothesis using discriminant and probit analyses.

1. Nominal instability, loans, and external enforcement

During the 1920s, most of the European countries undertook the task of returning to the pre-war gold standard and stabilizing their currencies. World War I had left these countries with relatively high rates of inflation and currency depreciation, which in turn were ignited by rapid monetary expansion to finance large continuing fiscal deficits. The gold standard was attractive because it had eliminated currency instability from the international financial system before World War I; therefore almost every country tried to establish convertibility of its currency into gold. The third column of Table 7 gives an idea of the magnitude of nominal instability suffered by various European countries, as proxied by the level at which these currencies were stabilized with respect to their pre-war gold value. There were only six countries whose currencies returned to the pre-World War I gold value, and these were mostly nations that had remained neutral during the war. Even these countries were able to achieve this objective only by implementing policies to deflate prices enough to compensate for the decline in the purchasing power of money that had occurred during the war years. At the other extreme of nominal instability, five countries suffered hyperinflations and had to implement draconian reforms to stabilize their currencies. 23/ Between these two extremes, the table shows that the currencies of fifteen countries were stabilized at levels ranging from 1/4 to 1/125 of their pre-war gold values.

Table 7.

European Stabilizations of the 1920s

article image
Source: Elaborated from League of Nations (1946).

The second year is the de jure stabilization when different from the de facto.

Second stabilization (first attempt was unsuccessful).

Stabilized in 1924 but abandoned in 1925.

Stable currency during 1925-1928.

Table 7 also indicates whether there was an external enforcement institution during these episodes. The six countries with programs endorsed by the League, together with Germany, are shown as having external enforcement. The German case is interesting because it is another instance during this period where stabilization was supported by external enforcement. The German hyperinflation was stabilized in 1923 without any initial external support. However, the final consolidation of a stable currency was attained with the assistance of a foreign loan floated under the Dawes Plan. This Plan was formulated by a committee of experts appointed by the Reparation Commission to study German reconstruction and reparation problems. The monetary restraint envisaged by the Dawes Plan was enforced through two different channels. The first was foreign intervention in the German central bank (Reichsbank): half of the fourteen members of the General Board would be foreigners who would monitor the Reichsbank’s note issue. The second was an agent for reparation payments appointed by the Reparation Commission to monitor the German’s currency policy. 24/

Although the Polish case is similar to the German experience, Poland is not shown as having external enforcement in Table 7. Like Germany, Poland relied entirely on monetary and fiscal reforms to stabilize its currency in 1924 after a terrible hyperinflation, with no external support at the beginning. However, inflationary pressures recurred and a second stabilization took place in 1927, this time assisted by foreign credits. The difference from the German case is that foreign interests were not directly represented on the board of the central bank. The external presence in Poland was reduced to an American financial adviser sent by the Kemmerer mission in 1926. 25/ This external involvement was so minimal that Poland is not classified as having external enforcement in Table 7.

Table 7 also shows external involvement through the provision of external credits. It indicates that only in five out of 25 countries was the stabilization carried out with specific currency stabilization loans. This is reproduced in the seventh column and refers to preliminary or de facto stabilization—that is, when the monetary authority is defending the parity. Temporary credits were extended to seven countries, though these were not necessarily used for the specific purpose of stabilizing their currencies. In contrast, thirteen countries did not have recourse to foreign loans to stabilize their currencies. With respect to the final legislation of the parity, or de jure stabilization, nine countries had temporary credits but made hardly any use of them. Only five consolidated their stabilization without any access to specific credit arrangements, while eleven stabilized their currencies de jure with the aid of specific loans.

Although the majority of European countries achieved preliminary stabilization without specific financial assistance from abroad, the following observations are relevant. First, the eleven countries that received loans for the final stabilization were among those with the highest rates of currency depreciation and included four economies that were suffering from hyperinflation. In fact, applying the straightforward non-parametric test of medians it turns out that the median currency depreciation in countries with specific external loans (tenth column of Table 7) was significantly higher than the median currency depreciation in countries without specific loans. 26/ This suggests that nations with higher nominal instability relied more heavily on external credits to stabilize their currencies than nations with lower nominal instability. 27/

Second, there was external monitoring of stabilization programs, as evidenced in Table 7, in six of the countries: Austria, Bulgaria, Danzig, Germany, Greece and Hungary. 28/ As before, these six countries were among those with the highest rates of nominal instability as measured by the exchange rate depreciation. Again applying the test of the medians it turns out that the median currency depreciation for the countries with external enforcement was higher than that for countries with no external enforcement. 29/ This observation supports the assertion that countries with higher nominal instability sought external enforcement to stabilize their currencies more frequently than countries with lower instability.

Third, we must take into account the possible effect that anticipated future external assistance may have on the initial or, de facto stabilization of the currency. For instance, Llach (1987) imputes to the announcement of the formation of the commission of experts some causal influence on the success of the German hyperstabilization. 30/ Similarly, in the Austrian case the anticipation of a League of Nations’ reconstruction scheme appears to have acted to stabilize the Austrian crown. 31/ Thus even if thirteen countries did not actually have recourse to foreign loans to stabilize their currencies de facto, the fact that future credits of some kind were anticipated in eight of them may have helped to effect the stabilization.

Finally, it is interesting to note that four of the five countries that experienced hyperinflation during the 1920s (Austria, Germany, Hungary and Poland) had access to foreign loans and to external enforcement of the conditionality attached to these loans. 32/ The sole exception was the stabilization in the Soviet Union in 1922. This latter case suggests, consistent with the view in the League’s own 1946 report, that external loans and foreign enforcement are not strictly necessary to implement a hyper-stabilization. 33/ Later episodes in Hungary (1945) and China (1949) furnish additional examples of socialist stabilization without external intervention.

2. Discriminant and probit analysis

The basic hypothesis that emerges from the theoretical discussion of Section II is that programs to stabilize economies that face more severe credibility problems and/or more pressing financial needs are more likely to require external enforcement. Moreover, the analysis in Section III of the programs endorsed by the League of Nations suggests that those programs were characterized by binding commitments that enhanced the credibility of the stabilization attempts. Hence, it seems natural to test whether countries with greater credibility problems and financial needs did indeed rely more heavily on external agents to stabilize their economies. The previous subsection indicated that countries with higher nominal instability did have recourse to such external involvement.

Unfortunately, the observed counterpart of a “credibility problem” is hard to find, and the investigator is forced to rely on proxies that may indicate the existence of such problems. The difficulty in finding suitable variables to measure the degree of credibility is exacerbated by the lack of reliable data for the 1920s. Moreover, as discussed above, credibility problems and financial needs are likely to appear together, increasing the difficulty of finding a proxy that isolates a credibility problem. We will, therefore, handle them as a single hypothesis. Given these limitations, three variables have been chosen to serve as proxies for the “credibility” of the economic policymaking authority. The first variable, BACKt, is the ratio of the sum of gold and net foreign assets held by the central bank to the amount of note circulation. The idea is that lower values of BACKt should mean that the government is less able to defend a given exchange rate objective. This measure was emphasized by the League itself as a good indicator of sound financial practices. The second variable, FISCt, is the ratio of the fiscal balance to the level of government expenditures. The value of FISCt is intended to capture the degree of sustainability of the fiscal policy: lower values should mean less sustainable (and credible) policies. This variable can also be interpreted as a measure of the severity of the financial needs that a government faces. Finally, the third value considered, DEPRt, is the percentage rate of depreciation in the end-of-period value of the domestic currency (New York market); the logic is that a higher rate of depreciation suggests that markets displayed less confidence in the policies followed by that country. All variables are measured for the year in which the stabilization took place (see Table 7 for the specific years).

With respect to the endogenous classification criterion of the episodes, a binary variable is considered. This is a dummy variable that takes a value of unity if the country stabilized during—or in clear anticipation of—participation by an external enforcing agency in the stabilization program, and zero otherwise. The countries for which this dummy takes a value of unity are Austria, Bulgaria, Germany, Greece and Hungary. Descriptive statistics for the three variables and data sources are provided in the Appendix Table 1, where the countries are classified according to whether or not there was external enforcement. Notice that countries that stabilized with external enforcement exhibit a distribution of credibility indices that lies clearly to the right of the distributions of countries without external enforcement. Indeed, countries with external enforcement not only performed worse on average, according to the distributions in that table, but they also suffered from a greater degree of variability in the indices that we have defined.

A first step in testing the credibility and financial needs hypothesis of external involvement is implemented through a discriminant analysis. The objective is to determine whether countries can be classified into those that did and those that did not rely on external assistance on the basis of the three variables described. The classification is carried out using the generalized squared distances between the two groups of countries, assuming equal prior probabilities. The results presented in Table 8 are very suggestive and indicate that most of the countries can indeed be identified through the use of these three variables. In particular, 15 out of the 16 countries that did not rely on external enforcement are correctly classified by the discriminant analysis, while 3 of the 5 that were supported by international institutions are also identified. The only countries that are misclassified are Bulgaria, Greece and Italy. 34/

Table 8.

Discriminant Analysis of the European Stabilizations of the 1920s

article image
Classifying variables: BACKt, FISCt and DEPRt

The second step in testing the credibility and financial needs hypothesis is to carry out a probit analysis to verify the robustness of the results obtained from the discriminant analysis. This approach also provides a way to estimate the individual predictive powers of each of the three variables chosen in the forecast of the event defined by the dummy variable. The results of this probit estimation (Table 9) are very interesting. All of the variables have the expected signs, except for BACKt in equation (5); both FISCt and DEPRt are always statistically significant when they are not included in the same regression. The likelihood ratio tests whether the coefficients of variables other than the intercept are all equal to zero; this null hypothesis is rejected for all specifications except equation (1). These results are further corroborated by McFadden’s R2. 35/

Table 9.

Probit Analysis of the European Stabilizations of the 1920s

(Maximum Likelihood Estimates)

article image

-2 times log of likelihood ratio.

McFadden pseudo R2.

Percentage of accurate predictions.

What is even more interesting is the predictive power of the explanatory variables: over 78 percent of the observations are classified correctly in the worst specification. In this respect, it is worth noting that the fiscal balance has 86 percent of the observations as correct predictions. The countries that are not correctly classified by FISCt are Bulgaria, Greece and Italy. When the fiscal balance variable is supplemented by the rate of depreciation, the percentage of correct predictions increases to over 90 percent, and Italy is no longer misclassified. However, BACKt is an indicator that performs poorly. 36/

In sum, the empirical tests described here support the hypothesis that the presence of credibility problems and financial needs, when proxied by fiscal inconsistency and instability of the exchange rate, did indeed influence the decisions of countries to rely on external agents to help enforce stabilization programs undertaken during the 1920s. Of course, the interpretation of the results presented here does not make a strong distinction between the two different arguments. However, as indicated in Section II, the need for financing is often aggravated by credibility problems, so even though the two explanations are conceptually different, in practice it is difficult to disentangle them.

V. Conclusions and Lessons from the 1920s

There are two main reasons why a government may rely on external assistance to stabilize its economy. The first is that an external agent can restore credibility to the government’s macroeconomic stabilization efforts. This has been referred to as the ‘commitment technology’ argument. The second explanation is that external assistance, in the form of foreign loans and credits, can help to finance both the imbalances of an economy in disequilibrium and the stabilization task confronting the economy. This paper has argued that both these incentives to rely on external intervention are likely to be more pressing in the case of countries that are experiencing high nominal instability. The broad conclusion that emerges from the analysis of the reconstruction schemes carried out by the League of Nations in the 1920s is that for a stabilization effort to succeed it must be effected through a fundamental change in the fiscal and monetary regime. To achieve such a regime change, the League of Nations supported a complete and comprehensive series of reforms in each country where it was involved. Of vital importance in these countries was the goal of achieving sound and sustainable public finances, the establishment of independent central banks, and the return to convertible currencies. Nevertheless, the attempt to carry out all of these reforms might have been in vain if the League had not had the means to control and monitor enforcement of the programs, i.e. if the League had lacked a suitable ‘commitment technology’. In fact the control machinery, together with the League’s concern to protect its own reputation, appears to have helped to ensure that the League’s reconstruction schemes constituted credible changes of regime. In sum, the stabilization programs of the League achieved an impressive record of success, at least in the period prior to the Great Depression.

This study of the historical experience of the currency stabilizations undertaken in the 1920s yields suggestive results in several areas. The paper finds preliminary evidence that, consistent with the analytical argument of Section II, countries that suffered from higher nominal instability displayed a higher reliance on both external loans and external enforcement. The empirical analysis of the role of external assistance in the stabilization programs was undertaken in this paper by comparing indicators of credibility and financial need for countries that resorted to external intervention with those countries that did not. Our results support the hypothesis that the presence of credibility problems and financial need, especially when proxied by fiscal unsustainability and exchange rate instability, influenced the decision of European countries in the 1920s to rely on an external agent to help enforce their currency stabilizations.

The currency stabilizations of the 1920s, and in particular the experience of programs implemented with the support and assistance of the League of Nations, provide several lessons for current problems. First, these episodes suggest the appropriate procedure for implementing a monetary stabilization: programs should be comprehensive and should place special emphasis on the right institutional setting in order to provide lasting stability. Second, these episodes give some insights into the sorts of circumstances where external involvement becomes more desirable. Countries with economies in profound disarray and which suffer credibility and financing problems are more likely to require external enforcement of an adjustment program. Third, the experiences also illustrate the degree of sacrifice that is needed in terms of a country’s sovereignty and discretion in order to enable the external enforcer of a stabilization to achieve the objectives of the program. The analysis suggests that in cases where policymakers are willing to forego the flexibility of a discretionary regime in favor of a commitment regime, they are more likely to succeed in implementing a credible, and hence successful, stabilization program.


Table 1.

Summary Statistics of the Credibility Indicators

article image
Note: See text for definition of the variables.Source: International Statistical Yearbook, League of Nations; Memorandum on Public Finance; Memorandum on Currency and Central Banks; Estonian Yearbook; League of Nations (1926ab, 1927), Young (1925), Mazower (1991) Department of Overseas Trade (1929), Lithuanian Ministry of Finance (1924) and Mitchell (1980).


  • Backus, D. and Driffil, J. (1985), “Inflation and Reputation,” American Economic Review 75(3), 53038.

  • Barro, R. (1986), “Reputation in a Model of Monetary Policy with Incomplete Information,” Journal of Monetary Economics 17(1), 320.

    • Search Google Scholar
    • Export Citation
  • Barro, R. and Gordon, D. (1983a), “A Positive Theory of Monetary Policy in a Natural Rate Model,” Journal of Political Economy 91(4) 589609.

    • Search Google Scholar
    • Export Citation
  • Barro, R. and Gordon, D. (1983b), “Rules, Discretion and Reputation in a Model of Monetary Policy,” Journal of Monetary Economics 12(1) 10122.

    • Search Google Scholar
    • Export Citation
  • Blackburn, K. and Christensen, M. (1989), “Monetary Policy and Policy Credibility: Theories and Evidence,” Journal of Economic Literature 27(1), 145.

    • Search Google Scholar
    • Export Citation
  • Bresciani-Turroni, C. (1937), The Economics of Inflation, Allen & Unwin, London.

  • Cagan, P. (1956), “The Monetary Dynamics of Hyperinflation,” in Friedman, M. (ed.) Studies in the Quantity Theory of Money, U. of Chicago Press, Chicago.

    • Search Google Scholar
    • Export Citation
  • Calvo, G. (1978), “On the Time Consistency of Optimal Policy in a Monetary Economy,” Econometrica 46(6), 141128.

  • De Bordes, J. van Walré (1924), The Austrian Crown. Its Depreciation and Stabilization, P. S. King, London. Reprinted by Garland Publishing, New York, 1983.

    • Search Google Scholar
    • Export Citation
  • Department of Overseas Trade (1929), Economic Conditions in Poland, H. M. Stationery Office, London.

  • Department of Overseas Trade (1930), Economic Conditions in Portugal, H. M. Stationery Office, London.

  • Dominguez, K. (1992), “Role of International Organizations in the Bretton Woods System,” NBER Working Paper 3951.

  • Dornbusch, R. (1992), “Monetary Problems of Post-Communism: Lessons from the End of the Austro-Hungarian Empire,” mimeo, MIT.

  • Dornbusch, R., and Fischer, S. (1986), “Stopping Hyperinflations Past and Present,” Weltwirtschaftliches Archiv 122, 147.

  • Dornbusch, R., Sturzenegger, F. and Wolf, H. (1990), “Extreme Inflation: Dynamics and Stabilization,” Brookings Papers on Economic Activity 2, 164.

    • Search Google Scholar
    • Export Citation
  • Edwards, S. (1989), “The International Monetary Fund and the Developing Countries: A Critical Evaluation,” Carnegie-Rochester Conference Series on Public Policy 31, 768.

    • Search Google Scholar
    • Export Citation
  • Eichengreen, B. (1992), Golden Fetters: The Gold Standard and the Great Depression 1919-1939, Oxford University Press, New York.

  • Fischer, S. (1980), “Dynamic Inconsistency, Cooperation and the Benevolent Dissembling Government,” Journal of Economic Dynamics and Control 2(1), 93107.

    • Search Google Scholar
    • Export Citation
  • Garber, P. and Spencer, M. (1992), “The Dissolution of the Austro-Hungarian Empire: Lessons for Currency Reform,” IMF working Paper WP/92/66.

    • Search Google Scholar
    • Export Citation
  • Harberger, A. and Edwards, S. (1980), “International Evidence on the Sources of Inflation,” University of Chicago, mimeo, Chicago.

  • Heymann, D. (1986), “Las Grandes Inflaciones: Características y Estabilización,” in Tres Ensayos sobre Inflación y Estabilización, CEPAL, Buenos Aires.

    • Search Google Scholar
    • Export Citation
  • Kydland, F. and Prescott, E. (1977), “Rules rather than Discretion: The Inconsistency of Optimal Plans”, Journal of Political Economy 85(3), 47391.

    • Search Google Scholar
    • Export Citation
  • League of Nations (1926a), The Financial Reconstruction of Austria, Geneva.

  • League of Nations (1926b), The Financial Reconstruction of Hungary, Geneva.

  • League of Nations (1927), Banking and Currency Reform in Estonia, Geneva.

  • League of Nations (1930), Principles and Methods of Financial Reconstruction Work undertaken under the Auspices of the League of Nations, Secretariat of the League of Nations, Geneva.

    • Search Google Scholar
    • Export Citation
  • League of Nations (1945), The League of Nations Reconstruction Schemes in the Inter-war Period, Publications Department of the League of Nations, Geneva.

    • Search Google Scholar
    • Export Citation
  • League of Nations (1946), The Course and Control of Inflation, Publications Department of the League of Nations, Geneva.

  • League of Nations International Statistical Yearbook, Geneva, various issues.

  • League of Nations Memorandum on Public Finance, Geneva, various issues.

  • League of Nations Memorandum on Currency and Central Banks, Geneva, various issues.

  • Leijonhufvud, A. (1983), “Keynesianism, Monetarism and Rational Expectations: Some Reflections and Conjectures” in Frydman, R. and Phelps, E. (eds.) Individual Forecasting and Aggregate Outcomes, Cambridge U. Press, New York.

    • Search Google Scholar
    • Export Citation
  • Leijonhufvud, A. (1984a), “Constitutional Constraints on the Monetary Powers of Government” in Mc Kenzie, R. (ed.) Constitutional Economics, Lexington Books, Lexington.

    • Search Google Scholar
    • Export Citation
  • Leijonhufvud, A. (1984b), “Inflation and Economic Performance” in Siegel, B. (ed.) Honey in Crisis, Ballinger, Cambridge.

  • Lithuania, Ministry of Finance (1924), Economic and financial situation of Lithuania in 1924, State Printing Office, Kaunas.

  • Llach, J. (1987), “La Naturaleza Institucional e Internacional de las Hiperestabilizaciones,” Desarrollo Económico 26(104), 52759.

    • Search Google Scholar
    • Export Citation
  • Meyer, R. (1970), Banker’s Diplomacy: Monetary Stabilization in the Twenties Columbia University Press, New York.

  • Mitchell, B. (1980), European Historical Statistics, 1750-1975, Facts on File, New York.

  • Osband, K. and Villanueva, D. (1992), “Independent Currency Authorities: An Analytic Primer”, IMF Staff Papers 39(4).

  • Pasvolsky, L. (1928), Economic Nationalism of the Danubian States, Macmillan, New York.

  • Persson, T. and Tabellini, G. (1990), Macroeconomic Policy, Credibility and Politics, Harwood Academic, New York.

  • Pullerits, A., The Estonian Yearbook, Tallinn, various issues. Rogoff, K. (1987), “Reputational Constraints on Monetary Policy,” Carnegie-Rochester Conference Series on Public Policy 26, 14182.

    • Search Google Scholar
    • Export Citation
  • Sachs, J. (1989), “Conditionality, Debt Relief, and the Developing Country Debt Crisis,” in Sachs, J. (ed.) Developing Country Debt and Economic Performance, 255-95, NBER, University of Chicago Press, Chicago.

    • Search Google Scholar
    • Export Citation
  • Sargent, T. (1986), Rational Expectations and Inflation, Harper-Row, New York.

  • Schacht, H. (1925), “The New German Currency” in Young (1925).

  • Schacht, H. (1927), The Stabilization of the Mark, Adelphi, New York.

  • Vaubel, R. (1986), “A Public Choice Approach to International Organizations,” Public Choice 51, 3957.

  • Young, J. (ed.) (1925), European Currency and Finance, Commission of Gold and Silver Inquiry, United States Senate, Washington, D.C.


This paper draws on two chapters of the author’s Ph.D. dissertation. Comments by Malcolm Knight, Delano Villanueva, Michael Spencer, Sebastian Edwards, Arnold Harberger, Axel Leijonhufvud, Jeffry Frieden, Kenneth Sokoloff, Jean-Leaurant Rosenthal and Daniel Heymann, among many others, are gratefully acknowledged.


A recent study of the economic history of the 1920s is Eichengreen (1992).


There has been a recent interest on the dissolution of the Austro-Hungarian empire and the similarities with the former Soviet Union. See Garber and Spencer (1992) and Dornbusch (1992).


Dornbusch, Sturzenegger and Wolf (1990) describe these issues. In this section we use the terms stabilization and adjustment interchangeably.


On the outcomes under different regimes see Barro and Gordon (1983b) and Fischer (1980).


A currency board is an example of such an institutional monetary arrangement. See Osband and Villanueva (1992).


See Barro (1986) and Backus and Driffil (1985). Blackburn and Christensen (1989) survey the literature. See also Persson and Tabellini (1990) for a comprehensive review of the literature on time inconsistency and modern political economy.


See Leijonhufvud (1983, 1984b) for a definition and discussion of a monetary regime.


See also the recent work by Dominguez (1992).


The credibility and financing arguments to rationalize external intervention are not the only possible explanations. Countries can also rely on an external agent to get technical assistance, whether fiscal, monetary or in other areas. A different explanation from the public choice literature suggests that governments use external agents as scapegoats and blame them for undertaking painful adjustment programs and other “dirty work”. See, for example, Vaubel (1986).


Garber and Spencer (1992) describe the League’s involvement in Austria and Hungary. In this paper, we emphasize the enforcement of each of the six reconstruction schemes as a necessary condition to increase the likelihood of the program’s success in effecting a change of regime.


The financing for the refugee settlement schemes was aimed at meeting specific expenses to settle them.


For example, in the two Greek schemes the International Financial Commission (IFC) that had been used by foreign creditors since 1898 to control some state revenues assigned as security to foreign loans, continued to function during the period of the League’s involvement. In the case of Danzig, a High Commissioner had already been appointed by the League to reside there because, according to the Treaty of Versailles, Danzig was made a free city under the protection of the League.


League of Nations (1930, 1945).


Portugal carried out a program of its own which was similar to the one recommended by the Financial Committee. See League of Nations (1930) and Department of Overseas Trade (1930).


Meyer (1970) discusses in detail the Polish and Romanian cases.


Some studies of the hyperinflation experiences are the seminal paper by Cagan (1956) or the more recent ones by Sargent (1986), Dornbusch and Fischer (1986) and Heymann (1986).


See Schacht (1925, 1927) and Heymann (1986).


See Meyer (1970). Dornbusch and Fischer (1986) also mention the presence of a British adviser to Polish authorities before the Kemmerer mission.


The level of significance is 0.002. The test was performed by ordering the countries according to their percentage currency depreciation during the period of non-convertibility (third column of Table 7) and sorting the countries that were above and below the median country. The test then analyzed the relative position of countries with and without external loans with respect to the median country to determine if their currency depreciation was the same (null hypothesis). Those countries that issued a national currency for the first time (Danzig and Lithuania) were excluded.


Further support for this statement comes from the fact that there was recourse to external loans in all the high-inflation episodes studied by Dornbusch and Fischer (1986).


The Eesti Mark was stabilized de facto in 1924, well in advance of the 1927 loan obtained by Estonia through the League. It is thus plausible that the external involvement was not anticipated at the moment of the stabilization. Danzing was dropped from the analysis due to lack of data for 1923.


With a level of significance of 0.024.


The same is admitted by the Reparation Commission (as quoted in Bresciani-Turroni 1937, p. 339).


The degree of external control in the Polish case seems to be very mild.


The Soviet stabilization was of a different type. Llach (1987) calls it “stabilization by coercion” as opposed to other cases that were “by consent”.


Bulgaria and Greece are probably misclassified because they did not exhibit exchange rate depreciations as large as the hyperinflation countries. In the case of the Italian misclassification, it is likely to be due to the substantial fiscal deficit.


Similar results are obtained by running logit regressions instead of the probit specification, as well as by lagging the variables on period instead of using the current values.


This performance seems to be compatible with the evidence found by Harberger and Edwards (1980) in their study of modern devaluation episodes. They found that countries that abandoned a fixed exchange rate reacted to a decrease in their equivalent measure of BACK only after it had reached a critical level. A similar scenario in which a critical value of BACK triggers external enforcement of a stabilization program may be possible in our sample.

Stabilization Programs and External Enforcement: Experience From the 1920's
Author: Julio A. Santaella