China, Botswana, Mauritius, and Australia ¿ four cases of success in our sample ¿ all owe their performance to the presence (or creation) of institutions that have generated market-oriented incentives, protected the property rights of current and future investors, and enabled social and political stability.

Consider the case of Botswana, presented by Daron Acemoglu, Simon Johnson, and James Robinson. Per capita income in Botswana grew at 7.7 percent annually between 1965 and 1998. The proximate reasons for this outcome are easy to list. Law and order were maintained. Diamond revenues were managed exceptionally well. The bureaucracy was efficient and run along meritocratic lines. Hard budget constraints were the rule (and not the exception) in the public sector. There were large public investments in education, health, and infrastructure. The exchange rate was set at a competitive level. However, policies were not uniformly "good" in the conventional, Washington Consensus sense of that word. The government in Botswana has intervened massively in the economy, and the public sector accounts for a much larger share of the economy than is true on average in Africa. The key to Botswana, Acemoglu, Johnson, and Robinson argue, is that institutional arrangements have adequately protected the property rights of actual and potential investors. The authors provide a rich, textured account of the political and historical roots of these arrangements.

In the absence of good public institutions, growth has been difficult to achieve on a sustained basis. And when growth has taken place, it has either proved fragile (as in post-1997 Indonesia) or incapable of delivering high levels of social outcomes in areas such as health, education, or gender equality (as in Pakistan). In his chapter on Indonesia, Jonathan Temple describes the Indonesian implosion of 1997 as a case of outgrowing existing, weak institutions. Pakistan's failures in social development, despite respectable growth until very recently, are documented in painstaking detail in William Easterly's chapter. Easterly attributes this failure to the "roving bandit" syndrome (Olson 2000): State institutions dominated by a highly fragmented set of military and landed elites have had little incentive to produce public goods and therefore have not done so.

State institutions are not the only ones that matter. Social arrangements can have equally important and lasting consequences for economic growth. Gregory Clark and Susan Wolcott's discussion of Indian economic history illustrates this. They argue that India's backwardness is due in large part to the inability to employ technology, and not to an inadequate diffusion of technology per se. Their evidence from the textile industry shows that while identical machines were used in India and in Britain, these machines were operated much less profitably in the former. The problem in India is neither allocative inefficiency nor inadequate technology; the problem is low technical efficiency despite access to state-of-the-art technology. The authors speculate that the answer lies in the nature of the employment relationship and its variation across societies. In productive economies, workers exert more effort in the workplace than can be justified purely by monitoring or by direct financial incentives because they expect everyone else to act in that manner. India, the authors argue, is characterized by a mutual-shirking equilibrium, rather than a mutual gift-giving equilibrium. In this view, India's poverty is largely unconnected to government policy or public institutions.

Trade -or, more specifically, government policy toward trade- does not play nearly as important a role as the institutional setting

All of the successful countries in our sample have benefited from trade and foreign investment. But as the narratives make clear, specific public policies that are directed at international economic integration or disintegration do not correlate very well with economic performance once one looks at the evidence carefully.

Take Australia, for example. Australia's relative decline vis-à-vis the United States or other rich countries is often attributed to the country's inward-looking policies. But as Ian McLean and Alan Taylor note, there is a timing problem in asserting this claim. While the Australian government sharply changed its policies towards integration in the first three decades of the 20th century (imposing higher tariffs, import licensing, and a stop to Asian immigration), Australia's relative decline compared to the United States and California took place largely before this change in "growth strategy."

Mauritius provides another illustration. According to Arvind Subramanian and Devesh Roy, the level of trade protection in Mauritius has long been among the highest even within sub-Saharan Africa and has come down appreciably only in the late 1990s ¿ more than two decades after the onset of high economic growth. India was able to double its growth rate in the 1980s prior to the liberalization of its highly restrictive trade regime, which came a decade later (see below). Yingyi Qian argues that the impact of China's growing openness to trade and direct foreign investment came mainly through domestic institutional changes.

Geography is not destiny

Consider Australia and Mauritius again. As McLean and Taylor stress, Australia is the only rich OECD economy that contains large areas of tropical land. Much of Australia is desert or arid, with low and highly variable rainfall. Soil quality is poor. Mauritius is a tropical country, with a high degree of dependence on an export commodity buffeted by terms-ofe shocks. Botswana, which has the added disadvantage of being landlocked, has obviously not suffered greatly from being geographically disadvantaged either. Botswana and Mauritius both started out with extremely poor initial conditions. Good institutions, it appears, can overcome geographical constraints and lousy initial conditions.

Good institutions can be acquired, but doing so often requires experimentation, willingness to depart from orthodoxy, and attention to local conditions

The narratives in this volume go beyond simply asserting that "institutions matter." Indeed, one advantage of case studies is that they can provide a richer account of where good institutions come from, the shape they take, and how they need to evolve to support long-term growth.

In Botswana's case, Acemoglu, Johnson, and Robinson speculate that the roots of Botswana's unusually good institutions lie in a combination of factors: tribal institutions that encouraged participation and imposed constraints on elite behavior; the limited effect of British colonization on these tribal institutions, as the colonizers had little interest in Botswana until relatively late; the relative power of rural interests, which created an overlap between Botswana's area of comparative advantage and the economic interests of the elites; and last but not least, the wise and foresighted leadership exhibited by post-independence political leaders. The final element in this list reminds us not to be too deterministic about the source of high-quality institutions. Choices made by political leaders make a big difference.

Perhaps nowhere has this been clearer than in China. Qian's discussion of China focuses on what he calls "transitional institutions" ¿ institutions that can differ greatly from off-the-shelf, "best practice" institutions that are often the object of institutional reform in the developing world. Transitional institutions can have the virtue of being more suited to the realities on the ground on both economic efficiency and political feasibility grounds. Qian shows that the Chinese leadership experimented and purposefully crafted imperfect, but feasible institutional arrangements. He discusses four specific examples: dual-track reform, which liberalized prices at the margin while maintaining the "plan track" in place; township and village enterprises, which represented an intermediate form of ownership between private and state ownership; Chinese-style federalism, which left the regions with significant autonomy and created healthy economic competition among them; and anonymous banking, which allowed financial development while restraining the capacity of the state to expropriate large depositors. These "transitional institutions" succeeded because of their high ratio of economic benefits to political costs. They improved economic incentives without requiring a significant redistribution of income, large-scale (and risky) institutional reforms, or the expenditure of large amounts of political capital.

The Chinese example demonstrates that successful institutions often have heterodox elements. This is a lesson that comes across also from the narratives on Botswana and Mauritius. As noted before, Botswana mixed up market-friendly institutions with heavy state intervention and a large public sector. Mauritius combined its outward export-processing zone with centralized wage bargaining and (for a developing society) an unusually generous welfare state.

The country narratives suggest that "good" institutions ¿ in the sense of institutions that promote and sustain growth ¿ must often have elements that are highly specific to a country's circumstances. An approach to institutional reform that ignores the role of local variation and institutional innovation is at best inadequate, and at worse harmful. China, Mauritius and Botswana are examples of countries that have done very well over extended periods of time with a heterodox mix of institutional arrangements. In effect, these countries have combined orthodox elements with local heresies. As some of the other cases discussed in this volume demonstrate, property rights, sound money, and open trade in themselves do not always do the trick. For example, Clark and Wolcott note that pre-independence (1873-1947) India's relative performance lagged despite institutional arrangements that would be regarded as ideal by many economists: secure property rights, free trade, open capital markets, and social and political stability. In his comparative analysis of Vietnam and Philippines, Lant Pritchett points to the paradox that the country whose policies and institutions best fit today's conventional wisdom (Philippines) is doing poorly, while the one with divergent institutions (Vietnam) does very well.

The experience of former socialist economies, discussed by Georges de Menil, further reinforces the role of local context. The three countries closest to Western Europe (Poland Hungary, and the Czech Republic) have done very well. What seems to have been key for these countries, as de Menil emphasizes, is their relationship with the European Union (EU). The EU provided a plausible institutional model for these countries, in view of a common historical heritage and relatively short experience under Communism. Furthermore, this model was backed up with the carrot of eventual accession to the EU. Consequently, structural reform was effective and took hold relatively quickly in Poland, Hungary, and the Czech Republic. For countries further to the east, this same type of institutional reform proved to have worse "fit" and less political feasibility. Hence the finding that distance from Düsseldorf and the number of years under Communism together are the best predictors of a transition economy's relative economic performance.

The narratives on Mexico and Bolivia complement these macro level analyses by providing more specific detail on how institutional arrangements matter to economic performance. Maite Careaga and Barry Weingast focus on fiscal federalism in Mexico. Their key point is that good institutions are those that provide public officials with the incentives to provide market-fostering public goods at least cost in terms of corruption and rent seeking. Thinking in such terms helps endogenize the concept of "good governance." The Mexican history with federalism provides a rich laboratory for studying the consequences of changes in legal provisions with respect to revenue sharing. Careaga and Weingast argue that greater dependence on locally generated revenues and greater electoral competition increase the provision of market-fostering public goods. They present evidence that is consistent with these expectations.

Bolivia has undertaken extensive macroeconomic reform, liberalization, and privatization since 1985. Yet economic performance has remained lackluster. Daniel Kaufmann, Massimo Mastruzzi, and Diego Zavaleta attempt to sort out the institutional reasons for this failure. Their main story is that the reform agenda has not been appropriately targeted on the most glaring trouble spots on the institutional front. Relying on a worldwide enterprise data set for benchmarking, they document the large variance in institutional quality that exists within Bolivia, with institutions relating to macroeconomic stability generally perceived as working much better than those relating to the rule of law. The authors identify petty corruption, uncertain property rights, and inadequate courts as the source of problems. Enterprises react to these by withholding investments and taking shelter in the official economy. An important virtue of the data set and approach taken in this chapter is that the authors are able to unpack "institutional quality" and show how aggregate indices or country averages can be misleading. The clear implication of the Bolivia story is that institutional and governance shortcomings vary across national contexts, and that institutional reform agendas have to focus on the constraints that happen to bind the most locally.

The onset of economic growth does not require deep and extensive institutional reform 

This is one of the most important (and encouraging) lessons that emerge from the country narratives. It is also a lesson sharply at variance with conventional wisdom on institutional reforms, which holds that their complementary nature requires a long list of such reforms to be pursued simultaneously.

To appreciate the logic of the conventional wisdom, here is a thought experiment. Imagine a Western economist had been invited in 1978 to give advice on reform strategy to the Chinese leadership. How would she formulate her advice, in light of what we think we know today? Being a sensible economist, she would presumably know that the place to start would be agriculture, as the vast majority of the Chinese population lives in the countryside. Liberalization of crop prices would be the number one item on the agenda. Cognizant that price incentives make little difference when farm incomes accrue to communes, she would immediately add that privatization of land must accompany price liberalization. Reminded that the obligatory delivery of crops to the state at controlled prices is an important implicit source of taxation, she would then add that tax reform is also required to make up for the loss in fiscal revenues. But another problem then arises: if the state cannot deliver food crops to urban areas at below-market prices, will urban workers not demand higher wages? Yes, that requires some reforms too. State enterprises need to be corporatized so they can set their wages and make hiring and firing decisions freely. (Privatization would be even better, of course.) But if state enterprises now have autonomy, will they not act as monopolies? Well, antitrust regulation, or trade liberalization as a shortcut, can take care of that problem. Who will provide finance to state enterprises as they try to restructure? Clearly, financial market reform is needed as well. What about the workers who get laid off from the state enterprises? Yes, that's why safety nets are an important component of any structural adjustment program. And so on.

The logic of the recommendations is impeccable, even if their practicality is questionable. The recipients of such advice would be excused if they reached the conclusion that this reform business is too hard to accomplish in one's own lifetime. Luckily, actual experience with successful reform provides a different lesson: an ambitious agenda of complementary institutional reforms is not needed to kick-start growth. As we know with hindsight, the Chinese reformers were able to take imaginative shortcuts that sidestepped the complementarities that might have otherwise ruined a partial and gradual approach. Dual-track price reform and the introduction of the household responsibility system enhanced agricultural production incentives at the margin without requiring ownership reform, undercutting fiscal revenues, and upsetting the social balance in urban areas. As Qian makes clear in his narrative, this may not have been an ideal reform by textbook standards, but it worked.

Is China a special case? Let's look at the world's next most populous country, India, which has recently managed to roughly double its rate of economic growth. How much reform did it take for India to leave behind its "Hindu rate of growth'" of 3 percent a year? J. Bradford DeLong shows that the conventional account of India, which emphasizes the liberalizing reforms of the early 1990s as the turning point, is wrong in many ways. He documents that growth took off not in the 1990s, but in the 1980s. What seems to have set off growth were some relatively minor reforms. Under Rajiv Gandhi, the government made some tentative moves to encourage capital-goods imports, relax industrial regulations, and rationalize the tax system. The consequence was an economic boom incommensurate with the modesty of the reforms. Furthermore, DeLong's back-of-the-envelope calculations suggest that the significantly more ambitious reforms of the 1990s actually had a smaller impact on India's long-run growth path. DeLong speculates that the change in official attitudes in the 1980s, towards encouraging rather than discouraging entrepreneurial activities and integration into the world economy, and a belief that the rules of the economic game had changed for good, may have had a bigger impact on growth than any specific policy reforms.

In short, the experiences of the world's two largest developing economies indicate that modest changes in institutional arrangements and in official attitudes towards the economy can produce huge growth payoffs. Deep and extensive institutional reform is not a prerequisite for a takeoff in growth. That is the good news. The bad news is that the required changes can be highly specific to the context. The "transitional institutions" of India and China, to use Qian's term, look very different. And for a good reason: the binding constraints on growth differed in the two countries. The mark of a successful reform is its ability to concentrate effort on the binding constraints.

Sustaining high growth in the face of adverse circumstances requires ever-stronger institutions 

India and China are both very low-income countries. So is Vietnam, which has been growing quite rapidly under a Chinese-style strategy that defies conventional wisdom on institutional reform. Pritchett, who analyzes the Vietnamese record and compares it to the Philippines', suggests that countries that are in the process of escaping from low-level poverty traps may be fundamentally different from middle-income countries. The policies required to initiate a transition from a low-income equilibrium to a state of rapid growth may be qualitatively different from those required to re-ignite growth for a middle-income country. At low levels of income, with reasonable institutions and reasonable policies, it may be easy to achieve high growth up to semi-industrialization. But the institutional requirements of re-igniting growth in a middle-income country can be significantly more demanding. Pritchett notes that per capita GDP in the Philippines remains lower than its level in 1982, even though institutional quality (with the transition to democracy after 1982) has increased significantly. Pritchett speculates that the trouble may be that uncertainty about the rules of the game has increased. In his words, "what trips countries up is the transition from one set of 'institutions' to another." The uncertainty over the rules of the game that accompanies comprehensive, but poorly managed, institutional change is a fundamental roadblock to sustained economic growth.

Indonesia provides an apt illustration of the dangers of letting institutional reform lag behind growth. Jonathan Temple describes Indonesia as a case of "growing into trouble." In his view, growth was not accompanied with the good fundamentals that would have provided the economy with the resilience to handle adverse shocks. Indonesia's economic performance since the mid-1960s has been facilitated by three fortuitous circumstances: oil, the green revolution, and high-growth neighbours. But rapid growth, Temple argues, made institutional weakness a great liability. Indonesia's political and economic institutions were unable to handle the adjustments required in the wake of the Asian financial crisis. The upshot is that Indonesia remains mired in a crisis that appears to have put a complete stop to its growth process. Perhaps what set countries like China and India (as well as South Korea or Taiwan) apart from Indonesia is that these countries have used economic growth as an opportunity to undertake further institutional reforms along the way.

The collapse of growth in the case of a country like Venezuela is much harder to explain on the basis of conventional indicators of institutional weakness. As Ricardo Hausmann explains in his narrative, Venezuela was seen as the most stable democracy in Latin America, with a strong party system, free press, and solid labour and business organizations to negotiate social conflicts. Yet Venezuela's growth rate, once Latin America's fastest at 6.4 percent per annum, has collapsed to the point where output per worker in the non-oil economy is almost half what it was in 1980. What happened? Hausmann focuses on two explanations. The neoclassical explanation is that the decline in the value of oil exports has reduced income and correspondingly (noned) output. But Hausmann's calculations suggest that this cannot account for more than half of the collapse. The second factor is a rise in country risk, reflected in Venezuela's country ratings and contractual interest rates, which has reduced the desired capital stock. What lies behind this, according to Hausmann, is the inability to settle distributive conflicts in the wake of a collapse in oil income. Venezuela has simply become a riskier environment, which in turn has eroded the quality of public institutions and their legitimacy. This argument is reminiscent of the importance Pritchett attaches to the rules of the game. It suggests that countries can trip up even when their institutions appear strong by conventional yardsticks.

Dani Rodrik is Professor of International Political Economy at the John F. Kennedy School of Government (KSG), Harvard University.

References

Rodrik, D. (2003) In Search of Prosperity, Princeton University Press.