Another Lost Decade?
Latin America's Growth Failure Continues into the 21st Century

Mark Weisbrot and David Rosnick[1]

November 13, 2003

Executive Summary

With less than a quarter remaining in 2003, and well-formed projections for 2004 now available, there is a reasonably good basis for assessing the economic growth record of the first five years of the current decade. This half-decade looks very bad for Latin America, with almost no growth of GDP per person over the years 2000-2004 (see chart below).

The 1980s were known as the "lost decade" for Latin America, with income per person, the most basic measure of economic well-being, actually shrinking from 1980-1989. The nineties saw positive per capita growth in the region, although still sluggish by historic standards for Latin America and for developing countries generally. Nonetheless there had been some hope that the region had finally reversed its long-term economic malaise.

Among policy-makers and economists in the United States it has been widely assumed that the economic policy changes which began to be implemented in Latin America in the early 1980s would eventually bear fruit, and lead to strong economic growth. A quarter century later, this has not yet happened.

This paper finds that:

        For the first 5 years of the current decade, 2000-2004, per capita GDP in Latin America is expected to grow by 0.2 percent annually, or about 1 percent for the whole period.

        This continues a long period of economic failure: for the prior 20 years, 1980-1999, the region grew by only 11 percent (in per capita terms) over the whole period. This is the worst 20-year growth performance for more than a century, even including the years of the Great Depression.

        By comparison, for the two decades from 1960-1979, Latin America experienced per capita GDP growth of 80 percent.

        The current decade will not be a success by any reasonable measure: to match the performance of the 1960-79 period would require annual per capita growth of 5.9 percent for the remainder of the decade. Most economists would view this outcome as nearly impossible. Even to achieve the slow growth of the 1990s would require annual per capita growth of 2.5 percent for the second half of the decade.

        There is reason to believe that growth in the near future, including 2004, may be worse than anticipated. Among these reasons are current weaknesses in the U.S. and several of the larger Latin American economies.

There has been very little mention of this long-term economic failure in the policy debates, and almost no discussion of its possible causes. The authors conclude that such discussion is overdue.

Introduction

The 1980s were known as the "lost decade" for Latin America, with income per person, the most basic measure of economic well-being, actually shrinking (by 3.1%) from 1980-1989. The nineties saw positive per capita growth, 1.4% annually, although still very sluggish by historic standards for Latin America and for developing countries generally. Nonetheless there had been some hope that the continent had finally reversed its long-term economic malaise.

Among policy-makers and economists in the United States it has been widely assumed that the economic policy changes which began to be implemented in the early 1980s would eventually bear fruit and lead to strong economic growth. These policy changes have included a lowering of tariff and non-tariff barriers to trade; removal of restrictions on international capital flows; the abandonment of industrial policy or government-directed development strategies; large-scale privatization of public enterprises and public pension systems; and in many countries, the implementation of monetary and fiscal policies approved by the International Monetary Fund.

These policy changes were intended to increase efficiency at the microeconomic levels of industry and agriculture, promote macro-economic stability, increase investment and productivity, and therefore lead to a higher rate of growth. It is important to emphasize this, because � ignoring questions of distribution, which these policy changes were not intended to address � economic policies are not generally considered successful if they do not increase GDP (or income) per person.[2] Furthermore, even from the standpoint of alleviating poverty or other social goals, it is difficult if not impossible to achieve such goals without economic growth. While in theory it might be possible to redistribute substantial amounts of current income and wealth from the upper classes to the rest of the population, as a practical matter this is often not feasible. Economic growth allows for the possibility of redistributing newly generated income towards the poor � often the majority in developing countries � without having to drastically reduce living standards among the non-poor.

This paper looks at the data for Latin American growth over a 55-year period beginning in 1950. As will be seen below, the years since 1980 have been characterized by a remarkable slowdown in economic growth, compared with performance in the previous 20 years. With five years of data for the first decade of the 21st century (including projections for the rest of 2003 and 2004), there is no sign so far that Latin America's long period of poor economic performance is coming to an end.

The annual growth rate of per capita GDP for Latin America and the Caribbean for 1960-1999 is shown in Figure 1. As can be seen there is a striking falloff in growth from the period 1980-1999, as compared with the previous two decades, 1960-1979. For 1980-1999, the region's per capita GDP grew at an annual rate of only 0.5 percent, a cumulative total of 11 percent for the two decades. By comparison, from 1960-1979, per capita growth was 3.0 percent, or 80 percent for these two decades.

The comparison between the growth rates of these two periods is a fair and appropriate comparison. The year 1979 was a business cycle peak for the United States economy, which has a large impact on Latin America because it is the destination for a large proportion of the region's exports (currently almost two-thirds), and also a large source of capital flows. On both sides of the divide, there is one decade of very good U.S. economic performance � the 1990s witnessed the longest running business-cycle expansion in U.S. history, and the 1960s were a similar period of high growth. Also, the 1970s were a period that contained severe oil shocks � 1974-75 and again at the end of the decade � high inflation, and a world recession. The 1980s also contained a world (and U.S.) recession. The comparison of 1980-1999 with the previous 20 years is therefore using an appropriate baseline.

Figure 2 shows the rate of growth of per capita income for the region by decade. The most recent data is for the first five years of the present decade (2000-2004). The data for 2003 and 2004 are projections from the IMF.[3]  This shows the first half of the decade registering extremely poor economic performance, at 0.2 percent annual per capita income growth. This is even worse than the 0.5 percent annual growth of the 1980-1999 period.

The numbers in Figure 2 include a downturn during 2001-2002, undoubtedly influenced by the U.S. recession, which was officially dated as March to November 2001. It is possible that the second half of this decade will show much faster growth, but there is no reason to assume that it will, or that there will be no economic downturn in the second half of the decade.

In any case the poor performance of the first five years will weigh heavily on the results of the current decade. The region would need an annual per capita growth rate of 2.5 percent in the second half, simply to match the slow growth of the 1990s; to match the more successful growth of the 1960-1979 period would require annual per capita growth of nearly 5.9 percent, a goal that most economists would see as nearly impossible.[4]

There is also reason to believe that the projected growth for 2004 may turn out to be lower than originally forecasted. First, IMF growth projections have historically turned out to be considerably higher than observed growth. For example, the IMF's spring projections for the following year overstated growth in 17 of the last 20 years. The average overstatement was 1.6 percentage points, which is a very large error relative to the size of annual growth projections.[5]

Furthermore, although the United States experienced rapid growth (7.2 percent) in the third quarter, it has numerous signs of weakness that could abort the current recovery, and fourth quarter growth is expected to be much slower. Business investment has yet to recover to its pre-recession levels, and consumers are holding record levels of mortgage and consumer debt relative to disposable income. Although the economy has begun to add jobs, the labor market remains weak, having lost 2.4 million jobs since the last business cycle peak of February 2001, and with wages barely outpacing inflation over the last year. State and local governments are trying to close budget gaps totaling more than $80 billion over the next fiscal year. And as the IMF noted in its most recent World Economic Outlook,[6] the U.S. suffers from a housing bubble, the collapse of which could have an enormous impact on demand � an estimated $3 trillion in private wealth could disappear.[7]

Figure 3 shows the average annual growth for 2000-2004, with the 2004 projections separated out. Assuming the projections for 2003 are accurate, it is clear that the even the barely positive growth for the first half-decade depends heavily on 2004. Growth in per capita GDP for 2000-2003 is negative at �0.2%; the projection for 2004 is 2.1%, pulling the 5 year average up to a positive 0.2%.

           

But even if these seemingly optimistic projections for 2004 come true, the first half decade will have been a failure by any plausible comparison, and the second half would need truly extraordinary performance in order to rescue the decade.

The weakness in the U.S. economy could potentially affect Latin American growth even farther out than 2004. The same is true for the effect of a fall in the U.S. dollar, which is overvalued. The U.S. is running an unsustainable current account deficit that has reached a record 5.0 percent of GDP.  This deficit makes it virtually certain that the dollar will continue to decline against other currencies. To the extent that the dollar falls against Latin American currencies, this will reduce Latin American growth by lowering the price of U.S. exports to the region and raising the price of imports from Latin America in the United States.

If we look at some of the larger individual countries, there are additional reasons for concern. Brazil is currently projected to grow at 0.5 percent for 2003,[8] which is negative on a per capita basis. There are no clear signs of a turn-around in the foreseeable future. Unemployment has hovered near 13 percent, and real wages have fallen by about 16 percent since the fourth quarter of last year.  More importantly, real interest rates are currently extremely high: inflation (Consumer Price Index) has run at 5.4 percent over the last quarter, while the central bank's official overnight rate is set at 19 percent. The government is also committed to maintaining a primary surplus in the central government budget of 4.25 percent. These tight monetary and fiscal policies, if not reversed, could limit growth for some time to come.

Growth for Mexico in 2003 is currently forecast at 1.6 percent, which would be flat or negative per capita. For 2004, the government is projecting between 3 and 3.5 percent growth, but there is currently no sign of acceleration. The Mexican Finance Ministry estimates a growth rate of 0.7 percent annually for the third quarter (GDP, not per capita). The Mexican economy is particularly sensitive to the strength of any recovery in the United States, which is the destination for 85 percent of its exports (about 21 percent of GDP).

In Argentina, forecasters are projecting real GDP growth of 5.8 percent (not per capita) for 2003, after a string of 17 quarters of negative growth. The IMF is projecting 2.8 percent growth for 2004. Nonetheless, there are clouds that remain over the economy. Most important is the future of negotiation on the country's debt. If the new government agrees to unsustainable debt payments on the estimated $94 billion of defaulted debt, and/or (as in the past) adopts contractionary monetary and fiscal policies, the country could lapse back into recession. For example, even if the government's current offer to pay 25 percent of the defaulted debt were accepted, the country would still have an unsustainable debt burden of over 90 percent of GDP.

But as noted above, even if all goes as well as forecast for 2004, the first half of the current decade, at 0.2 percent annual per capita growth, is certainly a failure. And it indicates that a sub-par performance for Latin America for the decade is extremely likely, if not inevitable.

Additionally, there can be no doubt about the terrible performance of the long-term trend. Again, using the 1960-1979 period as a baseline, the quarter century for 1980-2004 is dismal. Annual growth in GDP per capita registers a mere 0.5 percent, as opposed to 3.0 percent over the previous period. Countries that are now considered relatively successful are not doing very well compared to past performance. For example, Mexico registers 0.8 percent annual per capita growth for 1980-2004, as compared with 3.3 percent for 1960-79. For Brazil, which one had one of the fastest growing economies in the world, per capita growth is only 0.8 percent annually for 1980-2004, as compared with 4.9 percent for 1960-79.[9] Even for the years 1994-2002, which are generally perceived as a period of "successful, free-market reforms," per capita growth in Brazil was only about 1.7 percent annually, and the country piled up an enormous public debt (it increased from 29 to more than 60 percent of GDP).

Conclusion

Latin America's long-term growth failure has attracted very little attention, given the magnitude of the phenomenon and its importance to the well-being of the population there.[10] This is also true more generally of the sharp slowdown in growth in the vast majority of developing countries over the last two decades, as well as the reduction in the rate of progress on the major social indicators (life expectancy, infant mortality, literacy and education), which would be expected in a period of sharply reduced growth.[11] But in Latin America the slowdown is particularly striking as it coincides with a long period during which sweeping policy reforms were implemented, and follows on the heels of respectable average growth rates.

It is very difficult to show econometrically a relationship between any major economic policy change and economic growth (or the lack of it), primarily because of the difficulties in controlling for other changes that take place during the period in question. It is therefore appropriate to look at a long period of economic failure, such as the years 1980-2004 for Latin America, and question whether policy changes had anything to do with this failure. At the very least these policy changes cannot be said to have been successful in any meaningful economic sense. In some ways it is an historical coincidence that this obvious and compelling issue has barely been raised: most of the critics of the "Washington Consensus" or "globalization" who have gotten public attention in recent years have focused on issues of poverty and inequality, environmental degradation, or problems of governance with regard to multilateral institutions. And there is no doubt that these are serious problems � inequality of income and wealth in Latin America, for example, is far beyond the level of developed countries and has probably worsened over the last 25 years. But to the extent that poverty has worsened in some countries, or progress towards the alleviation of poverty has been arrested in Latin America during these decades, it is overwhelmingly due to the region's sharp slowdown in economic growth, much more than any increases in inequality.

There are several common responses to the growth slowdown among policy-makers and economists, other than mostly ignoring it. Perhaps one of the oldest has been the argument that the decline of the 1980s at least, and perhaps beyond, was the result of growth in the 1970s that was either unsustainable (excessive borrowing, current account deficits) or had run up against certain inherent constraints � for example, industrialization by means of import substitution had run its course. One problem with this argument is that the different countries in the region were at very different stages of economic development when the slowdown began, yet almost all were seriously affected by it. Of course the U.S. and world recession of 1980-82 could explain the onset of the slowdown for Latin America, but the prolonged slump over more than two decades indicates that some structural and policy changes are at work.

This argument also blurs the distinction between policy decisions in later decades and the results of prior decisions. For example, if the Brady debt restructuring of 1989 left some countries with unsustainable debt, and that led to chronic current account deficits, depreciation/inflationary spirals (e.g. Brazil) and other macroeconomic instability, then this restructuring can also be considered a policy mistake rather than simply a direct and unavoidable consequence of over-borrowing during the prior decades. And as time goes on it becomes more difficult to say that the continuing poor growth performance of Latin America in the 21st century is simply a result of bad decisions made 30 or more years ago.

Another common argument is that the policy reforms have not had enough time to work. While it is true that various reforms were implemented in different countries at different times, there is still no post-reform time period during which we can say there was a success for the region, or even a significant part of the region. Even if we assume that the reforms could not have taken effect until the 1990s, the growth of the 1990s (1.4 percent annually per capita) was slow, and is now practically absent for the first half of the current decade. At some point there has to be some questioning of whether these reforms actually lead to growth, and a debate over which reforms may have contributed to this prolonged period of economic failure.

In the last few years there has been a more general response that "globalizing" countries have been successful, while others have not.[12] This research runs into the econometric problems noted above � most importantly, countries that grow fast tend to increase their share of trade in GDP, so the causality can run in the other direction � i.e. fast-growing countries increase trade as a percentage of GDP.[13] In any case, even if it were true that there were some positive relationship between some liberalizing reforms and growth, this argument still cannot explain why so many countries have done so much worse over the last 20-25 years, regardless of how much they adopted these reforms. This is true for the vast majority of developing countries in the world, but it is especially true for Latin America.

In sum, an economic failure � looking simply at growth � of the length and magnitude that we have seen since 1980 is unprecedented in the history of Latin America, at least in the last 100 years. There has been very little debate at all about the possible causes of this failure. That debate is long overdue.

Appendix       


[1] Mark Weisbrot is the Co-Director and David Rosnick a research associate at the Center for Economic and Policy Research.

[2] An exception would be if productivity (output per unit of labor) increases but people choose to take these gains in the form of more leisure time, i.e. longer vacations, shorter hours, and/or earlier retirement.

[3] International Monetary Fund.  World Economic Outlook, September 2003.

[4] Rates of per capita income growth of this magnitude � and even higher � have been achieved and sustained over decades in countries such as South Korea and Taiwan, under very different policies than are being implemented in Latin America today. South Korea has grown by an average annual rate of 6.0 percent, per capita, per year from 1960 to 2002.

[5] See Dean Baker and David Rosnick (2003), "Too Sunny In Latin America? The IMF�s Overly Optimistic Growth Projections and Their Consequences." The projections used in the current paper, for 2004, are from the World Economic Outlook released in September 2003; they may therefore be more accurate than those examined by Baker and Rosnick, which were estimates published by the IMF in the spring of each year.

[6] International Monetary Fund. World Economic Outlook.  September 2003. Chapter 1: Economic Prospects and Policy Issues, p. 6-12.

[7] "The Run-Up in Home Prices: Is It Real or Is It Another Bubble?" by Dean Baker, August 5, 2002.

[8] This is the most recent number from Fitch Ratings; the IMF projects a positive 1.5% per capita growth for Brazil in 2003, which is used in this paper.

[9] The cumulative numbers for per capita income growth over the two periods (1960-1979 versus 1980-2004) show the contrast between rapidly rising living standards for a generation versus stagnation, even ignoring that the second period is five years longer: in Brazil, 160 percent vs. 22 percent; in Mexico, 93 percent vs. 21 percent.

[10] One rare exception was the Economist, �Wanted: a new regional agenda for economic growth.� April 24, 2003.

[11] See Weisbrot, Baker, Kraev, and Chen, "The Scorecard on Globalization 1980-2000: Twenty Years of Diminished Progress" (2001).

[12] See e.g. David Dollar and Aart Kraay, "Trade, Growth, and Poverty," (World Bank, 2001)

[13] Rodrik (2000), "Comments on Trade, Growth, and Poverty," and Rodriguez and Rodrik (2000, "Trade Policy and Economic Growth: A Skeptic's Guide to the Cross-National Evidence) have demonstrated flaws in Dollar and Kraay's arguments, as well as other studies purporting to show a causal relationship between "openness" and growth.