Reform: what pace works best? Gradual, says John McMillan, who argues for a step-by-step...

Avoid Hubris and other lessons for reformers

THE PAST two decades have seen a worldwide shift to markets. Globalization has opened domestic markets to international competition. The former communist countries have converted themselves, to varying degrees, into market economies. In the

The Asian financial crisis of 1997-98 caused severe hardship for millions of ordinary Indonesians, Koreans, and Thais. Russia's 1991 market reforms were followed by a decade of negative growth. In Latin America, privatization was often so corrupt that, in polls, a clear majority now say it was not beneficial.

The good news, however, outweighs the bad. India and China are growing. The average Chinese is three times as well off now as in 1980; the average Indian is twice as well off. It is impossible to exaggerate the consequences of rapid growth in two very large, very poor countries.

The good news goes beyond China and India. Trade opening and privatization in various developing and former communist countries, from Uganda to the Czech Republic, have mostly brought their intended benefits. However, the improvements, while positive overall, have been spotty: often slow in coming and relatively small. The gains from market reform are inconspicuous; you have to look hard at the data to see them (as the studies reviewed below do).

One of the lessons for reformers is that the efficacy of any one policy depends, in ways that are often unforeseeable, on other policies. The chief lesson is this: Avoid hubris.

Reform gains

Studies of countries lowering their trade barriers have found that, although the consequences have varied from country to country, trade reform has usually improved economic performance and consumers have benefited from the lower prices. A substantial trade opening boosts economic growth, on average, by an estimated 1.5 percentage points (Wacziarg and Welch, 2003). The measured benefits from trade opening are significant but not huge.

Developing countries that have opened themselves to investment from overseas have similarly benefited. In countries that began to allow foreigners to hold shares in domestic firms (a diverse group, including Brazil, Indonesia, and Nigeria), investment rose. Growth averaged 1.1 percentage points higher after liberalization than before (Henry, 2003).

With privatization, the story is similar: genuine but unspectacular improvements. After being privatized, the typical firm raises its labor productivity, increases its investment, and lowers its prices. Studies of 211 firms in more than 50 countries comparing firm performance for the three years before privatization with the three years after (Megginson and Netter, 2001) find that investment as a percentage of sales rose an average of 5 percentage points. Because state firms produce only a fraction of GDP, such improvements translate into a small gain in aggregate growth.

Is tackling corruption, by improving the rule of law, another part of the solution? The cross-country data show that corruption significantly slows growth (Mauro, 1995). If India were to cut its corruption level to, say, Italy's, then, according to these regressions, its growth would rise by 1 percentage point. As with any other single policy, cracking down on corruption helps but is no panacea.

The sum of these estimated growth effects of trade opening, financial liberalization, privatization, and corruption lowering is a little less than 4 percentage points. This total should be taken with a grain of salt. The estimates are rough. Also, adding them may overestimate their combined effect (some countries liberalized trade and investment simultaneously, so there may be some double counting) or underestimate it (as discussed below, the whole of reforms may be more than the sum of their parts). Caveats aside, how much difference would such an increase in growth make?

Boosting the rate of growth from, say, 3 percent to 7 percent means that it would take 10 years for national income to double, instead of 23 years. (To calculate how long it takes for income to double, divide the growth rate into 69.) If an African country with zero growth were to liberalize and attain 4 percent growth, people's standard of living, instead of stagnating, would double every 17 years. Africans would steadily become less poor but would still be very poor.

In the United States, per capita income is $35,000. In Tanzania, it is $550. The gap in living standards is 64 to 1. The measured gains from trade opening, privatization, and anticorruption policies do not go far in shrinking that gap. Any solution to global poverty will entail transforming the poor countries' economies. What have we learned from two decades of reform experience?

Iraq suggests we may not have learned much at all. In 2003, the American-run Coalition Provisional Authority announced "big bang" reforms for Iraq. The intention, according to The Economist (September 27, 2003), was to "abruptly transform its economy into a virtual free trade zone." The reforms, gushed The Economist's reporter, were "bright economic news," which promised to "yank Iraq back into the global economy in record time" Unfortunately, even given an end to the violence, these reforms probably wouldn't. The fact that, in 2003, the American officials advising Iraq would recommend big bang reforms demonstrates, I contend, the remarkable longevity of bad economics.

Reform is hard to do because we cannot predict its effects. The big bang approach presumes we know where we are going and how to get there. We may know where we should be headed, but there is much we do not know about how to get there. No recipe for success has yet been written. Acknowledging our ignorance means moving step by step rather than betting everything on a comprehensive blueprint (McMillan and Naughton, 1992). The whole point of the market economy, after all, is that it handles, better than any more centralized alternative, the unforeseen and the unforeseeable. If we could plan the reforms, we could have planned the economy.

System renovation

Why is market building difficult? Reform means transforming a country's entire economic system. The various parts of a reform package reinforce each other. A pair of reforms may be complementary, meaning that unless one is already in place, the other is ineffective. A potentially worthwhile reform could even be harmful if its complementary reforms are missing.

Trade liberalization is a case in point. For some countries, trade opening did not bring its intended benefits. In a sample of 24 countries that substantially liberalized their trade (Wacziarg and Welch, 2003), about half subsequently achieved faster growth. For example, Chile grew 2.8 percentage points faster and Uganda grew 2.2 percentage points faster. One-fourth obtained no improvement in growth and one-fourth actually grew more slowly than before. For example, the Philippines grew at the same rate as before, and Mexico grew 2.2 percentage points more slowly.

The effects of trade reforms have varied widely from country to country. There is a lot of variation around the average 1.5 percent boost to growth cited earlier. It is widely agreed among those who have studied reform that opening an economy to trade is good policy--but only if the country is ready for it.

Trade opening might not be beneficial if the labor market is distorted. The gains from trade reform come, in principle, through reallocating labor. Workers move within a given industry, from less efficient firms to more efficient ones, and across industries, from those that had been protected to those in which the nation has a comparative advantage. After Brazil lowered its trade barriers in the early 1990s, for example, firms revamped themselves. However, most of the efficiency gains came from workers moving within their own industry. Contrary to trade theory textbooks, there was little movement of workers across industries. Moreover, many of those thrown out of work were not rehired but became chronically unemployed, scraping by in the informal sector. The workers bore high transitional costs (Muendler, 2003). In other countries, the pattern was similar. A study of 25 countries opening their trade found that little structural change had occurred (Wacziarg and Wallack, forthcoming). The transition from declining industries into industries of the future is held back if the labor market is not up to the task. Labor market reforms complement trade reforms.

Similarly, whether privatizing a state-owned firm improves its performance depends on its economic environment. Ownership incentives alone are not enough to induce large firms to be run efficiently. Also needed are the oversights that come from well-functioning (and well-regulated) financial markets. As the collapse of the energy company Enron illustrates, these oversights sometimes fail even where financial markets are sophisticated. Where financial markets are underdeveloped, Enron-type problems are pervasive. After Mexico's banks were privatized, for example, their managers engaged in "related lending," making loans to themselves on generous terms. The Russian natural-gas giant Gazprom has a market value far below its assets, reflecting stockholders' skepticism after the managers sold the firm's gas deposits for a fraction of the billions they were worth. While asset stripping is not universal--the data show that privatization has usually improved firm performance--it can occur if privatization is pursued in isolation of other reforms. Financial-market oversights complement privatization.

Wide-ranging reform is what is needed--but it is hard to do, even when the setting is favorable. When New Zealand radically deregulated in the mid 1980s--slashing trade barriers, removing agricultural price supports, flattening taxes, trimming financial-market regulation, corporatizing and privatizing state-owned firms--it suffered a major recession, with negative growth and high unemployment. It took more than a decade for any benefits to show. An industrial country, New Zealand had started with a full set of market-supporting institutions, like laws of contract, and its labor and financial markets had been operating reasonably effectively. The reform process can be expected to be even more painful in a country like Mexico or Turkey, where the legal system is creaky and labor and financial markets suffer from high transaction costs, and more painful still in a country like Tanzania or Bangladesh, where market mechanisms are stunted.

Social engineering

The need for a package of reforms could be a reason, on the face of it, for doing everything at once. But here is the rub: it is hard to predict how the pieces of the market system will fit together. Some systemic interactions may be complex and indirect, making it difficult to anticipate their existence. Others are straightforward, like the interaction between trade reform and the labor market or between privatization and the equity market, but we have little data on the magnitude of interaction.

Within each individual market, also, there are systemic complexities. Building a smoothly functioning labor or financial market is a far harder task, and more time-consuming, than freeing up trade or privatizing firms. Imagine being charged with creating an equity market in a country that lacks one, like Myanmar. The equity market relies on trust: investors hand over their money to managers with little direct assurance that it will not be misused. To sustain that trust, it is necessary to build a mix of private sector and government based mechanisms (Black, 2001). The firms that oversee the market--accountants, investment banks, law firms--must have, and want to keep, a reputation for trustworthiness. Such a reputation is developed only over time. Self-regulating organizations--the stock exchange, with its rules on financial reporting and its sanction of delisting, and various voluntary industry associations help keep the players honest. A vigorous business press is needed to scrutinize companies' dealings. To prevent managers from expropriating investors' funds, the government must write laws and train judges. To ensure that investors receive accurate information, the government must set up a regulator like the U.S. Securities and Exchange Commission. The workability of each of these equity-market mechanisms is contingent, in ways we cannot fully anticipate, on the presence of the others.

Half a century ago, in The Open Society and Its Enemies, Karl Popper contrasted two modes of reform. Utopian social engineering, with a grand blueprint for society, "pursues its aim consciously and consistently" and "determines its means according to this end." Piecemeal social engineering, by contrast, involves tinkering with parts of the system, with no overall plan. Whereas piecemeal reform entails "searching for, and fighting against, the greatest and most urgent evils of society," utopian reform entails "searching for, and fighting for, its greatest ultimate good."

Although Popper acknowledged that the utopian approach was "convincing and attractive," with its appeal to rational thought, he argued that it was folly. The utopian approach "demands a strong centralized rule of a few." By contrast, the piecemeal approach can succeed. Because it acknowledges that "perfection, if at all attainable, is far distant," the piecemeal approach is "the only method of improving matters which has so far been really successful, at any time, and in any place" (Popper, 1971, pp. 157-59).

Popper's argument that piecemeal methods work better than utopian--already tested and affirmed by (utopian) communist central planning--was further tested by the move away from communism. China tried piecemeal reform; Russia, utopian.

Big bang reform was characterized by its architect, Jeffrey Sachs, as "a rapid, comprehensive, and far-reaching program of reforms to implement 'normal' capitalism." Sach's characterization fits Popper's definition of utopian social engineering, in being "comprehensive" and having a stated endpoint, "normal capitalism." Russia's income declined after reform. The numbers overestimate the drop in living standards, as the unmeasured underground economy grew, but after correcting for that, Russia's growth was still negative. If retail sales and electricity consumption are used as indicators of economic activity, Russia's 1999 economy was operating at about 80 percent of its 1990 level (Shleifer and Treisman, 2004). The consequences of Russian big bang reform corroborated Popper's dismissal of grandiose schemes.

China's reforms, bringing spectacular income growth of around 8 percent per capita for 30 years, were piecemeal. Each reform was tried out on a small scale and expanded if it worked. The endpoint was undefined. In Deng Xiaoping's folksy formulation, China was "crossing the river by feeling each stone" When pressed on where China was headed, its leaders said they wanted a "social market economy with Chinese characteristics"--a phrase that is empty of meaning, and presumably intentionally so.

Where to start?

The modest (or piecemeal) reform prescription is to start with something that seems feasible and sensible. In 1980s China, for example, early success came in agriculture. When the collective farms were abolished and replaced with individual plots, China's food supply doubled. The collectives were a hopelessly inept way to organize farming, as the farmers themselves were well aware, so easy gains were to be had.

A willingness to experiment with novel incentive devices and new organizational forms is sometimes part of successful reform (McMillan, 2002). China's other early success was in the creation of village enterprises: small manufacturing firms owned and run by village governments. The reformers failed to anticipate the village enterprises' rapid growth, which "took us by surprise completely," as Deng later said. With hindsight, though, these firms' success is explicable: they found a ready sale for their products by filling empty market niches; they created jobs; they obtained finance, otherwise unobtainable, via the villages' powers of taxation; and they were disciplined by the intense product-market competition among themselves that quickly developed.

In reform, one size doesn't fit all. What works varies with the country's initial conditions. That village-owned firms were the main source of China's first decade of growth demonstrates that the specifics of time and place matter. Village-owned firms are unsuited to present-day China and are unlikely to be the solution in another country.

Nonstandard policies sometimes work, at least for a while. Much as India and China have benefited from connecting themselves to the global economy, they are not free traders. They have grown despite keeping sizable trade barriers. (Through the 1990s, China maintained bureaucratic controls over imports of goods and currency, while India's tariffs averaged 40 percent and nontariff barriers were widespread.) India and China are counterexamples of the proposition that full trade opening is necessary for economic growth. They do not represent an argument against open trade--the theoretical and empirical case for it is well substantiated--but they do warn against seeing free trade as a universal remedy. The case for open trade is not helped by overselling.

The honest approach to economic reform is to be deliberately experimental. To claim we can do something more premeditated than trial and error is to exaggerate our knowledge of reform processes: Avoid hubris.

References

Black, Bernard, 2001, "The Legal and Institutional Preconditions for Strong Stock Markets," UCLA Law Review, Vol. 48 (April), pp. 781-855.

Henry, Peter Blair, 2003, "Capital Account Liberalization, the Cost of Capital, and Economic Growth," American Economic Review, Vol. 93 (May), pp. 91-96.

Mauro, Paolo, 1995, "Corruption and Growth," Quarterly Journal of Economics, Vol. 110 (August), pp. 681-712.

McMillan, John, 2002, Reinventing the Bazaar: A Natural History of Markets (New York: Norton).

--, and Barry Naughton, 1992, "How to Reform a Planned Economy," Oxford Review of Economic Policy, Vol. 8 (Spring), pp. 130-43.

Megginson, William L., and Jeffry M. Netter, 2001, "From State to Market," Journal of Economic Literature, Vol. 39 (June), pp. 321-89.

Muendler, Marc, 2003, "Trade and Growth: The Case of Brazil," Department of Economics, University of California, San Diego, http://econ.ucsd.edu/seminars/muendler_brazil.pdf.

Popper, Karl, 1971 [1945], The Spell of Plato, vol. 1, The Open Society and Its Enemies (Princeton, New Jersey: Princeton University Press).

Sachs, Jeffrey, 1994, "Shock Therapy in Poland," University of Utah (April), http:www.tannerlectures.utah.edu/lectures/sachs95.pdf.

Shleifer, Andrei, and Daniel Treisman, 2004, "A Normal Country," Foreign Affairs, Vol. 83 (March), pp. 20-38.

Wacziarg, Romain, and Jessica Seddon Wallack, forthcoming, "Trade Liberalization and Intersectoral Labor Movements," Journal of International Economics.

Wacziarg, Romain, and Karen Horn Welch, 2003, "Trade Liberalization and Growth," Stanford University's Graduate School of Business, www.stanford.edu/~wacziarg/downloads/integration.pdf.

John McMillan is the Jonathan B. Lovelace Professor of Economics at Stanford University's Graduate School of Business in California.

Avoid Hubris but Acknowledge Successes

Lessons from the postcommunist transition

JOHN McMillan gives a wide-ranging critique of market reforms around the globe, arguing that they have provided, on the whole, "spotty" benefits, in many cases at considerable cost. In response, I will not attempt to cast as wide a net; rather, with the humility of my more limited recent experience, I will focus on the postcommunist transition, showing that comprehensive and rapid reforms have led to greater success than a gradual, step-by-step approach.

The evidence of postcommunist transformation (China's case deserves a brief but separate treatment) shows many successes, many mistakes--but rarely only in the economics of reform--and several cases where the process is unfinished and hence not subject to declarations of victory or defeat. I agree with McMillan that the experience of "big bang" reforms justifies the caution to "avoid hubris" because some mistakes--such as privatizing too fast--created a new problem of state capture and underdeveloped institutions, which I discuss later. Nevertheless, the insufficiently recognized successes of comprehensive and rapid reform can be a cause for pride.

A yardstick for progress

First, let me define how I measure progress and the countries that can be considered examples of the big bang approach. Many measures of progress in transition exist, such as indicators of economic growth, financial stability, democratization, and establishment of institutionalized rule of law. But, as I show elsewhere (Havrylyshyn, 2004), the index of transition indicators (TI), updated annually by the European Bank for Reconstruction and Development (EBRD), serves as an excellent proxy, despite certain shortcomings. Where relevant, I add measures of more direct benefits and costs of transition.

By this yardstick, how far have the transition economies progressed since the fall of the Berlin Wall? Chart 1 summarizes progress with reform on a TI scale of 1-4.3 (the last value representing a fully functioning market economy). It shows five distinct groupings: Central Europe (CE) is most advanced, with the Baltics at virtually the same level; Southeast Europe (SEE) is distinctly less advanced but, in general, somewhat ahead of countries in the Commonwealth of Independent States (CIS). The CIS cleanly divides into two groups: those with moderate progress (CISM) and those with very limited movement away from the Soviet economic regime (CISL). The latter comprise Belarus, Uzbekistan, and Turkmenistan.

I am not aware of any effort to categorize systematically the countries that undertook big bang, as opposed to gradual, reforms. But the TI measure allows identification of "early reformers," the ones with the highest scores by 1994, when the EBRD began its index. That list is identical to the leading Central European and Baltic groups shown in the chart. I propose to identify these countries as big bang cases. Perhaps the only exception is Hungary, which had quietly done so much in the communist period that it was able to continue at a relatively gradual pace. I will therefore put Hungary in the "gradual reform" category (which hurts my argument and helps the critics of big bang reforms).

How should we consider Russia? Russia's reform effort of 1992-93, in particular, is considered by critics as a big bang case, though most proponents argue that it was a far cry front Poland's big bang approach, and, in any event, much of the reform was undone in 1994-95 and had to be recouped later (Aslund, 1995). Nevertheless, I will classify Russia as a big bang case, again helping the argument of the critics.

Within the CISM grouping, only Armenia made an early attempt at rapid reform, although it was soon stalled by civil conflict and a yielding to populist politics; all the others in the CISM category followed a gradualist path. Thus, I identify 9 rapid reformers, 15 gradual reformers, and 3 with very limited reforms.

What the record shows

How did these groups fare economically, socially, and politically? The already visible economic benefits of transition are summarized in Tables 1 and 2, which show the index of output recovery since 1989, latest inflation levels, and per capita values of foreign direct investment. It is evident that the economic performance rankings are much the same as the TI measure of progress in reforms, save, perhaps, for recovery of output. The CISL value of 96 is well above that of the CISM at 55. This is often a key piece of evidence against rapid reforms, with critics noting that these countries avoided a huge output fall by pursuing gradual reform. Even if true, this may mean only that the restructuring other countries experienced has not begun in the lagging group. Recent data in IMF reports for Belarus and Uzbekistan show much lower growth than in the CISM group. Furthermore, there is good evidence that growth in earlier years was considerably overestimated because of favorable barter terms and statistical issues.

While it appears that only Central Europe has fully recovered and begun to see positive output benefits, this is not really the case because the official Soviet GDP in the 1989 bench mark year underestimated the value of services, overestimated unusable and unsalable products, and took no account of the frequent nonavailability of consumer goods or the time consumers spent standing in line for products. Aslund (2001) adjusts the output index for only some of these problems, finding that in Central Europe virtually no fall in output occurred and that most countries were well beyond the initial level by 2001. Though Aslund is surely correct about the direction of bias, one study is insufficient to be conclusive. Therefore, I have cut by half his implicit adjustment factors and calculated some illustrative values shown in the second column of Table 1. Even these smaller corrections suggest that, by 2002, Central Europe may have reached 150 percent of initial GDP; the Baltics, 111 percent; SEE, 97 percent; and CISM, 74 percent. For the CISL--essentially unreformed Soviet economies the original index value is arguably appropriate. These conservative adjustments reinforce the evidence that more rapid reformers performed better than gradual ones. The conclusion is, surprisingly, even stronger if one looks at the social costs and democratization progress.

Grun and Klasen (2001) voice a common view of critics when they state that "socialist countries had enjoyed relatively high levels of well-being [but], in the transition, rising inequality and falling incomes have led to a dramatic deterioration." There was undoubtedly a deterioration, but one should ask how big it was and, of relevance here, what differences can be found between rapid and gradual reformers. For this purpose, the Human Development Index (HDI) of the United Nations Development Program is a very useful tool. Its values for the period are shown in Table 3.

Central Europe's performance is even better when measured by the HDI than by standard economic indicators, with HDI showing no decline even in the first half of the decade. As one goes down the groups, an increasing deterioration is seen. In the Baltics, a strong deterioration is evident, but, by 2001, it is fully recouped and surpasses the initial value; the same applies to Southeast Europe. The sharpest deterioration is seen in the CISM group of moderate reformers, though a partial recovery has taken place since 1995. The CISL group did not avoid social costs, but less adjustment meant that costs were lower; I leave to future analysts to determine if this means a successful minimization of social costs or merely a postponement. On balance, it seems clear that the correlation between rapidity of reforms and social costs is not positive but negative.

On progress toward democracy, political scientists are as divided as economists. But the important prediction of Przeworski (1991) that market reforms can be effectively implemented only in a less than fully democratic regime does not seem to be upheld. As McFaul (2002, p. 221) explains it: "Many had predicted ... that reorganization of economic institutions would undermine democratic transition.... To the contrary, those countries that have moved the fastest on economic transformation also showed the greatest success in consolidating democratic institutions." One broad-brush indicator of this view that economists and others can understand is the striking result in Chart 2, showing a strong positive relationship between the degree of democracy and progress in market reforms, rather than a negative one. There is room for discussion on the endogeneity of the relationship, but recent Freedom House indicators of democracy, showing deterioration for many CISM countries after an initial improvement, suggest the causation is not in all cases simply from democracy to reforms, or from gradual reforms to democracy.

The pitfalls of privatization

McMillan seems to accept the reformist contention that privatization is generally beneficial economically; this is especially valid for transition economies where very recent microlevel studies nearly always show that some benefit comes even from privatization to insiders. I have summarized these findings (2004) but offer two important caveats. First, the benefits of privatization without a proper accompanying climate of open competition and rule of law may be very small or even zero (Zinnes, Eilat, and Sachs, 2001). This is probably why the empirical studies for CIS reformers do not show as strong results as in Central Europe. Second, where privatization has resulted in a strong concentration of ownership, a class of rent-seeking vested interests develops and "captures" the state to ensure that policies work in their favor, prevent competition, and slow further institutional and democratic development (Hellman and Schankermann, 2000, reporting a World Bank study). These unintended consequences of speeding up privatization by co-opting insiders, an approach common to many gradual reformers (CISM and SEE), may have been the most important error of reform advocates and certainly one area where humility is called for.

Notably, state capture by large vested interests did not occur primarily in rapid reformers. In the World Bank studies designating 10 captured states in 1998, only 2 were clearly early, rapid reformers--Croatia and Slovakia--and both have experienced a major political change since then that has arguably reversed the situation. If, for the sake of avoiding bias, Russia is defined as a rapid reformer, then at most three captured countries might be big bang cases, whereas at least seven pursued gradual reforms.

Emulating China?

McMillan, like many critics, suggests that China's gradual reforms and its far superior economic growth performance would have been the better path to follow. There certainly are some lessons from the Chinese experience, but, for the most part, it is difficult to imagine it could have been repeated to good effect in the European cases, given the different political environment. Centralized control did not, as McMillan (adducing Karl Popper) suggests, lead to a utopian approach in China but, on the contrary, afforded it the luxury of a piecemeal approach. In Europe, the collapse of communist polities made gradualism politically risky.

China's large agricultural sector at the outset (about 70 percent in the late 1970s compared with 20 percent in 1989 for the Soviet Union) allowed a costless gradualism. The Chinese situation was clearly akin to Arthur Lewis's famous development model in which surplus labor from agriculture was available for transfer to new industrial activities with no loss of agricultural output. For the Soviet bloc economies, labor for a new private sector could come only from a sharp reduction of the state industrial sector. As the differences between Central Europe and the CIS show, where gradualism was applied, if anything, it made things worse, because the conditions for new enterprises to take root and create new jobs were slow in coming.

Don't sell the big bang short

How does all this relate to the criticism that the big bang approach leads to greater output losses, higher social costs, and slower introduction of good institutions? The updated statistical evidence suggests the contrary. A major error of many critics, including economists like McMillan and Joseph Stiglitz and political scientists like Peter Reddaway and Dmitri Glinski, has been to concentrate too much on Russia, its institutionally inimical privatization process, and early dramatic declines in well-being, and too little on the successful cases of big bang reforms. These should include not only the well-studied and well-publicized first case of Poland but also most of Central Europe and the Baltics. Even if one cedes to the critics that Russia was a case of big bang and Hungary followed a gradualist path, the cross-country facts remain that, of nine rapid reformers, eight have done very well. Most of those that chose a more gradual path have had fewer economic benefits and, surprisingly, higher costs, manifested in the deterioration in indicators of well-being and stalled progress in democratization.

Could the recent surge in growth in many CISM countries to annual rates of about 5 percent reverse this conclusion? Perhaps, but only if these rates are sustained for a long time. This is still a big if: a key determinant will be whether the vested interests (or oligarchs, in popular parlance) that developed under slower reforms and through insider privatization will continue to wield their strong influence on state policies to protect their privileges or will be willing to give these up and press to ensure their rights over property already acquired. Establishing the general rule of law, without special privileges, would benefit entrepreneurs as a whole and contribute to the welfare of society at large. Until that happens, the evidence indicates that rapid reform has generally been better.

Table 1
Fruits of transition
Central Europe and the Baltic states have made the most progress
since the end of central planning.

                          Recovery: Estimated index of GDP 2002

                         (1989 = 100, with arbitrary adjustment
                                for Soviet accounting (1))

                     Unadjusted 2002 value     Arbitrary adjustment

Central Europe                120                      150
Baltics                        82                      111
Southeast Europe               81                       97
CISM                           55                       74
CISL                           96                       96

Sources: Unadjusted data from the EBRD's Transition Report 2003;
adjusted data from author's calculations based on Aslund (2001).

Notes: CISM = Members of the Commonwealth of Independent Stales (CIS)
that have made moderate progress. CISL = CIS members that have made
limited progress.

(1) Many have argued that Soviet income accounting overstated GDP,
hence, the recovery from 1989 levels has been much higher than simple
arithmetic (Column 1) shows. I have made an arbitrary adjustment to
the EBRD index value of 2002 as follows: Central Europe, 25 percent;
Southeast Europe, 20 percent; Baltics and CISM, 35 percent; CISL,
no change.

Table 2
Bleak spots
Soviet bloc countries that reformed the least had the highest
inflation and attracted the least foreign investment.

                                         Inflation
                            (consumer price index, 2002; percent)
                    Group
                   median   Lowest country       Highest country

Central Europe        2.3   Poland (2.1)         Slovenia (7.4)
Baltics               2.3   Lithuania (0.9)      Estonia (3.8)
Southeast Europe      5.9   Macedonia (3.6)      Romania (22.7)
CISM                  5.2   Ukraine (1.6)        Russia (15.4)
CISL                 12.7   Turkmenistan (9.6)   Belarus (41.4)

                            Foreign direct investment per capita
                               (cumulative 1989-2002; dollars)
                   Group
                 Average    Lowest country       Highest country

Central Europe      1,600   Poland (1,000)       Czech Rep. (3,400)
Ball                1,400   Lithuania (1,000)    Estonia (1,800)
Southeast Europe      384   Serbia (190)         Bulgaria (547)
CISM                  279   Tajikistan (21)      Kazakhstan (950)
CISL                  142   Uzbekistan (36)      Turkmenistan (210)

Source: EBRD, Transition Report 2003.

Note: For definitions of CISM and CISL, see Table 1.

Table 3
Varying performance
From a human development perspective, Central Europe fared well, but
moderate reformers in the Commonwealth of Independent States saw a
sharp deterioration.

(average Human Development Index value; maximum 2001, Norway = 0.944)

                               1990      1995      2001

Central Europe                0.812     0.819     0.845
Baltics                       0.812     0.779     0.822
Southeast Europe (3 of 6)     0.752     0.749     0.774
CISM                          0.768     0.717     0.737
CISL                          0.760     0.739     0.760

Source: UNDP, Human Development Report 2003.

Note: For definitions of CISM and CISL, see Table 1.

References:

Aslund, Anders, 1995, How Russia Became a Market Economy (Washington: Brookings Institution).

--, 2001, "The Myth of Output Collapse after Communism," Working Paper 18 (Washington: Carnegie Endowment for International Peace).

European Bank for Reconstruction and Development, 2003, Transition Report 2003 (London: EBRD).

Grun, Carola, and Stephan Klasen, 2001, "Growth, Income Distribution, and Well-being in Transition Economies," Economics of Transition, Vol. 9 (July), pp. 359-94.

Havrylyshyn, Oleh , 2004, "Uncharted Waters, Pirate Raids and Safe Havens: A Parsimonious Model of Transition Progress," paper delivered to the 10th Dubrovnik Economic Conference of the Croatian National Bank, June 23-26, Dubrovnik, www.hnb.hr.

Hellman, Joel, and Mark Schankermann, 2000, "Intervention, Corruption, and Capture," Economics of Transition, Vol. 8 (November), pp. 545-76.

McFaul, Michael, 2002, "The Fourth Wave of Democracy and Dictatorship: Non-Cooperative Transitions in the Post-Communist World," World Politics, Vol. 54 (January), pp. 212-44.

Przeworski, Adam, 1991, Democracy and the Market: Political and Economic Reform in Eastern Europe and Latin America (Cambridge: Cambridge University Press).

United Nations Development Program, 2003, Human Development Report 2003 (New York: UNDP).

Zinnes, Clifford, Yair Eilat, and Jeffrey Sachs, 2001, "The Gains from Privatization in Transition Economies: Is Change of Ownership Enough?" IMF Staff Papers, Vol. 48 (Special Issue), pp. 146-70.

Oleh Havrylyshyn was Deputy Director of the IMF's European II Department and is now on sabbatical at the University of Toronto. He served as Ukrainian Deputy Minister of Finance in 1992 and was Ukraine's Alternate Executive Director at the IMF in 1993-96.

[GRAPHICS OMITTED]