Journal Issue

Letters to the Editor

International Monetary Fund. External Relations Dept.
Published Date:
December 2003
  • ShareShare
Show Summary Details

Deeper into the institutions debate

I read the articles in the grouping “Institutions in Development” (June 2003) hoping to find some thinking outside the box of extractive economic growth, the sustainable development oxymoron. Seeking economic prosperity does not work well in locations unable to command extractive resources to process and sell. In such places, the goal is more fundamental—human well-being found in the synergistic measures of long life expectancy, low infant mortality, high educational attainment, and low fertility.

The United Nations Development Program report for 1997 noted that “many aspects of deprivation—from poor health to discrimination to domestic violence—have little to do with income.” It cited the example of Haryana, a North Indian state, whose infant mortality rate (68 per 1,000 live births) was four times that of Kerala, a South Indian state. Yet Haryana’s economic growth rate was 3.2 percent, while Kerala’s was 0.3 percent. Kerala has achieved a higher level of well-being than Haryana because of traditional differences in family institutions—Kerala has weak patriarchal family structures, while women in Haryana suffer “systematic deprivation.” Those prepared to consider human well-being a desirable goal can learn from a large society like Kerala (32 million people), which has already achieved the two behaviors required for human sustainability—modest consumption of ecosystem resources and small families.

William M. Alexander

Emeritus Professor of World Food Politics California Polytechnic State University

You selected a felicitous quote from my 1973 article in F&D on climate (“In Brief,” June 2003): climatic factors (in the tropics and the polar regions, for example) can severely hamper development. Daron Acemoglu, in his article in the same issue, argues that geographic factors do not influence development. His proof is that, five centuries ago, the Mughals, Aztecs, and Incas, living in tropical areas, were richer than societies in temperate areas and they are now poorer. His argument is irrelevant: Babar, the founder of the Mughal empire, extended his reign from Kabul and established the capital in Delhi—both in the temperate zone; the Aztecs and Incas inhabited high elevations—altitude overrides latitude. Africa is the preeminent tropical continent, and tropical Africa has never, in recorded history, been richer than Europe.

Acemoglu states that Europeans introduced the wrong kinds of institutions in the Congo, the Caribbean, and Central America, but the right kinds in Australia, Canada, New Zealand, and the United States. Note that the countries in the first list are in the tropics, while those in the second list are in the temperate zone. A coincidence? Or was there something about the climate that influenced European decisions? Acemoglu overlooks the Dutch settlement of tropical Northeast Brazil around 500 years ago. Even though the Dutch brought their wonderful institutions with them, it is one of the poorest parts of Brazil. Could it be that the area’s tropical climate had some influence?

Trying to identify a unique determining factor for the economic fate of all countries is a futile exercise. My book The Tropics and Economic Development (Johns Hopkins Press for the World Bank, 1976) identifies what it is in a tropical environment that causes problems for economic development. My book Economics for the Twenty-first Century (Ashgate, 2001) discusses other factors that influence economic development: civil society, national cultures, social capital, the public sector, and corporate governance.

Andrew M. Kamarck

Brewster, Massachusetts

The novelty of the views expressed in the articles on the ultimate causes of economic development—in particular, the role of geography and institutions—is that they are based on empirical facts derived from multiple regressions. Linear regression has become one of the major (if not the major) sources of proof in economic analysis, although it is often blind to facts familiar to other social scientists, either because these facts cannot easily be quantified or because economists simply choose to ignore them. Economists should pay more attention to the complex social processes shaping economic development.

When the Spanish arrived in the West Indies and South America in the early sixteenth century, they brought with them a system of agricultural production based on a mixture of forced and bound labor on large estates. This system, which prevailed in most of southern Europe until the twentieth century, failed to generate economic development primarily because it maintained very low wages that discouraged productivity, kept farmers’ incomes low, and hampered capital accumulation by farmers. It has been one of the major causes of underdevelopment in Latin America and, though disappearing fast, is still present in various forms. In contrast, when the British settled North America in the early seventeenth century and, later, Australia and New Zealand, they brought with them a very different production system—namely, individually owned farms, mercantilism, and, eventually, capitalism. This system boosted production, increased farmers’ incomes, and sustained economic growth.

Jeffrey Sachs refers to Adam Smith’s argument about Central Asia’s geographic isolation. Central Asia was, for many centuries, at the center of the trade route known as the Silk Road. Egypt and the Roman Empire imported silk from China for a long time. The Silk Road further developed when the Arab empire pacified most of the Middle East. The region had large, wealthy metropolises like Samarqand and Bukhara, which were major trade, political, and cultural centers (it suffered a brief setback during the Mongol invasion of the thirteenth century) until the development of sea trade between Europe, South Asia, and the Far East in the seventeenth and eighteenth centuries. When Adam Smith wrote, Central Asia had lost its comparative advantage not because of changes in geography but because of the revolution in transportation.

Other examples also illustrate that there is no “geographic fate” that can explain economic development or underdevelopment. Landlocked, mountainous Switzerland is the wealthiest country in Western Europe despite its geographical handicap. Japan, which could have been handicapped by its peripheral location relative to major international trade routes, its mountainous landscape, and its lack of natural resources, is the wealthiest country in eastern Asia. Both owe their wealth to smart economic policies. Botswana, Africa’s success story, is landlocked, and far from any trade roads, and largely desert. It has natural resources, but less than nearby Angola. Given its natural resources, climate, location, and demographics, Angola should be better off, but Botswana is way ahead because of several decades of efficient economic policies and investment, good management, and institutions that promote peace and the rule of law.

Other arguments supporting the geography thesis have to do with the impact of health on development: malaria, AIDS, tuberculosis, malnutrition, and violence are seen as major causes of underdevelopment. But deadly diseases (smallpox, measles, cholera, plague) have existed in other regions (Europe, for example) without stalling economic development. To the contrary—meeting the challenge of these diseases gave rise to modern public health systems (clean water and sanitation, vaccination), which, in turn, had a salutary economic effect.

The quality of leaders is a critical element of sustained development. Competent leaders are able and willing to meet the challenges posed by geography, history, and international competition. Some countries have evolved smoothly, finding the leaders they needed, while others have suffered political crises and turned to charismatic leaders to rescue them. Countries also need people who work hard at managing the state and improving economic policies. Historical evidence shows that economic development is not possible without such leadership, and it is certainly impossible in the face of extensive corruption.

When we compare levels of economic development in different countries, we also need to consider the time frame. It took about two centuries for European countries and their offspring (Australia, Canada, New Zealand, and the United States) to reach the level of development we now use as a reference (per capita income equivalent to 10,000 U.S. dollars or euros in purchasing power parity). Countries starting later have an advantage: they have a model to follow. It took Japan only about a century to reach a similar level of income per capita and industrialization. In our own day, it is happening even faster. It took Taiwan Province of China and the Republic of Korea only about 50 years, after adopting the right policies, to get there.

Africa is the biggest challenge. Africa is just starting to recover from hundreds of years of political, social, and economic instability and domination: European penetration, the slave trade, colonization, and the Cold War. Since 2001, however, an African Union is being built and a new program of development has been undertaken (NEPAD). How long it will take Africa to raise incomes remains to be seen. In any case, Africa is well located and widely open to sea trade routes to the other continents. It has an abundance of natural resources and large rivers that permit deep penetration of the continent. So far, its development has been limited primarily by political instability and lack of social capital. This can be changed rapidly if proper policies are followed. There is no geographical, ethnographical, or historical fate. It is a matter of time, institutions, leadership, and proper economic policies.

Michel Garenne

Institut de recherche pour le développement

Paris, France

Eneas Gakusi

Centre d’études et de recherches sur le développement international Clermont-Ferrand, France

No global middle class

In the September issue of F&D, Prakash Loungani discusses various measures of inequality. He rightly concludes that inequality measured by looking at cross-country mean incomes (or, rather, GDP per capita) has been rising over the past quarter of a century. Thus, at the international level, we have growing divergence between countries instead of the much-heralded convergence. He goes on to discuss global inequality, which is inequality between all individuals in the world, regardless of where they live, and states that we observe “convergence, period.” This statement is inaccurate. It is probably based on recent work by Surjit Bhalla and Xavier Sala-i-Martin, who both use the Deininger and Squire (DS) database, which is inappropriate for the calculation of global inequality. To mention just a few problems: (1) quintile shares (five data points) are used to approximate entire country distributions; (2) for 85 percent of the countries/years, even quintile shares are not available, so the authors assume either that there was no change in distribution between the years for which the data are available (sic!) or that the change was linear; (3) while, originally, quintile shares were calculated from survey data, the authors multiply the shares, not by survey means as should be done, but by GDP per capita (thus mixing up survey and national accounts data); (4) the quintile shares given in the DS database refer to both households and individuals; the authors mix these categories up and treat households as if they were individuals; and (5) quintile shares of expenditures and income are treated as equivalent.

These results are thus driven not by the data but by the authors’ assumptions. They ignore changes in income inequality within countries (see point 2) and assume, in the face of all the evidence to the contrary (see Martin Ravallion “Should Poverty Measures Be Anchored to the National Accounts?” Economic and Political Weekly, Vol. 34 (August 26), pp. 3245–52), that the difference between GDP per capita and survey means is distribution-neutral. In plain language, it means that the excess value of GDP per capita compared with survey means—which we know is due to underreporting of property and self-employment income and undersurveying of the rich—is allocated equally across the board. Thus, the poor are, as it were, “given” property income they never received. For sure, the poor would not be poor if we assumed them to be rich!

Bhalla and Sala-i-Martin do not calculate global inequality but cross-country inequality adjusted for population size. The difference between the two is precisely within-country inequality. To put it more simply, while Bhalla and Sala-i-Martin account for the fact that global inequality is shrinking because China and India are growing faster than the rich countries, they disregard the effect of widening inequality within China and India. Their results are, at best, unproven and, at worst, misleading. This is a point I argued in my paper “The Ricardian Vice: Why Sala-i-Martin’s calculations of global inequality are wrong” (available at

In the only study of global inequality based directly on survey data (which is methodologically the only correct way to calculate world inequality), I found a significant increase in global inequality between 1988 and 1993 and a decline between 1993 and 1998 (“True world income distribution, 1988 and 1993: First calculation based on household surveys alone,” Economic Journal, Vol. 112 (January), pp. 51–92). My study is also affected by problem (5) listed above, but it certainly shows that if we use the proper procedure, it is much more difficult to come to the conclusion that global inequality is decreasing.

Finally, disagreements about the direction of change in global inequality, which—whether positive or negative—is, in any case, small and not necessarily statistically significant, detract from the key issue on which all authors agree: that global inequality is incredibly high, with a Gini coefficient [editor’s note: a measure of inequality with 0 representing perfect equality and 100, perfect inequality] in the upper 60s for purchasing-power-parity-adjusted incomes and a Gini of 80 if one uses exchange rates. The global distribution of welfare is even more unequal than distribution in the most inegalitarian countries, like Brazil and South Africa. To speak, as some authors do, of a “world middle class” is just plain wrong.

Branko Milanovic

World Bank

Washington, D.C.

Prakash Loungani replies

My characterization of the evidence on global inequality and the emergence of a global middle class was based, in part, on the views expressed in surveys by Stanley Fischer (2003 Ely Lecture to the American Economic Association) and the World Bank’s David Dollar and Aart Kraay (Foreign Affairs, January 2002). I disregarded Milanovic’s evidence of an increase in inequality between 1988 and 1993 because of the limited time period of the analysis; as David Dollar has pointed out, “the period from 1988 to 1993 was the one in the past 20 years that was not good for poor people in China and India.” Milanovic is right, nevertheless, to remind us of the limitations of the so-called Deininger-Squire data set that is provided by the World Bank.

Why isn’t Africa growing?

In “Unlocking Growth in Africa” (June 2003), Kenneth Rogoff’s first recipe for growth is that “African countries must become more open to trade and foreign direct investment, and their efforts must be reciprocated.” Trade and investments are key to growth in Africa. Aid—even in the form of grants—may breed dependency. Loans, when properly managed and when there are clear paths for repayment, may be useful. When, however, industrial countries provide grants and loans principally to create employment in their domestic economies, the capacity of such aid to help growth in Africa may be limited. For example, when aid in support of agricultural projects in Africa comes in the form of vehicles (sometimes posh), the projects contribute minimally to growth and the loans are rarely repaid. And aid in the form of technologies—especially machinery that the recipient country does not have the capacity to maintain—often turns into a tale of woes and abandoned scraps.

Opening up to trade, in Nigeria’s case, has meant massive imports of goods ranging from toothpicks, cereals, livestock products, and health drinks to new and dilapidated vehicles. The proportion of Nigerians employed in the manufacturing sector has shrunk. Businesses and firms have closed. Poverty is rampant. Despite the absence of reliable statistics, it is doubtful that Nigeria has experienced real growth, at least since structural adjustment in the mid-1980s. One proof of this has been the increased rate of illegal immigration by Nigerians, pressured by economic conditions at home, to Europe and North America.

While the United States subsidizes its steel industry, and the industrial countries spend $300 billion a year on agricultural subsidies, Nigeria and other less developed countries eliminate subsidies and reduce spending on health and education. The result of these actions, as well as of the devaluation of domestic currencies and trade liberalization, is negative real growth and poverty. Presently, Nigeria is dependent on imported rice and poultry. Nigeria’s poultry industry has largely collapsed.

Dr. Chinedum Nwajiuba

Department of Agricultural Economics

Imo State University, Owerri, Nigeria

Congratulations to the F&D team on the new look of the magazine, which is a delight to read. I was particularly struck by the quality and clear analysis in the Straight Talk article entitled “Unlocking Growth in Africa.” This excellent analysis reminds us that aid (accumulation of physical capital) is not an engine of growth. Growth, it notes, depends much more on “soft factors” (institutions and governance). This has been confirmed by a recent IMF study showing that differences in growth rates between countries are attributable to the quality of their institutions. The author of the article shows clearly that, despite the large loans that characterized the 1970s, economies did not take off for reasons we know all too well, including the choice of projects that turned out to be white elephants, a significant proportion of credits allocated to “indirectly productive” social sectors that could not generate the resources needed for debt repayment, and problems in managing aid. Finally, as the quality of institutions becomes a factor of growth, the teams preparing the much talked-about poverty reduction papers must be tearing their hair out!

Jean Pierre Kpata

Bangui, Central African Republic

IMF should blow the whistle

Kenneth Rogoff correctly suggests crises that the IMF’s future hinges on its being more candid in its country assessments, even if that sometimes means bringing forward a crisis that might have been delayed (“More Cheerleading or More Whistle-Blowing?” September 2003).

A crucial issue that Professor Rogoff does not address is how the IMF’s case-by-case lending policies since 1995 might have played a role in past financial crises and how they run the risk of amplifying future emerging market financial crises. In the effective absence of access limits on IMF lending since the 1995 Mexican peso crisis, financial markets have often been encouraged by actual and prospective large-scale IMF lending programs to engage in “moral hazard” lending. As the recent Argentine and Russian crises illustrate, this has tended to delay debt crises and to make them more severe when they eventually occur. It has also resulted in countries being saddled with large quantities of senior debt, which cannot be restructured. This has the unfortunate effect of constraining the debt-burdened country’s future macroeconomic policy options in the postcrisis period.

One would hope that, beyond introducing greater candor in its surveillance activities, the IMF would restore more transparency to its lending operations. A good place to start would be a return to the access limit policy the IMF followed before the Mexican crisis.

Desmond Lachman

Resident Fellow

American Enterprise Institute Washington, D.C.

If countries follow the suggested policies, they should be automatically eligible for IMF loans. If they don’t, they shouldn’t get any loans and the IMF wont bail them out.

Allan H. Meltzer

The Allan H. Meltzer Professor of Political Economy and Public Policy

Carnegie Mellon University

distinguished economist, consultant to the World Bank and various central banks and author of The History of the Federal Reserve discusses the Fed and current controversies from deflation to international bailouts.

Only in The Region, the quarterly magazine of the Federal Reserve Bank of Minneapolis

Other Resources Citing This Publication