The Speed of Adjustment and the Sequencing of Economic Reforms: Issues and Guidelines for Policymakers

Contributor Notes

Authors’ E-Mail Addresses:,,

This paper reviews the issues involved in determining the appropriate speed of adjustment and the sequencing of economic reforms, focusing on considerations relevant to policymakers. It points out that the debate between the protagonists of a high-speed approach and those favoring a gradualist approach is based primarily on the weights given to adjustment costs, policy credibility, reform feasibility, and risk assessment. It underscores the importance of appropriate sequencing and the impact of sequencing on the speed of adjustment and reforms. The paper concludes by highlighting factors that policymakers should consider when selecting their approach toward speed and sequencing.


This paper reviews the issues involved in determining the appropriate speed of adjustment and the sequencing of economic reforms, focusing on considerations relevant to policymakers. It points out that the debate between the protagonists of a high-speed approach and those favoring a gradualist approach is based primarily on the weights given to adjustment costs, policy credibility, reform feasibility, and risk assessment. It underscores the importance of appropriate sequencing and the impact of sequencing on the speed of adjustment and reforms. The paper concludes by highlighting factors that policymakers should consider when selecting their approach toward speed and sequencing.

I. Introduction

The ongoing adjustment and reform efforts of developing countries, the transformation in Central and Eastern Europe and other transition countries of the former Soviet Union, and the recent crisis in East Asia have brought to the fore the importance of the speed and appropriate sequencing of adjustment efforts.2 For policymakers, these issues are critical to their decision making. In fact, over the past several decades, observers of internationally supported adjustment programs have tended to disagree more on the pace and sequence, than on the content, of the reform packages (see, for example, Collier and Gunning, 1999).

Numerous studies have analyzed the factors influencing the speed and sequencing of adjustment (see Box 1 for definitions). Conceptually, the problem is to find the optimal adjustment path that will maximize the intertemporal social welfare function of a country, subject to financial and structural constraints (Nsouli, 1996). However, estimating the social function and quantifying the financial and structural constraints are not simple tasks. An attempt to do so, in purely hypothetical simulations, was undertaken by Feltenstein and Nsouli (2001). In general, the extensive literature on speed and sequencing has been inconclusive and often conflicting in the policy advice it provides. An overview, which takes into account all three aspects of this question—speed considerations, sequencing requirements within each sector, and sequencing requirements relative to other sectors—is largely missing from the literature. This paper seeks to fill this gap and to provide a practical guide for policymakers on speed and sequencing issues.

The paper is divided into four sections. Section II reviews the factors that are at the center of the debate on the speed of adjustment. Section III focuses on the three key questions: What are the sequencing requirements within a sector? What are the major sequencing requirements relative to other sectors? What are the speed implications of sequencing within and across sectors? The last section provides practical guidelines for reform and discusses factors that need to be considered in determining the speed of adjustment and sequencing of reforms.

Definitions of Speed and Sequencing

The debate on the speed of adjustment and sequencing of reforms is often carried out in a vacuum. Articles typically do not fully define the concepts; hence, their analyses are not always comparable. The definitions used in this paper are as follows.


The speed of adjustment can be defined as the time elapsed between the move from an initial set of macroeconomic variables to a targeted set of such variables. If in period zero, for example, the rate of inflation is at X0 percent, and it takes t time periods to reach a targeted rate of inflation set at Xt, then the speed of adjustment refers to the number of periods to go from X0 to Xt. The relations, however, could be more complex, involving preset targets for growth, inflation, and the external sector accounts, as well as intermediate targets, such as the budget balance and the rate of credit expansion.

The speed of adjustment also refers to the time elapsed in moving from one organizational economic structure to another. For example, in a broad sense, the speed of adjustment refers to the time involved in moving from a centrally planned to a market-oriented economy. In a narrower sense, it refers to the time involved in reducing price controls, changing the tariff structure, privatizing public enterprises, introducing financial sector reforms, and establishing the relevant institutions.

Thus, the speed of adjustment refers to the total time required to move from one set of macroeconomic variables to another and to introduce economic reforms and make them operational.


The sequencing of reforms refers to the order in which either macroeconomic policy actions or specific reforms are introduced. Sequencing involves the order in which reforms are undertaken across sectors (for example, whether fiscal adjustment or stabilization should be a prerequisite for introducing current account liberalization or decontrolling prices) and the order in which reforms are undertaken within sectors (for example, whether in the case of capital account liberalization, foreign direct investment or short-term capital flows should be liberalized first).

The sequencing across sectors and within sectors, to the extent that it requires time, will necessarily impact on the speed of adjustment.

II. The Speed of Adjustment

A fairly clear-cut argument divides the proponents of a high speed of adjustment—what is often referred to as the shock, big bang, or cold-turkey approach—and those supporting a gradual approach. The debate centers on four major issues: the costs of adjustment, the credibility of the reform program, the feasibility of the approach, and the risks associated with the strategy. Table 1 summarizes the opposing views, focusing on broad considerations. This section reviews the arguments of the two camps.

Table 1.

The Speed of Adjustment: Contradictory Views?1

article image

To highlight different views, this table focuses on the main arguments rather than on the underlying details. The citations often refer to earlier papers, as subsequent papers of the same author usually contained similar views. The author to whom the argument is attributed is not necessarily a proponent of one or the other approach.

A. Adjustment Costs

Under circumstances where product and factor prices adjust immediately and resources can be reallocated without cost, the optimal policy is clear, namely the simultaneous removal of all distortions. In the real world, however, resources cannot be reallocated simultaneously without incurring costs among different sectors of the economy. Moreover, different markets adjust to policy changes and price signals at different speeds.

Proponents of both the shock and the gradual approach base their arguments on lower adjustment costs. Supporters of shock therapy argue that rapid reforms increase the incentives to reallocate resources, resulting in lower adjustment costs. In the presence of rational expectations and the absence of distortions, speedy reforms lead to a more socially desirable adjustment path for the economy (Mussa, 1984). When distortions are widespread, rapid adjustment can lead to income and wealth losses for those affected by the reformed industry. Those losses, however, would be acceptable if the marginal social costs of the losers were less than the marginal social benefits derived from the productivity gains of moving resources out of protected industries.

Proponents of gradualism note that labor and capital are sector specific and thus not readily transferable between sectors. Thus, rapid reforms can result in larger temporary contraction of economic activity than if a gradual approach had been followed. Earlier studies along these lines mostly focused on trade reforms. Little, Scitovsky, and Scott (1970), in their pioneering work on sequencing, argued that fast reforms generate higher short-term costs in terms of unemployment and income distribution effects, engendering stiffer political opposition than a gradual approach.3 In contrast, a gradual approach would produce less severe losses of rent (or gain) by owners of factors of production during a particular phase of reform (Mehlum, 2001). Part of the gains obtained by the initial reforms can be transferred to the losers of subsequent reforms (Lian and Wei, 1998). According to this view, a gradual, multistage reform is superior to a one step removal of all controls. Dewatripont and Roland (1992,1994) argue that gradualism may be preferable to the big-bang approach when the budgetary cost of reform (for example, the cost of maintaining the income level of workers laid off from inefficient state enterprises undergoing reform) is an important policy variable.

The above discussion shows that a major concern in the use of shock therapy is the potential higher rate of unemployment that can result during the transition to the new equilibrium (Ahluwalia, 1996; Agénor and Montiel, 1999). The shock approach is more likely to lead to reduced production and significantly higher unemployment if rigidities in labor and product markets are high. In general, adjustment costs are expected to be lower when labor markets are less rigid, capital and labor are less specific, and entrepreneurs are more flexible and adaptable. Supporters of the shock approach recognize the limitations of executing reforms quickly and concede that certain aspects of reform may take several years to implement.

B. Credibility

The credibility of government policy and the reform process is essential for the success of the reforms and the control of adjustment costs (Bhattacharya, 1997; and Cukierman and Tommasi, 1998). Credibility prevails when private expectations about future policies do not deviate from the authorities’ explicit or implicit announcements. Credibility is an important determinant of the magnitude of the adjustment costs. If reforms are credible, private agents will align their behavior with the announced policies. The more credible the reforms, the faster will the reallocation of resources be, all other things being equal. By contrast, if announced policies and reforms are not credible, agents will be reluctant to respond to the announcement and the adjustment process will take longer or may even fail. The importance of credibility increases in a highly distorted economy. In such an environment, intertemporal speculation can derail efforts to gradually undertake reforms when these reforms are perceived to be uncertain or reversible (Van Wijnbergen, 1991).

Credibility of a reform package can cut both ways. Broad-based reforms can enhance the credibility of reforms in developing countries if the public views them as a clear signal by the authorities to break with past traditions (Hiemenz and others, 1992; and Funke, 1993). Rapid and broad reforms can also have a greater probability of success if they help to overcome more easily the resistance of vested interest groups. Partial reforms, on the other hand, may not only raise credibility questions, but may also fail if the government lacks credibility initially.

By contrast, proponents of a gradual approach argue that gradualism can enhance credibility if the results achieved in the short term are sufficiently favorable (Rodrik, 1989). Successful initial reforms can enhance the authorities’ reputation and facilitate the implementation of subsequent reforms. In some cases, such as in current and capital account liberalization, several reforms may have to be tackled simultaneously without necessarily liberalizing the accounts completely in one go. In such an environment, a gradualist approach can also enhance credibility (Johnston, 1994).

C. Feasibility

The feasibility of the reform approach is also crucial to the debate. On the one hand, it can be argued that it is almost impossible to design a detailed sequence of reforms (Funke, 1993). Therefore, in periods of distress, the best approach may be to implement broad-based reforms as quickly as possible. On the other hand, proponents of a gradual approach stress that it simply takes time to implement reforms (Fischer and Gelb, 1991). Both the scope of the reforms and the country’s administrative capacity dictate a phased rather than a simultaneous approach to reforms.

Furthermore, the government’s ability to manage the reform process is vital (World Bank, 2002). It may take time to obtain the relevant information about the likely reform outcomes and associated probabilities (Erbas, 2002). Thus in spite of strong theoretical support for rapid and comprehensive reforms, many economists and policymakers believe that gradualism is the inevitable approach (Gelb and Fischer, 1991). The fundamental reason for gradualism (and thus sequencing) in reform emanates from the prerequisite nature of some policy actions.

Gradualism is also dictated by the “competition of instruments.” McKinnon (1973), in particular, argues that different reform elements require different instruments, some of which may conflict with each other. For example, a simultaneous implementation of trade and capital account liberalization may conflict with exchange rate policy. Trade reform may need to be supported by a depreciation of the real exchange rate whereas capital account liberalization, which may lead to capital inflows, can result in a real appreciation of the exchange rate. Feasibility considerations suggest that reforms must be introduced sequentially in light of the possible competition of instruments. Thus, groups of reforms involving complementary rather than competing policies must be identified and implemented sequentially.

D. Risks

The risks associated with each adjustment and reform strategy are also important. As already noted, the concern with shock therapy is that it can lead to lower production and higher unemployment. By contrast, the gradual approach can fail because of the longer time frame needed to bring about significant change. Limited information and general uncertainty about possible reform outcomes add to the challenges the authorities face.

If the shock approach does entail significant short-run increases in unemployment, political support for reforms may waver, eventually prompting the authorities to abandon their efforts (Agénor and Montiel, 1999). Moreover, when reforms are far reaching, the interrelations between the various elements may pose policy conflicts. For example, if a set of reforms does not succeed in one reform area, the public may become skeptical of the outcome of other reforms, reducing the effectiveness and likely sustainability of the reform effort.

On the other hand, partial reforms may fail to significantly reduce distortions and hence not deliver gains to the private sector quickly enough. Partial reforms that undermine the efficient reallocation of resources from controlled to liberated industries not only reduce output and welfare, but also fail to promote competitive markets (Lipton and Sachs, 1990; and Murphy and others 1992). The experience of some East European countries has shown the shortcomings of piecemeal approaches.

E. Other Factors

In addition to the speed of adjustment and reform, several other factors may have an important bearing on the final outcome:

  • Reform area: the sector matters, because the speed of adjustment differs between the real sector and the financial sector (Mussa, 1984).

  • Initial situation: initial conditions are decisive in determining the speed of reform (Bruno, 1992).

  • Cost of failure: the higher the cost of failure, the larger are the incentives for rapid adjustment (Dornbusch, 1991).

  • Affordability: countries that can no longer afford the negative implications of existing distortions, need more rapid approaches to reform (Johnston, 1994).

The overall speed of adjustment is affected by the pace of reform in different sectors. Whereas financial stabilization policies can generally be implemented relatively rapidly, the response of the production structure, investment, and ownership patterns to economic reforms tends to be much slower. Furthermore, reforms in such areas as privatization, tax, and trade require considerable preparatory work and time to implement.

The initial conditions are also critical in determining the appropriate speed of reform. It is argued that in a crisis situation, rapid adjustment is preferable. Difficult initial situations require fast and broad-based action. When the economy is deteriorating quickly, agents tend to revise their expectations more rapidly. Rigidities in the economy may weaken, allowing for more drastic action to bring about rapid change. In high-inflation countries undergoing stabilization, a big-bang approach is recommended to break the inflationary expectations (Bruno, 1992), or, in the case of financial crises, to restore confidence.

The higher the cost of failure, the greater is the incentive for rapid adjustment. In a two-period model, Dornbusch (1991) derives the following conclusions: the higher the cost of failure, the greater the adjustment effort and the probability of success; the higher the marginal cost of adjustment, the lower the effort and the probability of success; the higher the financing requirements, the higher the probability of program failure; and the higher the reserves or the ability to borrow, the lower the probability of program failure. Based on this analysis, failure of governments to reform in the first period implies that they must pursue the next phase of reforms with new adjustment costs and worse initial conditions. This two-period analysis highlights the incentive to front load the adjustment effort—policymakers have a strong incentive to perform well up front in order to forestall future higher adjustment costs and a lower probability of success.

The speed of reform may also be linked to affordability. Countries with low saving rates—such as in some Latin American countries—were inclined to speed up the implementation of reforms of the financial sector, whereas typically more gradualist approaches were followed in countries with high saving rates, such as Japan and Korea (Johnston, 1994). Efficient financial systems were less critical to the mobilization of savings in the latter group, which could afford to move gradually, while the former group had to move quickly to enhance intermediation.

III. Sequencing

The debate on sequencing initially surfaced in the late 1970s and early 1980s with the reform experience of Latin American countries. The predominance of reform failure generated an intense debate among policymakers and academics to determine whether the failures were due to the policy prescription or the sequencing of reforms. This debate was reinvigorated in view of the experience with the reform efforts of the transition economies.4 Many economists have argued for the appropriate sequencing of reforms without necessarily indulging in the big-bang versus gradualist debate.

Table 2 shows the different views on the sequencing of reforms, distinguishing between proposals for developing countries and transition economies.5 This distinction is important because different initial conditions require different reform tasks. In developing countries with, say, repressed financial markets, the liberalization of domestic markets stands central. In the case of a planned economy that is moving toward a market economy, reforms include liberalization of price controls, privatization of public enterprises, and relaxation of financial and labor market controls.6 Obviously, the classification of the various proposals in such a broad framework is limited by overlapping reform categories. Also, more specific recommendations related to one or two reform areas cannot be represented appropriately. Notwithstanding these caveats, the table highlights areas of agreement as well as the main controversies surrounding sequencing issues.

Table 2.

Schematic Views of Selected Sequencing Proposals

article image
Source: Funke (1993).

Only broadly defined reforms are considered. If they do not match exactly with the analysis in the proposals, the closest category is chosen. Number 1 (5) stands for the reform that should be initiated first (last). The ranking refers mostly to the starting time of reforms so that different reform steps may overlap. More than one ranking indicates that no unequivocal grouping appears possible.

As Table 2 suggests, most researchers agree that fiscal and monetary stabilization and institutional reforms should occur early in the reform process. More controversial are the timing of the other reform areas, in particular, the timing of trade reform, financial sector reform, and capital account liberalization. One drawback of this broad schematic view is that it gives no guidance on specific reform priorities within one sector and it is not helpful in gauging how fast reforms in any one area should take place.

This section focuses on the major reform areas: trade liberalization, capital account liberalization, financial sector reform, and price and market reform.7 The analysis centers on three questions: What are the sequencing requirements within each sector? What are the major sequencing requirements vis-à-vis other sectors? What are the speed implications of sequencing within each sector and across the relevant sectors?

A. Trade Reform

Trade reform has been debated extensively but without full agreement on speed and sequencing. However, there is widespread agreement on the benefits of dismantling trade barriers. Numerous theoretical and empirical studies show that open economies promote more efficient uses of resources, higher income levels, and higher growth (Dollar, 1992; Sachs, Warner, 1995; Sharer, 1998).8

The literature strongly favors the removal of quantitative restrictions at the initial stages of reform (Michaely and others 1991),9 providing several arguments in support of this strategy: the arbitrary nature of many quantitative restrictions creates uncertainty for consumers and producers, prompting both to overstock. The distortions reduce efficiency in domestic production and dampen competition in the domestic market. Quantitative restrictions support the development and maintenance of monopolies, and quotas often result in rent-seeking and corruption in the allocation of rights to imports.

Conversion from quotas to tariffs thus has several advantages. It reduces rent-seeking, and the rents associated with imports are transferred from the importers to the government. This provides additional support to a government attempting to carry out a stabilization program in conjunction with trade reform. In addition, dismantling quantitative restrictions enhances transparency and predictability of the system and reduces hoarding practices. Tariffs are also more direct, thus easier to measure, facilitating the design of a trade reform package and the monitoring of its progress.

The two policy reforms have, however, different implications. When quantitative restrictions are binding, removing the restrictions on a good improves welfare. Welfare does not increase when quantitative restrictions are imposed on other goods. This is because imposing quantitative restrictions on other goods does not lead to greater imports of the good that already has quantitative restrictions (because of the binding quota). In contrast, imposing a tariff on a good j when there is already a tariff on a good i can improve welfare because a tariff on i when the two goods are net substitutes leads to greater imports of j than there would have been otherwise. Imposing a tariff on j would improve welfare if it results in a structure of imports that is closer to what would have been had there been no tariffs.

Gradual tariff reform becomes more complex, particularly when the aim is to improve welfare at each step of reform. In general, the following adjustments in tariffs improve welfare: a proportional reduction in all tariffs and a cut in the highest ad valorem tariff if the good under consideration is a net substitute for all other goods (Falvey and Kim, 1992).10

In the case of trade restrictions through tariffs and quotas, an equiproportional reduction in all distortions improves welfare. The initial measure could also take the form of converting quotas to tariffs equal to the tariff implicit in the quotas. This measure will not change the volume of imports, but it will have income distributional effects as quota rents are transferred from importers to the government (Falvey and Kim, 1992).

Following a conversion of quantitative restrictions, it would be desirable to attain uniformity and a lower level of tariff protection early in the reform process. This allows for a quick reduction in the cost of protection, including the costs associated with a widely varying pattern of protection. Uniformity in the tax structure is attained by first raising very low tariffs and then lowering tariffs progressively, starting with high tariffs, while leaving the lower rates until later on in the reform process (Papageorgiou and others, 1986; Wolf, 1986). In essence, the objective is to create a neutral trade regime that provides equal incentives for exports and domestic sales.

On sequencing of trade reforms relative to other sector reforms, there is strong support for initiating stabilization before embarking on trade reform (Rodrik, 1992; Sachs, 1987; and Wolf, 1986). As long as high inflation significantly distorts relative prices, trade reform will take place under the wrong market signals. Resources will thus still be directed inefficiently. High and variable inflation rates, for instance, confound relative price changes with movements in the price level generating serious distortions by sending inappropriate signals to the market. Furthermore, inflation associated with a real exchange rate misalignment is likely to put pressure on the external sector.

Trade reform often necessitates large nominal devaluations to restore equilibrium in the exchange market, which in turn can exacerbate inflation when fiscal and monetary policies have been insufficiently tightened. Reforms are more credible and easier to carry out if unemployment, inflation, and external deficits are reduced beforehand. Other arguments supporting initial stabilization focus on the paucity of instruments to implement stabilization and trade reform simultaneously.

The argument for stabilization before trade liberalization is strengthened by the historical evidence provided by the Asian experience. When trade reform is combined with stabilization, adjustment costs can increase significantly, making the cost of the combined package much higher than that of sequential implementation (Falvey and Kim, 1992). Moreover, economic imbalances may constrain trade reform, necessitating prior adjustment (Agénor and Montiel, 1999).

Empirically, most reform programs that began with a depreciation of the real exchange rate and prudent economic policies were more successful and the adjustment costs were much smaller than previously believed (Papageorgiou, and others 1990; Michaely, Papageorgiou, and Choksi, 1991). Moreover, sustained trade reform cannot be achieved without reducing the public sector deficit to a noninflationary level, as evidenced by the slowing or backsliding of liberalization in countries with poor economic performance (Thomas and Nash, 1991).11 An early announcement and implementation of a trade liberalization program could increase efficiency thus helping to build political support to sustain the reform effort.

On the order of reforms, the literature supports liberalizing the domestic factor markets, particularly the labor market, before liberalizing the commodity markets. Some argue that factor prices play a more important role in resource allocation than output prices, as some factor prices play a significant role in production decisions. The entire price system could become distorted if factor prices do not reflect their true opportunity costs.

There is fairly widespread support for carrying out trade liberalization before capital account liberalization. The commonly cited reasons include higher cost and welfare implications of reverse sequencing.12 If domestic factor markets and foreign trade are still heavily distorted when the capital account is liberalized, capital could flow into sectors heavily affected by distortions, further increasing inefficiencies in domestic production.

The speed of reform depends on several factors. First, policymakers must consider the possibility of a temporary output loss due to trade liberalization. If domestic industries are heavily protected, full trade liberalization can lead to large-scale economic disruption and recession. Second, the simultaneous implementation of other required policy changes can create significant adjustment costs, in particular when the economic outlook is heading toward recession. Third, producers’ interests may contribute to political pressure for gradual tariff reform, especially when a program of stabilization is needed concurrently with liberalization.

The speed implications associated with reform across sectors often make reference to the likely losses of revenue for many developing countries, as it ultimately involves deep cuts in import and export taxes. Therefore, the speed with which the government can restructure the tax system and broaden its base will be an important consideration in determining the optimal speed of trade reform.

The speed of trade reform is also affected by the willingness of authorities to adjust the exchange rate. Although trade reform makes imports readily available, the export response normally takes longer to materialize. Thus, correcting the real exchange rate misalignment in conjunction with a comprehensive trade reform becomes an imperative. However, a real depreciation may not be viewed as a practical option, because it means a fall in real wages, unless the depreciation is offset by a rise in living standards brought about by reduced protection. Chile, for instance, pursued comprehensive liberalization together with an overvalued managed exchange rate in the 1970s, the result of which was a rapid rise in imports and a subsequent collapse in its exchange rate (Dornbusch, 1992).

B. Capital Account Liberalization

The liberalization of the capital account of the balance of payments has attracted considerable attention over the past decade, particularly following recent financial crises in several countries.13 Some experts have attributed these crises, at least in part, to failure to recognize distortions in domestic markets when countries liberalized international capital flows. How then should the authorities proceed with capital account liberalization?

It is widely recognized that the free flow of capital can be beneficial to economic development in economies devoid of distortions. From a theoretical perspective, capital account liberalization can improve welfare, as long as financial markets are efficient and foreign funds are used to support the development process. Foreign capital can reduce the cost of capital and help make capital intermediation more efficient. However, as long as domestic financial markets are mired in distortions, foreign capital can be detrimental to welfare. Most empirical studies find either a positive or a nonsignificant effect from capital account liberalization on economic development. Important channels include positive effects on domestic investment, technology spillover, and domestic financial development (Bekaert, Harvey, and Lundblad, 2001; Edwards, 2001; IMF, 2001; and Quinn, 1997).

Liberalizing the capital account involves the lifting of restrictions on inflows and outflows of foreign direct investment flows and the use of long and short-term financial instruments. The liberalization of long-term capital inflows, in particular foreign direct investment, is often among the first steps taken to open the capital account. Such a step can bring considerable benefits, including transfer of technology and know-how, which may promote more efficient business practices. It may also speed the learning processes of foreign investors and lenders about future profit opportunities in the country. Liberalizing long-term capital flows may thus help smooth the reform processes and facilitate adjustments to post reform incentive structures. Empirical studies tend to confirm that foreign direct investment flows can promote overall investment (Lim, 2001; and World Bank, 2001). Because direct foreign investment flows are more stable than portfolio flows, thus less prone to sudden shifts in investor sentiment, they are less likely to contribute to financial crises.

There is considerable debate on the optimal time to liberalize portfolio flows. In particular, there is concern that the higher volatility of short-term capital flows may increase the country’s vulnerability to crises. Long-term portfolio flows may therefore be liberalized before short-term flows.14 In general, portfolio inflows in debt and equity instruments are more likely to produce positive effects, if domestic financial markets and the financial infrastructure are well developed. A sound institutional environment, including adequate accounting rules, auditing, disclosure practices, and efficient payment systems are key prerequisites for its success (Eichengreen, Mussa, and others, 1999).

In many cases, debt inflows are liberalized before equity inflows because the domestic debt markets tend to develop before domestic equity markets, mirroring their different economic roles. Recent studies indicate that equity market liberalization can also have independent positive effects on private investment and growth. Again, these effects are more likely to materialize if an appropriate institutional framework is in place (Fuchs-Schündeln and Funke 2001).

Restrictions on outflows may reflect existing macroeconomic imbalances or distortions in the financial sector. In fact, they may be in place in order to maintain existing financial imbalances. For example, restricting outflows can be used to hold interest rates artificially low. In the event of a liberalization that is accompanied by capital outflows, the authorities must be prepared to cope with higher domestic interest rates (Eichengreen, Mussa, and others 1999). Similarly, an outflow may result in a real depreciation of the exchange rate, prompting the authorities to address the underlying imbalances that are supported by that restriction.

The optimal sequence of capital account liberalization is complicated and varies widely depending on initial conditions. Researchers broadly agree that capital account liberalization should come after successful stabilization as the information content of prices tends to be higher in stable environments. High and volatile inflation distorts relative prices and reduces the value of market signals. Liberalization of capital flows in such unstable environments could increase the volatility of capital flows, with adverse economic effects.

Regarding capital account liberalization and trade reform, the initial view argued for trade reform to precede capital account liberalization. McKinnon (1973) focuses on the problem of competing instruments. As noted earlier, current account liberalization may require a real depreciation of the exchange rate to offset the initial negative impact on the balance of payments emanating from reduced protection. In contrast, capital account liberalization may produce a real exchange rate appreciation if liberalization leads to capital inflows. Because the responsiveness of foreign demand for local assets is expected to be faster than the responsiveness of trade flows to the opening of the current account, an unsustainable appreciation of the exchange rate may result. Because goods markets and financial asset markets clear at different speeds, the goods market, which takes much longer to clear, should be liberalized first.

To avoid unnecessary resource shifts, several authors (Edwards, 1990; and Khan and Zahler, 1987) favor opening the current account before the capital account. McKinnon (1982) argues that the experience of the Southern Cone countries shows that Chile’s superior performance was due to having kept the capital account closed while tariffs were reduced. He associates Argentina’s poor performance with the country’s adoption of the opposite sequencing.

In contrast to this view, a number of authors favor a simultaneous liberalization of the current and capital account. In this case, foreign funds that result from capital account liberalization can be used to reduce or offset the short-run costs of trade adjustment (Krueger, 1981 and 1984; and Little and others 1970). There is also an argument in support of the liberalization of the capital account even before the current account. Lal (1987) proposes to open the capital account along with an announcement of a phased trade liberalization. To the extent that the announcement is credible, long-run investment decisions are determined by relative world market prices instead of distorted domestic market prices.

Reform of domestic financial markets before liberalization of the capital account is generally considered to be critical to the success of reforms (Hanson, 1992; Johnston, 1998; Mathieson and Rojas-Suarez, 1992; and McKinnon, 1991). In a financially closed economy, the domestic banking system often suffers from extensive and severe regulations. Once the capital account is liberalized, excessive domestic regulation could weaken the competitiveness of domestic banks relative to international banks. Johnston, 1998 emphasized the importance of developing and strengthening domestic financial institutions, markets, and instruments before opening the capital account. As long as banking supervision is weak, access to foreign funds could increase the vulnerability of the economy. Mathieson and Rojas-Suarez (1992) also underscored the importance of strengthening the safety and soundness of the domestic banking system before liberalizing the capital account.

The appropriate speed of capital account liberalization is dependent on the development of domestic financial markets. Liberalization of the capital account occurred gradually in most industrialized countries (IMF, 2001), whereas in emerging markets the gradual and shock approaches have both been used. If the domestic financial sector is already fully liberalized and has an institutional framework in place to ensure smooth operations, then the capital account can be liberalized immediately. Most developing countries lack such an institutional framework. However, postponing capital account liberalization until domestic financial markets function properly would delay the benefits of access to foreign capital. In such cases, the liberalization of the capital account and the development of the domestic financial framework could go in tandem. However, even though in more advanced emerging markets the domestic framework is generally more developed and capital account liberalization can advance fairly quickly, it can still take an extended period of time to ensure that capital account liberalization does not outpace the corresponding domestic liberalization (Eichengreen, Mussa, and others 1998).

C. Financial Markets

It is well established that a well-developed financial sector helps promote investment and growth (Khan and Senhadji, 2000). The literature on capital account liberalization has already highlighted the crucial role of a well-functioning financial market.15 Components of financial sector reform encompass such areas as monetary, exchange, regulatory, and structural reforms. More specifically, components of financial sector reform include (Johnston and Sundararajan, 1999):

  • Reform of the interest rate regime and monetary operations (for example, liberalizing interest rates, introducing market-based monetary control procedures, and increasing central bank autonomy);

  • Development of the banking sector (for example, reforming selective credit regulations, restructuring and recapitalization weak financial institutions, and reforming prudential regulation);

  • Development of money and interbank markets (including reform of the clearing, payment, and settlements systems);

  • Fostering of competition in financial markets (promoting the institutional development and non-bank financial intermediaries);

  • Development of long-term capital markets (including domestic public debt management and government securities markets); and

  • Development of foreign exchange markets.

Sequencing strategies must be guided by the objective of improving financial sector efficiency while strengthening financial stability. There is fairly broad agreement that key monetary control reforms should be among the first reforms pursued in financial markets (Johnston, 1994). Countries with repressed financial systems typically face problems because of the ineffectiveness of direct credit and interest rate controls. The introduction of market-based instruments, such as open market operations, improves the effectiveness and efficiency of macroeconomic policies. At the same time, these initiatives not only make possible, but require a freeing of direct controls on credit and interest rates. To compensate for weak institutional capacity of financial markets, monetary policy reform could initially be accompanied by an active use of reserve requirements and refinance facilities.

In the financial sector, reforms typically begin with the banking sector, followed by money markets and capital markets. As long as credit allocation is targeted to favor priority sectors rather than respond to market signals, the banking sector will remain vulnerable and problem loans will accumulate. In such a case, cleaning up of firms’ and banks’ balance sheets must be given high priority. At the same time, the banks’ incentive system must be changed to stop old lending habits.16

The introduction of indirect monetary policy instruments can stimulate the development of money markets. But it also requires concurrent reforms to strengthen the structure of these markets, to beef up trading arrangements, and to improve clearing settlement procedures.

The development of capital markets usually occurs at a later stage, because of the far-reaching reforms required to establish a sound institutional framework. Adequate accounting rules, as well as auditing and disclosure practices, strengthen market discipline and increase investors’ confidence in domestic capital markets. At the same time, institution building and prudential regulation and supervision are critical to reduce the risk of future financial instability (Johnston, 1994).

An important prerequisite for domestic financial market liberalization in an inflationary environment is to regain control over the fiscal deficit (McKinnon, 1991). The financing of large fiscal deficits by the expansion of the stock of high-powered money contributes to inflationary pressures. The monetary authorities would have to limit the expansion of domestic liquidity by raising successively the reserve requirements of banks and/or engaging in open market operations. This would lead eventually to an unsustainable financial situation.

As noted in the previous section, financial sector reform is also closely linked to the timing of the liberalization of the capital account. As long as the banking system and prudential regulations are still weak, allowing banks to expand their resources through external borrowing can make the banking sector more vulnerable. Financial sector reform should thus precede or accompany capital account liberalization. Most of the countries that avoided a crisis after opening capital flows had a sound financial system in place (IMF, 2001a). Financial market reform must not be seen in isolation, however, but rather in the context of a broader macroeconomic stabilization and reform program.

The early literature on the speed of financial sector reform mostly favored a gradualist approach. However, protracted financial repression, by allowing inefficient mobilization and allocation of financial resources to continue, is associated with high costs. Speed considerations must be guided by finding a balance between the potential benefits and risks associated with financial liberalization.17 In this regard, several factors are generally considered critical to the speed of financial sector reform (see Johnston, 1994).

First, the initial conditions matter, influencing greatly the approach chosen. The different experiences in Asian and Latin American countries, for example, suggest that countries with high savings rates, such as those in Asia, may find it easier to pursue a gradual approach. In contrast, countries with low saving rates may be forced to move faster, because the urgency to encourage domestic savings is higher. Not only does the official economy matter, but the actual conditions matter as well. In highly regulated financial markets, the incentives to circumvent existing regulations by developing parallel or curb markets are high. In such environments, where the unofficial sector plays an important role, broad-based and rapid changes are not only important for reducing distortions but are also more likely to generate greater benefits as the official sector takes over from the unofficial sector.

Second, the impact of reforms tends to be faster in financial markets than in other reform areas. Therefore, financial reforms may be well suited to signal the authorities’ willingness to accept market discipline. Asia’s experience suggests that a comprehensive, well-sequenced approach is needed that coordinates the strengthening of prudential supervision, the restructuring of the banking sector, and the liberalization process. Partial liberalization, without supporting banking sector reform, increases the risk of a financial crisis (Johnston and Sundararajan, 1999).

Third, it is important to reach a critical mass of reforms in bank restructuring and prudential supervision and then to progressively improve on these with the development of financial markets and governance of commercial banks. These reforms need to be supplemented by institutional reforms to avoid the pitfalls of liberalization (Khatkhate, 1998). With reforms inevitably taking time, the challenge is to accelerate financial reforms while limiting the potential risk for macroeconomic losses and distortions in resource allocation of poorly sequenced reforms (Johnston and Sundararajan, 1999; Williamson and Mahar, 1998).

D. Price Liberalization, Market Reform, and Privatization

Price liberalization, market reform, and privatization are closely linked. When markets are not functioning correctly and private enterprises are inefficient, price liberalization alone will not create the proper incentives.

Price liberalization is essential for the efficient allocation of resources within and across sectors. Without a rational price system, profit and losses alone cannot signal what industries should expand and which ones should shrink. In both transition and developing economies, price liberalization led to a rapid increase in the availability of products.

In addition to price liberalization, private ownership of firms is also key to a properly functioning market economy. Because the simultaneous privatization of all state-owned enterprises is rarely feasible, the optimal sequence of privatization becomes important. Relevant selection criteria can include the choice of specific sectors and the size and profitability of enterprises. The Chinese experience suggests that private initiatives in specific sectors, such as agriculture and services, can produce positive supply effects quickly.18 The experience of Central and Eastern European countries shows that privatization of small enterprises (small-scale privatization) may be a sensible focus of initial reform for transition economies. Small-scale enterprises are more adaptable, have fewer internal control problems, and have lower capital requirements. Enterprises that need less restructuring may be expected to generate positive employment effects sooner. Privatization of large enterprises may be economically and politically more difficult because the transitional unemployment that can result will raise political opposition, making it difficult to impose hard budget constraints on large companies (Kornai, 1986).

Recent theoretical and cross-country empirical research shows that the sequence of privatization is influenced or even depends on the initial conditions and government objectives (Gupta, Ham, and Svejnar, 2000). Depending on the initial conditions, the authorities’ primary objectives may be to increase efficiency, maximize privatization revenues, or both. To enhance efficiency, privatizing inefficient firms first may prove more useful. The success of this strategy, however, will depend on the institutional framework. Efficiency gains occur only if the institutional framework is conducive to fast restructuring and improved enterprise performance. The method of privatization also matters for efficiency gains. The empirical literature indicates that efficiency gains tend to be larger for start-up firms and firms that after privatization are dominated by outsiders, possibly with foreign involvement. Efficiency gains are less likely if firms continue to be dominated by insiders (Havrylyshyn and McGettigan, 1999). If quick privatization revenues are an objective, it may be reasonable to privatize profitable firms first. However, selling public enterprises requires adequate property rights legislation, agreed on accounting standards, a liberalized market for factors and goods, and a reasonable level of financial market development.

Price liberalization and privatization need to be coordinated with other reforms. Financial sector reform, for example, is critical for improving the supply response to price reforms. Price deregulation can only lead to an efficient allocation of resources when bank lending and other investment and consumer financing decisions are also guided by market principles. Price decontrol can have substantial temporary effects on prices, but in general, it has no lasting adverse effect on inflation (Hernández-Cata, 1999). Thus, price decontrol does not hinder medium-term macroeconomic stabilization, and both are closely linked to successful privatization (Barnett, 2000). With the higher information value of prices in a stable macroeconomic environment, it may be easier to sell companies at their real market value.

In terms of privatization, important reforms should precede or be implemented simultaneously with privatization: institutional reforms, macroeconomic stabilization, and price and market reforms (Ahluwalia, 1996; Fischer and Gelb, 1991). The preconditions for a market economy are necessary—above all, a well-functioning institutional infrastructure that encompasses modern legal and regulatory frameworks and information systems (accounting and auditing), the right to unrestricted acquisition of private property, and the freedom to carry out economic activity protected by a legal system that provides the possibility of sanctions and redress.

The formulation of adequate regulatory frameworks is a key determinant of privatization’s success. The Latin American experience demonstrates that countries that put in place a modern regulatory system in advance of privatization (as, for example, Venezuela did when privatizing the national telephone company) are more successful than those that do not (Edwards, 1996). Institutional reforms have to encompass a tax system appropriate to a market economy and should ideally be in place prior to privatization.

Many arguments favor rapid liberalization of prices. In an environment of significant price controls, partial price reforms can promote intertemporal speculation, low supply response, and supply diversion. Moreover, partial reforms promote hoarding in anticipation of further price liberalization. As a result, inventories increase and the net supply of the product actually reaching the market declines. The shortage, in turn, creates rents and attracts lobbyists, who push for the continuation of policies that create rents. Thus, the political pressure against price decontrol mounts, increasing the probability that the government will prematurely abandon partial liberalization. For example, partial price decontrols in Brazil in the mid-1980s led to hoarding by suppliers and shortages in stores, strengthening political opposition against the reforms (Bresser, 1987). Partial liberalization can also lead to multiple price practices. Partial reforms encourage the diversion of essential inputs away from their traditional uses to enterprises that are not constrained to pay the regulated price and are willing to pay higher prices. Such diversion disrupts production in traditional industries, causing a decline in overall output and welfare.

With rapid liberalization, most of the disadvantages of a gradual approach disappear. Hording incentives decline, and market prices start to guide resource allocations earlier in the reform process. But if other distortions, such as rigidities in labor markets, exist, rapid adjustment could lead to greater misallocation of resources or even increased unemployment.

A number of arguments have been made in favor of rapid privatization (Hinds, 1991; and Lipton and Sachs, 1990). First, rapid privatization is the only way to weed out inefficiencies. The speedy transfer of control of enterprises to private shareholders is viewed as an indispensable condition for efficient management performance.19 Second, government reform and reorganization of public enterprises before liberalization are technically and politically difficult to coordinate, implement, and sustain. Third, in an economy with massive state intervention, quick privatization may be necessary to form a critical mass of private ownership to avert stalemate or reverses in the entire process. Fourth, support for rapid privatization is also motivated by the financial burden inefficient public enterprises impose on government budgets and the banking system. Without full-scale privatization, stabilization can be burdened by government subsidies for and bailing-out of public enterprises.

Despite these considerations, a number of practical factors limit the government’s ability to privatize rapidly. For example, redistribution of the assets of public enterprises is a thorny issue. In several transition economies, the process pitted the general public against workers demanding a distribution plan in their favor. Such disputes paralyze privatization. For instance, the Polish mass privatization program to redistribute public enterprises to the population at large was blocked by a coalition favoring greater redistribution to workers.

Other serious political constraints relate to uncertainty. The restructuring required to move to an incentive-based regime initially increases uncertainty because the improvements in enterprise performance the reforms seek to bring about take time to show themselves. The experience of transition economies—Lithuania, Russia, and Poland—that attempted rapid privatization and restructuring confirms the view that, ex ante, the quick approach had little chance of success because of the major restructuring problems encountered. In many cases, former managers retained control of the privatized enterprises. In the presence of aggregate uncertainty related to restructuring, a well-balanced approach, starting with the restructuring of only a subset of enterprises or sectors, may avert radical disruption of output, surge in unemployment, and steep fall in demand (Dewatripont and Roland, 1995).

IV. Guidelines for Reform

The economic thinking on the speed of adjustment and the sequencing of reform has been changing over time. Early recommendations were largely based on the experience of Latin American countries. More recent analyses have taken into account the experience of transition economies of Central and Eastern Europe and former countries of the Soviet Union, and recent crisis episodes. Overall, the foregoing discussion makes clear that the arguments favoring a gradual or a shock approach are not absolute. Each country has to choose the proper speed of adjustment and sequencing of reforms by examining country-specific factors. This concluding section builds on the previous discussion to provide a checklist of key questions for policymakers in determining the appropriate speed of adjustment and sequencing of reform programs.

  • What are the initial conditions? What are the extent of financial disequilibria and the degree of the structural bottlenecks? Other things being equal, the larger the disequilibria and the more severe the structural constraints, the longer it will take to reestablish financial balance and to alleviate bottlenecks. Large initial disequilibria and severe bottlenecks, however, do not necessarily call for a more gradual reform process. The effects of other considerations, as noted below, are also important.

  • How much financial resources are available to the country? The availability of financial resources can influence the speed of adjustment in two conflicting ways. On the one hand, if financial resources are scarce and financial imbalances large, the economy will need immediate and fast action, because a gradual process cannot be financed. On the other hand, for example, a country that is already burdened with massive debt may not have the financial resources to quickly implement all necessary structural reforms, thus reducing the speed at which reforms can be introduced.

  • What is the country’s institutional capacity to formulate and implement policy packages? When institutional capacity is weak, adjustment costs increase and the reform process becomes more difficult. Thus, all other things being equal, the greater the capacity limitations, the slower will be the speed of adjustment.

  • What are the political economy considerations? Political economy considerations affect the opportunity for and, thus, the speed of adjustment and reform. Factors that influence the window of opportunity include the time remaining in office before the next election for a democratically elected government, the perceived reform mindedness of the authorities, and the public tolerance to accept the transitional costs before reforms show positive results. If elections are close and short-term adjustment costs high, the government may be reluctant to embark on a far-reaching reform program. If special interest groups are directly affected, they are likely to campaign vigorously to slow down or stop the reform process. The authorities should always ask themselves what options they have to improve the opportunity for reform. Although some factors, such as reelection time, may be difficult to change, the authorities may be able to improve the environment for reform, for example, by introducing social safety nets.

  • How well can the authorities sustain the reform momentum? What can the authorities do to increase the reform momentum? Sustaining the reform momentum can greatly increase the speed with which policy objectives are achieved. The authorities’ commitment to reform and their ability to achieve broad public support and to maintain social consensus greatly influence their ability to sustain the momentum of reform. Opposition to reform, either from powerful interest groups or from particularly vulnerable groups adversely affected by the reform program, endangers the entire reform process (Olson, 1982). The more broadly the vision and the details of the reforms are shared with the wider public, the greater are the chances of the reform package’s success. In this regard, too, transparency and the communication strategy are important. To enhance transparency, the authorities must make their reform intentions public, clarifying their priorities and the rationale for the sequencing of reform elements. Public support is also enhanced by the government’s communication strategy, as the investor relations offices, for example, of Brazil and Mexico show. Similar to public relations programs of multinational companies, these programs help the domestic community and international investors to better understand the specifics of a reform program and economic development in general (Institute of International Finance, 1999). An effective communications strategy is particularly important for countries that are already actively participating in the international flow of capital or countries that are starting to have access to international capital markets.

  • How can the authorities increase the credibility of the reform program? The credibility of the authorities is likely to depend largely on their track record. Once credibility is lost, it takes a long time to regain. Although a new government might benefit from a “starter bonus,” the government’s credibility to pursue reforms may need to be established as its determination has never been tested. When credibility is weak, private agents may fail to respond to the reform measures as expected, resulting in higher adjustment costs. This means that if the authorities lack credibility, a more front-loaded approach may be needed to regain credibility. By contrast, a government with high credibility can generate a positive public response to reforms, even if the reforms are being introduced gradually. Thus, the authorities must always review the manner in which they attempt to raise the credibility of their efforts by putting in place a consistent and realistic program. For example, the authorities can enhance the credibility of a program by enacting new laws or introducing institutional changes, or by making commitments in the context of programs or projects supported by international financial institutions.

  • How ambitious are the reform objectives in terms of quality and quantity? Reform objectives may be of a qualitative or a quantitative nature. Qualitative reform objectives refer to such changes as the transformation of an economy from a centrally planned to a market economy or to improvements in health and education services. Quantitative objectives refer to specific targets for macroeconomic or structural variables, such as budget deficits in percent of GDP or number of students per teacher. Both the scope and the quantitative targets of reforms affect the speed with which reform objectives can be attained. There is always tension between politically motivated reform promises and realistic reform objectives. However, to increase the chances of success, the authorities need to specify reform priorities and be realistic in terms of what is achievable over a specific time.

  • What are the policy options? The same overall reform objectives can be achieved through different sets of policies, with different adjustment costs and distributional effects. For example, in a fixed exchange rate environment, external adjustment may be envisaged either through a devaluation of the exchange rate and supporting policies or solely through restrictive fiscal and monetary policies. Depending on the structural rigidities in the economy, restrictive fiscal and monetary policy can result in higher short-term adjustment costs, particularly in terms of output, but also lower inflationary pressures than the first option. This example highlights the importance of comparing options and identifying trade-offs. The careful and public consideration of alternative policy options can enhance program ownership, as the debate on the alternative options can help to create a common understanding among the authorities, the general public, and the international community on the most appropriate policy package for the economy with a generally agreed upon speed.

  • What are the contingency provisions needed to address potential reform slippages or unforeseen shocks? Contingency provisions relate to specific actions to be taken if the expected results of certain measures are not realized. For example, if the revenue effects of certain tax measures are not fully realized, the contingency provisions may provide for a cutback in specific expenditures. If no specific contingency provision is included, the program may provide for a mechanism, such as a policy committee, to decide on the policy response. In the case of an external shock, for example, the response will depend on the likely nature of the shock, distinguishing between demand or supply shocks, transitory or permanent shocks, and the size of the shock. To the extent that adequate contingency provisions or mechanisms are in place, the policy response to events that could adversely affect the adjustment and reform effort would be speeded up resulting in the achievement of the objectives in a faster time frame.

  • While the consideration of the above questions can help policymakers in determining mainly the speed of the adjustment and reform process, four key factors can guide policymakers in their sequencing decisions:

  • Preparatory time: The time needed to prepare and implement different reforms can vary substantially. For example, the introduction of a value-added tax requires the formulation and approval of the requisite legislation, the build-up of the administrative infrastructure, and the establishment of the appropriate infrastructure to transfer payments. It would thus be unrealistic to try to implement a value-added tax without giving adequate weight to the time required for sequencing all of the preparatory steps.

  • Prerequisites: Specific reforms are often prerequisites for other reforms. Certain institutional reforms are necessary before other reforms can be implemented, and some progress in stabilization may be necessary before pursuing other reforms. For example, the lifting of a government monopoly in the import and distribution of certain products will first require the liberalization of import procedures and price controls to remove distortions. Similarly before opening the capital account, the financial sector must be strengthened. Financial stabilization is also a fundamental prerequisite for other reforms.

  • Complementarities: The understanding of reform complementarities is critical in determining appropriate sequencing. For example, there can be a complementarity among fiscal policy, trade liberalization, and public enterprise reform. Efforts to reduce the budget deficit can be facilitated by replacing quantitative trade restrictions with tariffs, by receipts from the privatization of public enterprises, and by reducing subsidies to unprofitable public enterprises. The same applies to the introduction of a social safety net to address the transitional employment and distribution effects of the adjustment process and specific reforms.

  • Conflicting instruments: Sequencing must take into account the potential conflict between different policy instruments. For example, trade liberalization may need to be supported by a real exchange rate depreciation, whereas capital account liberalization can result in a real exchange rate appreciation.

In conclusion, speed and sequencing considerations are closely linked. The optimal speed and sequence is country specific. The checklist outlined above should help policymakers evaluate the different policy elements in a comprehensive package to determine how best to sequence reforms and at what speed to proceed.

The Speed of Adjustment and the Sequencing of Economic Reforms: Issues and Guidelines for Policymakers
Author: Mr. Saleh M. Nsouli, Mr. Mounir Rached, and Mr. Norbert Funke