Radical/Shock Versus Gradualism/Sequencing Versus Minimum Bang

This issue does not bear lengthy discussion: it is another example of slogans displacing analysis. There is confused, misleading argument over whether Russia did or did not try shock therapy; over whether socalled Polish radicalism has been more successful than so-called Hungarian gradualism, with no attention paid to the initial conditions in each country. This rhetoric is understandable when it comes from politicians, inexcusable from economists.

Here, I have not abandoned my early view that a “robust sequence” should follow a “credible regime change.” The opening package of measures must be sufficient to make the regime change credible, but it is administratively impossible and economically unwise to try to do everything at once. The legal, institutional, economic, and behavioral infrastructure of the capitalist market economy cannot be installed quickly and by fiat, except possibly by a government as authoritarian as the worst of those overthrown in the revolutions of 1989. Thus, policy-makers must choose a sequence. It is better to design that sequence consciously, insofar as possible, rather than just to follow the political winds.

This position may well be “gradualism,” but that is widely (often intentionally) misinterpreted: a gradualist program for a sequence of policy measures is just the opposite of the uncoordinated improvisation of which gradualism is often accused; if it really is drift, then it is not deliberate policy. On the other hand, avoiding drift in a gradualist program does require some ability to precommit, as well as to retain some credibility when political imperatives or exogenous shocks force reoptimization. That is hard, so trying to do everything at once may actually seem easier. It is not, nor is it feasible.

We should therefore beware of those who take the conventional refuge of the extremist ideologue in claiming that the program has not failed, it has just not been implemented fully and consistently. That is a hypothesis that is in principle not testable, because it will always be possible to find a “key” element of the program that could not be applied.

In any case, the range of sensible strategies is limited, and there may be little margin for choice. Some elements of stabilization and liberalization make sense only when done simultaneously. The range observed across countries is in fact surprisingly limited and is mainly a function of initial conditions. Foresight, sequencing, implementation, and political support are the key differences and potential weaknesses. And no country fits particularly well the stereotype that the ideologues assign to it: “gradualist” Hungary was much more “radical” than Poland in its implementation of bankruptcy legislation, and Czechoslovakia was more “radical” than either with its voucher privatization program.

The choices posed by sequencing are not merely sources of academic dispute. There are serious, major decisions: about the priority to be given to privatization of large state-owned enterprises relative to other institutional changes and demonopolization; the urgency of creating a healthy structure of financial intermediation; and many others. These choices are conditioned by political feasibility, but for an economic policy program, the political considerations fall under tactics rather than strategy—and as our models demonstrate, political feasibility is endogenous.

Country Specificities

The range of countries that can now be classed as “in transformation” is vast: Hungary, the Czech Republic, Poland, and Slovakia (the “Central European” or “Visegrád” four); with Bulgaria and Romania, we have the “core six”; adding Slovenia and the Baltic states, we arrive at the likely full list for European Union association agreements (the “Europe agreements”)—the medium-run candidates for accession to the Union; Albania and the rest of the former Yugoslavia complete “Eastern Europe”; farther east, we have Russia, Ukraine, the rest of the former Soviet Union, Mongolia, China, and Viet Nam. Most of my observations will be drawn from the experience of the “core six,” but we want the conclusions to have more general relevance. Is that possible?

Comparisons are undermined by differences in initial conditions. Time series are still short, and we cannot rerun history, so it is especially tempting to compare the experiences of different countries. But they (their policymakers) all think they are quite specific and cannot learn much from each other, much less from countries not undergoing this particular form of transformation.

There are indeed significant differences in initial conditions (see Islam and Mandelbaum, 1993): foreign debt, the history of workers’ power in enterprise management, industrial history, agricultural institutions, the sectoral structure of output and employment, macroeconomic imbalance, natural resources, trade structure, political culture and institutions, previous duration of communism and central planning, previous “reform” efforts, and doubtless much more. There are equally significant differences in the policies these countries have followed.

Politicians and policymakers understandably stress the specificity of their own countries—their circumstances, needs, and achievements. Every country is of course different. But economics aspires to generality, and noting specific successes and failures here and there is little help unless we can interpret what they imply. It is unhelpful to select only the comparisons and generalizations that suit one’s prejudices: for example, those who do see important lessons from the Chinese case for Eastern Europe often reject any analogies with Latin American countries, whereas those relying on the general features of stabilization in Latin America find such strongly specific characteristics in China as to make it irrelevant for Eastern Europe.

Our models and methods help us to make allowances for the clearly relevant differences, and cross-country comparisons are our only hope of constructing counterfactuals. In any case, it is certainly not my purpose here to evaluate various countries’ policies and their relative success—that is hard enough even when the initial conditions are much more similar than they were in Eastern Europe in 1989—but rather to look for generalizations (a representative sample of comparisons and generalizations in the literature might include European Commission, 1991; Portes, 1993; Blanchard and others, 1994; and European Bank for Reconstruction and Development, 1994).

The Importance of Rents

In a fundamental sense, transformation is all about economic rents and income distribution. That is in good part why it is transformation, rather than a smooth transition. Distortions create rents, which in turn create rigidities, resistance to change. In these conditions, economic agents do not so much seek rents but rather seek to protect those rents they have. That is a major source of inertia. It is hard to identify the rents; but if one could, and then proceed to eliminate them all, the government would fall immediately. And such a wrenching redistribution would doubtless be highly inequitable, too.

More concretely, some claim that the initial capital stock in Eastern Europe, both physical and human, was almost entirely unusable or wrongly allocated but not transferable. Experience has shown that this is false. At the same time, however, clear losers are emerging, such as the many who are forced to retire prematurely or simply to withdraw from the labor force when their jobs disappear, because they cannot be moved or retrained. This, in turn, gives some justification, not merely on political but also on equity grounds, for not trying to reduce pension entitlements too drastically even though they are an immense and growing burden on state budgets. The older generation suffered longer under communism, may have lost their savings in a burst of “corrective” inflation, and will have less time to enjoy the fruits of a protracted transformation.

Initial Successes

Considerable progress has been achieved, even in Russia. Farther west, the relative stability of transformation-oriented economic policies through elections and major political swings testifies to credible regime changes in several countries besides the Central European four. Liberalization has been a considerable success, though some of its distributional consequences have been drastic and unfortunate. Price adjustment has been remarkably rapid, except for the remaining controlled prices; in general, goods prices are more flexible than we might have thought and are more likely to adjust to international relative prices quickly. This may be partly a result of the comprehensiveness of the liberalization and the concomitant substantial increases in the overall price level. Stabilization has also succeeded in most of the cases in which it was necessary and feasible, following fairly similar “heterodox” approaches with multiple anchors.

Opening up to market-oriented trade, with relatively liberal trade policies, has brought a successful reorientation of trade flows to the West, especially the European Union. Current account convertibility has proved sustainable where it has been introduced. And correspondingly, policymakers in these countries have been able to conduct a constructive dialogue with the Bretton Woods institutions. There is more good news, including substantial institutional change—but that is not what this paper is about.

Initial Errors

Why were the transformation programs not more successful, more quickly? Two reasons stand out: some of their architects thought it would be easy; and the specificities of these economies were not fully understood, nor their implications appreciated, even by domestic policymakers. Here, I mean not so much the country-specific differences in initial conditions, but rather the differences between economic transformation and either postwar reconstruction or the experience of developing countries.

What were the consequences? First, all forecasts and projections were way off the mark: they underestimated the size of the initial price level shock (“corrective” inflation), the extent of the fall in output and its persistence, and the initial improvement but subsequent deterioration of the fiscal balance and the current account; and they overestimated the amount of foreign investment and aid and, overall, the speed of the transformation process (especially privatization). From precise estimates to broad assessments, almost all these projections were quite overoptimistic. The major exception to that generalization is perhaps the general political stability and progressive character of political development, which are superior to expectations and remarkably restrained in the light of the economic results. Yet the errors were considerable and had political consequences, as shown by the recent Polish and Hungarian elections. Nor can we associate these mistaken expectations with variables overshooting their long-run equilibrium values, whatever those might have been—these were just bad judgments.

The policy errors were closely related to the forecast errors and their causes. Evidently—and fortunately—not all can be observed in any given country, but several appear frequently. Again, they are not features of policies imposed from the outside, by the IMF or anyone else; in most cases, whatever the sources of the arguments that influenced them, the domestic policymakers themselves did want these policies.

Here, it is most convenient just to list the major policy errors briefly before elaborating on them:

• Programs of restitution of physical property—both real estate and plant and equipment—to pre-nationalization owners, which created great uncertainty and discouraged investment;

• Overemphasis on macroeconomic relative to microeconomic policies;

• Excessively tight monetary policies (overestimating the importance of the “monetary overhang” and having to control the exaggerated price shock caused by excessive devaluation);

• Excessive devaluation and inadequately specified exchange rate policy;

• Misunderstanding of the capacities and behavior of the state-owned enterprises (and many who thought these enterprises would collapse quickly also thought the private sector could easily step in to compensate for their absence);

• Overly complex and ambitious plans to privatize state-owned enterprises, which lacked incentives for stakeholders;

• Sequencing errors—especially the delay in dealing with undercapitalized and technically backward banks and the overhang of bad debt of state-owned enterprises, as well as the initial overemphasis on the potential role of stock markets;

• Deliberate early dissolution of the Council for Mutual Economic Assistance (CMEA), which contributed to the dramatic fall in inter-regional trade;

• Overly ambitious and hasty elimination of tariff protection and rejection of any kind of “industrial policies”; and

• Inadequate emphasis on debt reduction (except Poland) for those countries that required it, and excessive delay in all cases—an error committed by not only domestic policymakers (as in Hungary) but also the international financial institutions and the banks.

Macroeconomic Policies

We could say monetary policy was too tight here, fiscal policy too loose there, and go into further detail. The main error, however, was simply the overemphasis on macroeconomic policy itself. This is not merely because some countries did not need to stabilize or would not have done so if they had not devalued excessively (for example, Czechoslovakia). The preoccupation with stabilization may be partly a reflection of the leading role of the IMF and its macroeconomic conditionality. This was reinforced by advisors who believed in the endemic macroeconomic disequilibrium of socialist economies and saw hyperinflation around every corner. It was accepted voluntarily by politicians who needed external support and in any case often found IMF policies wholly convincing.

I believe this was a serious strategic error. It was right for the Fund to focus on short-run macroeconomic policies, internal and external balance—that is its mission, if not its deformation professionnelle—but it was wrong to make all other aid conditional on agreement to a stabilization program with the Fund and thereby to put its priorities at the top of policymakers’ agendas. The Fund could not devote much effort to the microfoundations of the macro policies; and until lately, its performance criteria were put entirely in terms of macroeconomic indicators (while the recently imposed conditions on structural reforms appear not to have been enforced—see Gomulka, forthcoming).

The initial conditions described above should have indicated, however, that mere liberalization would not create the microeconomic, institutional basis for the macroeconomic policies appropriate to a market economy. Policymakers have limited time, administrative resources, and political capital. For too long, they devoted an excessive share to only one dimension of the transformation process, and too little to anticipating and dealing with the many other obstacles that would impede it.

That might nevertheless have been justified if macroeconomic stabilization were a necessary condition for microeconomic, structural reforms and for opening the economy to trade. There is ample evidence, however, that this is not so. Recent work on Latin America suggests that this aspect of the conventional wisdom on sequencing has been ignored by several countries, with considerable success (Edwards, 1994). Similarly, in Eastern Europe itself, it has proved possible to implement substantial structural change in high-inflation environments.

Of course it would be better still if there was stabilization, and there may be no medium-run trade-off between stabilization and either output or the speed of transformation. The issue then is feasibility. If immediate stabilization is not feasible (see the “war of attrition” models), there may still be very useful things that can be done, which, by constructing parts of the microfoundations, may also make stabilization that much easier when it does come onto the political agenda.

Fiscal Policy

The initial fiscal deficits in Eastern Europe were not huge by Western European standards, but they were much more difficult to finance in a noninflationary manner. It was relatively easy to cut expenditure in the short run, mainly because price liberalization automatically eliminates large subsidies. Revenue then drops, partly because of recession, partly because of the difficulty of taxing private firms. In the medium run, the basic problem is expenditure—subsidies fall, but social spending rises, especially pensions (early retirement) and unemployment benefits. Unless fast growth finally arrives, the long-run problem will be servicing domestic debt, with large deficits and growth rates below real interest rates.

External Policy

Here, the mistake was not early convertibility or fixed nominal exchange rates, where these were implemented. The big errors were over-devaluation and insufficient precision and clarity regarding the exchange rate regime. The danger of serious overshooting with the initial devaluation arises because there is no reliable indicator of an equilibrium rate, so policymakers take too seriously the preliberalization “free market” rate, which is always deeply undervalued. And even the “equilibrium” rate suggested by purely monetary considerations is likely to be significantly undervalued. The fall in exports to CMEA partners releases resources for export elsewhere, the fall in output reduces the demand for imports, the rationalization of distortions raises the efficiency of trade, and the required expansion of the nontraded services sector should be led by an increase in the relative price of services—hence, a real exchange rate appreciation.

The macroeconomic costs of serious overshooting are an excessive initial price shock, requiring, in turn, excessive monetary stringency to stop it from setting off rapid inflation at the outset and, hence, a strong negative demand shock, coupled with a negative supply shock to import-dependent firms; and, in due course, because the initial real wage is unsustainably low and domestic firms still have significant market power, a catch-up, cost-push inflationary process with real appreciation is required. In Czechoslovakia, the excessive initial devaluation may have actually fostered a shift to lower value-added production, inimical to competitiveness in the medium run.

The trade-off between a fixed and a crawling peg is well understood: a nominal anchor for monetary stability versus a real exchange rate target to maintain competitiveness. Initially the needs of stabilization are paramount and accentuated by the role of a fixed nominal rate in anchoring the price structure, which is shifting radically. Subsequent real appreciation is inevitable and desirable, however, so it is best to announce at the outset that the crawl will begin when the authorities judge that the real rate is in an appropriate longer-run range, which itself is likely to shift gradually.

Current account convertibility frequently has come early in the sequencing of the transformation process—and rightly, at least for the smaller economies with relatively high trade participation and some initial access to international reserves. It is more important in the formerly planned economies than elsewhere because it can play such a key role in creating markets and helping them to function properly. It eliminates bureaucratic influence on the allocation of foreign exchange. And early convertibility, if it can be sustained, is popular and highly visible, so it greatly enhances the credibility of policymakers. Payments union proposals (and their surrogates, like the “Interstate Bank”) were never an appropriate substitute for convertibility, either for Central Europe or the former Soviet Union.

These countries should not, however, suddenly open up their capital accounts, and in general they have not. There are dangers in uncontrolled capital inflows, as well as outflows. Capital account inconvertibility need not discourage foreign direct investment as long as current account convertibility guarantees the freedom to remit profits. Opening up successfully to capital account transactions requires positive real interest rates, a realistic exchange rate, and some depth in domestic financial markets.

Although the postwar European experience was not fully relevant for the Central and Eastern European countries, the Marshall Plan did demonstrate the importance of external assistance and debt consolidation in economic transformation and recovery. Several Eastern European countries, including Poland, Hungary, Bulgaria, and Russia, had serious external debt problems at the outset of the transformation. Given these debt burdens, it has been regrettable that the major western countries’ severe budgetary constraints have made large-scale grants unrealistic; debt relief, which might have been an adequate substitute, has been quite insufficient.

Poland simply did not pay and eventually—in a deal finalized almost five years after the revolutions of 1989—achieved substantial debt reduction. Bulgaria, too, has finally reached a deal; it has meanwhile suffered considerably from the constraints on its external transactions. Hungary presents the most interesting case: walking into the trap, springing it, and struggling valiantly without (yet) crying out.

The initial Hungarian story was that exports were growing so fast that there was no difficulty in financing debt service. This was true until 1993, but only for external finance. The domestic fiscal burden of debt service is crushing, and investment is severely depressed by the impact on financial markets. Fiscal reforms that would imply austerity (in pensions, etc.) are much more difficult, given the austerity already imposed by debt service: debt relief is complementary to, not substitutable for, the “tough policies” needed to put the Hungarian fiscal house in order.

The denouement was foreseeable, and it is partly the fault of the West. The Fund claims unconvincingly that the high proportion of Hungarian debt in bonds would limit the benefit to Hungary of debt reduction relative to the cost of impaired capital market access. But debt consolidation can take many forms, and history suggests a solution could be found if Hungary wanted it and the international institutions offered help rather than opposition. If Mexico deserved and could benefit greatly from debt reduction—and gained rather than lost capital inflows—so also could Hungary. Perhaps policymakers inside and outside Hungary will learn something from the recent announcement that J.P. Morgan expects to receive the “highly prized” mandate for Poland’s forthcoming $500 million Euromarket issue and to offer significantly finer terms than Hungary could get (Financial Times, December 1, 1994). So much for the “elephantine memory” of capital markets and the rewards for “good behavior”!

The external constraint is likely to bite soon and may prematurely inhibit expansion in those economies where recovery is finally beginning. The initial trade balance successes were temporary, the result of special factors: abnormally depressed domestic demand, temporary undervaluation, excess stocks to liquidate, the opening of European Union markets, and the need of state-owned enterprises to maintain cash flows in the face of monetary tightness. Trade balances already deteriorated in 1993, on both exports and imports (except for the Czech Republic and Russia). The good outcome assumed in the early writing on transformation would involve much larger foreign capital inflows to support the imports of investment goods that these countries need. So far, however, foreign direct investment has been disappointingly low. Portfolio capital inflows may pose (familiar) problems in a few countries, including undesired upward pressure on the exchange rate; and in more normal circumstances, Russia’s natural resources might present it with a Dutch disease. These problems of success would be welcome.

Household Behavior

Here, there were no mistaken perceptions: households in these economies have behaved rationally and adapted quickly. Moreover, some problems that were expected have not been significant. Household saving has in most countries held up well, as precautionary motives appear to have dominated life-cycle considerations (even there, while permanent incomes should have risen, the perceived need to provide privately for retirement might also have risen). Migration was a potential vicious circle, insofar as departure of the best and brightest would reduce the attractiveness of investment and, hence, reduce the growth of output and real incomes and stimulate further outflows. For those trying to get the West to increase aid to the East, this specter was a useful lever; but so far, the flows have not been large.

The serious problem arising from household behavior is widespread and strong consumer preference for imports, with little regard to the price differential. That is due partly to long-standing conditioning and the attractions of the previously unattainable, but it is proving remarkably persistent.

State-Owned Enterprises

The role and prospects of these enterprises were underestimated—often with colorful images of dinosaurs producing obsolete goods for the region’s bloated industrial sectors, in particular their rapidly contracting defense establishment. There was no coherent policy toward them, except to privatize as rapidly as possible or simply to make them fold by cutting off finance, so little thought went into trying consciously to modify their behavior before privatization. The result was “state desertion,” sometimes going further to involve explicit tax and other discrimination against state-owned enterprises vis-à-vis private firms.

Recent research suggests that many of the state-owned enterprises can in fact adapt, at least enough to warrant reconsidering policies that explicitly seek to starve them quickly. For example, after an initial lag, they are in most countries reducing employment in advance of privatization, so productivity is beginning to recover. The key is to stop open-ended subsidies and to create incentives with the prospect of ultimate privatization that will not necessarily involve sacking all the managers. Fiscal pressures can actually be useful here, in convincing managers that there is no money so they cannot expect any. Then they must discover that they cannot get it from the banks (see below).

Very severe financial tightening at the aggregate level can have mixed effects through state-owned enterprises on the development of the private sector. The limited credit available tends to go to these enterprises, but it is not enough, and they respond by releasing both labor and capital stock to the private sector. Nevertheless, an excessively severe demand shock weakens the ability of these enterprises to respond, and the shortage of credit hampers restructuring. The pressure can go too far, as it demonstrably has in many cases.

Continuous pressure must be maintained, however, perhaps for a long time to come, because in most countries comprehensive privatization of state-owned enterprises is still not in prospect, and the examples of rapid privatization in Czechoslovakia and Russia are at best problematic. It is much too early to evaluate the results of the Czech voucher privatization scheme, except to applaud the efficiency and determination with which it was executed. It does appear, however, to have created an extraordinary system of corporate governance. The firms are owned by the investment privatization companies to which most individuals entrusted their vouchers. Most of these are managed by the large banks, of which the state still owns (on average) 40 percent, and to which the former state-owned enterprises are heavily indebted. The banks are naturally reticent to call in their loans, even when they look bad, and the authorities have pursued a conscious anti-bankruptcy policy through the activities of the National Property Fund and the Consolidation Bank. The Russian case is simpler, in that in most firms control seems to have gone to insiders; but, again, it is too soon to judge.

Thus the delay in privatization of large state-owned enterprises may not have been as unfortunate as most of us thought, or at least not disastrous, given the constraints and the alternatives. There is still time to do some of the restructuring necessary to make privatization successful.

Financial Intermediation

It now seems to be commonly accepted that the single most important error in sequencing, at least in the core six countries, was not to have implemented urgently a financial cleanout: recapitalizing the banks, canceling debts of state-owned enterprises at the time of privatization, and instituting cash-limited fiscal subsidies to them meanwhile, so that the banks would not have to extend them new loans or capitalize their arrears. Capitalism cannot function without capital markets; stock markets were and are clearly not going to be a major mechanism for financial intermediation or corporate control for many years to come. The banks had to take on the task, and they were too weak to do so. Investment and the growth of new private firms have thereby suffered. Both monetary tightness and the need to maintain high bank intermediation margins to build up a capital base, in the absence of recapitalization, have contributed to high real interest rates, and thus to low investment.

This extremely serious problem was avoidable, but not widely foreseen or understood. The exercise could have been done quickly and comprehensively, and hence better than the partial and repeated efforts that have followed in several countries and that have created considerable moral hazard. It need not have impaired the credibility of the authorities—suggesting they were “soft” financially—if the debt cancellation had been tied explicitly to the one-off event of privatization. The fiscal impact, looking at the accounts of the consolidated public sector, is zero. As long as both banks and firms are state-owned enterprises, the public sector borrowing requirement remains unchanged. For some time, however, the IMF cautioned against the supposed effect on the deficit, and even now one sees warnings against the “high fiscal cost” in World Bank publications. This is nonsense, as the Fund is now willing to state publicly.

Alone among Eastern European countries, Hungary sprang the bankruptcy trap. Any firm unable to meet payments to any creditor had to file for bankruptcy, and a financial restructuring required unanimous consent by creditors. The system could not cope with the resulting 17,000 cases filed for financial restructuring or liquidation in 1992 and early 1993. Those who have been calling for vigorous enforcement of bankruptcy to force true restructuring should examine the Hungarian case and tell us whether any positive longer-run consequences will outweigh the negative short-run effects. In contrast, the Czech Republic has assiduously sought to avoid bankruptcies and the consequent restructuring, at least as long as voucher privatization was still incomplete. There, one can only conjecture how the complex, financially unsound structure of indebtedness and corporate control will unwind.

Labor Markets

In view of the lag in adjustment, unemployment will not control inflation in the early phase of the transformation, so incomes policies are necessary. In the macroeconomics of central planning, incomes policies played the role of nonexistent monetary policies, and until monetary policy can develop its full role as a macroeconomic policy instrument, it will need that help. Most of the initial stabilization programs satisfied this requirement with some version of the “tax on wage increases” first introduced in Hungary in the late 1960s. The inefficiencies created by such a tax may not be too damaging, since there is a case for maintaining a constraint on relative wages (an effect of most versions of this tax), and a perception that wage-setting decisions in state-owned enterprises may be interpreted simply as allocating rents.

Despite real wage moderation, unemployment began to rise. Contrary to some conjectures, the registered unemployed in Eastern Europe really are unemployed, and there are more, according to labor force surveys. Benefits rules are no longer generous. In most countries, there are small inflows into the pool of unemployed, but very small outflows. Hence the long-run problem: the creation of a pool of long-term unemployed, having no moderating influence on wage pressure but constituting a growing fiscal burden, which, in turn, brings restrictive measures that tend to increase unemployment. Starting from extremely high levels, participation rates have already fallen below the norms for industrial countries; that is a social as well as a fiscal burden.

What can be done to promote job creation? Not much in the state-owned enterprises or newly privatized larger firms, and after the first wave, job creation in small and medium-sized enterprises will require investment and, hence, functioning capital markets. Active labor market policies have succeeded in the Czech Republic, and public works offer a more direct alternative: public investment has been abnormally depressed, while infrastructure and public services (health, education, etc.) have deteriorated drastically, bringing a significant perceived decline in welfare. The fiscal burden of paying wages rather than unemployment or other social benefits may not be so great.

Trade and Industrial Policies

The first trap here opened with the deliberate destruction of the CMEA (Csaba, 1994, p. 15, refers to “their [the Visegrád countries’] thrust to get rid of Comecon from late 1989 onwards”) and the disintegration of the former Soviet Union. Eastern Europe compensated, however, by raising exports to the West, mainly the European Union, and especially Germany before its recession took hold.

The next step was to negotiate agreements with the European Union on a bilateral basis. The resulting hub-and-spoke pattern of trade relations marginalizes the individual Eastern countries, artificially discourages economic relations among them, and favors investment in Germany or Austria to supply the East rather than in the East itself. Meanwhile, accession to membership in the Union is likely to take much longer than hopeful early proposals (including my own) had suggested: these countries are too poor, too agricultural, and too populous for the European Union budget to accommodate under anything like present rules. Yet the economic advantages to the Union of opening up to the East are considerable (Faini and Portes, 1995). And if these countries do not have clear prospects for accession to the Union, there is the danger of another vicious circle of low investment, slow growth, and indefinitely receding prospects of membership.

All this argues for mapping out a clearer path to accession that would multilateralize the Europe agreements to create a comprehensive free trade area, then deepen integration in an approach to something like the planned “European Economic Area” comprising the European Union and the European Free Trade Association (Baldwin, 1994). This would be attractive to foreign investors, would promote institutional transformation, and would help to raise intraregional trade further toward the potential levels that gravity models suggest.

Two other policy orientations in this domain testify to exaggerated rejection of governmental action: the rush not only to discard import quotas and other nontariff barriers but also to lower tariffs to negligible levels; and the explicit avoidance of any policies that might possibly be labeled “industrial policies,” no matter how market-friendly, nonselective, nondiscretionary, and transparent they might be. A moderate tariff can yield useful revenue while temporarily protecting “senile” industries, as the demand-shift explanation of output decline would suggest is desirable. The initial overshooting weakened these countries’ negotiating position vis-à-vis the European Union, and it led in any case to irresistible pressures to re-establish various forms of protection.

Nonselective export promotion services and partial state guarantees for credits to small and medium-sized enterprises are examples of industrial policies that might be helpful and not lend themselves to abuse. Anything that helps to create markets or to deal with market failures systematically is surely to be welcomed. The financial cleanout and active labor market policies recommended above are equally examples of such structural policies. It is economic illiteracy or misplaced ideological purity to equate microeconomic policies with detailed state intervention in economic decision making.