James Vreeland
At the writing of this book, 49 developing countries around the world – whose populations account for more than one billion people – are participating in economic programs supported by the International Monetary Fund (IMF or Fund). These “IMF programs” grant the governments of these countries access to IMF loans, but access to the loans can be cut off if the governments fail to comply with specific policy conditions. IMF policy conditions impact the lives of individuals living in these countries in intimate ways: the policy conditions address government expenditures, so IMF programs help determine whether roads, schools, or debt repayment take priority. The policy conditions also address interest rates, so they may affect one’s ability to borrow to purchase a home or invest in a business. IMF policy conditions often address the value of the national currency, so IMF programs may impact the very purchasing power of the money in people’s pockets.
Not surprisingly, the IMF is well known throughout the developing world – to the elites and the masses alike. The organization often appears to exercise as much or even more authority than their own governments. Yet, the IMF is less familiar to average citizens in the developed world. And, to many throughout the world, the actual functioning of the organization is unknown or misunderstood. Unfounded opinion about the IMF abounds among people who often lump it together with other international institutions like the World Bank and the World Trade Organization, even though the administration and purposes of the IMF are quite distinct from these other international institutions.
Founded in the wake of the Great Depression, the IMF can be thought of as an international credit union with access to a pool of resources provided by the subscriptions of its members, which include nearly every country in the world. The size of a country’s contribution depends on the country’s economic dominance, hence, the bulk of the resources of the IMF come from the developed world. The Fund can lend from this pool of resources to countries facing economic problems. These days, the only countries that borrow from the IMF come from the developing world.
IMF loans can be thought of as a form of insurance for governments against the possibility of an economic crisis. Such insurance, however, introduces something economists call “moral hazard”: the prospect of receiving assistance in the face of an economic crisis in the form of an IMF loan may itself lower a government’s incentive to avoid the bad economic policies that cause economic crises in the first place.
To counter moral hazard, the IMF imposes conditionality: governments are required to follow what the IMF deems as “good” policies in return for the continued disbursements of the IMF loan. Thus, one can think of an IMF program as having two components: the loan and the conditions attached to the loan. The goal of this arrangement is to first stabilize a country facing a balance of payments crisis and then to promote growth and the reduction of poverty.
Yet, conditionality is controversial. If the policies imposed by the IMF are so good for countries, why must governments be enticed through conditional lending? At the heart of this question is national sovereignty, and beyond purely economic guidelines, the imposition of IMF programs is heavily influenced by international and domestic politics.
International politics play a role because powerful members sometimes use their influence at the IMF to pursue political goals. Votes at the IMF, like contributions, are pegged to a country’s economic size, so economically powerful countries have more say at the IMF than other countries, and can pressure the Fund to do their bidding. Governments who are considered important allies of the IMF’s most influential members – like the United States – sometimes receive preferential treatment from the IMF. The IMF may bail them out of economic crises with large loans even if they fail to comply with IMF conditions of changing economic policy.
Yet, at the domestic level of politics in developing countries, there are other cases where governments actually want IMF conditions to be imposed. These governments seek the assistance of the international institution to get around domestic political constraints and force changes in economic policy. Governments can use IMF conditionality to gain leverage over domestic opposition to policy change. Sometimes, such policy changes result in superior outcomes for society, but often IMF leverage is used to protect elites and make others bear the cost of an economic crisis.
Unfortunately, there is scant evidence of the success of IMF conditionality. Studies have even found that IMF programs hurt economic growth. A further effect of IMF programs is the increase of income inequality. This is not just because the IMF is involved with countries that already have economic problems – even accounting for this fact, these disappointing results hold.
There is little consensus over why IMF programs have the perverse effects that they do. Some argue that the influence of international political pressures has led to low levels of compliance with IMF conditionality. As a result, IMF lending simply subsidizes the continuation of bad economic policies. Others argue that the economic policies imposed by the IMF are the wrong ones. Instead of imposing austerity, the IMF should promote economic stimulus packages so that developing countries can grow their way out of economic problems. Still others argue that failure is due to domestic politics. Policy may change under IMF programs, but governments implement only selected reforms or impose partial reform with the goal of insulating domestic political elites and placing the burden of the economic crisis on labor and the poor. Strangely, with all of these various points of view, there is a broad based consensus that the IMF should scale back its operations. Many feel that the IMF should get out of the development business.
Recently, however, the IMF has made a bold new commitment to promote economic development through continued conditional lending. Thus, IMF programs remain a presence throughout most of the developing world. In some countries, participation in IMF programs is business as usual, a routine way of life. […]
If IMF programs have any effect, it should be on the balance of payments (BOP). First and foremost, the Articles of Agreement mandate the IMF to address problems in this area. What is the balance of payments? The IMF defines a country’s overall balance of payments as the sum of the “current account,” the “capital account,” and the “financial account” plus “net errors and omissions.” The current account of the balance of payments is the credits minus the debits of goods, services, income, and current transfers. The capital account refers mainly to transfers of fixed assets and nonproduced, nonfinancial assets. The financial account is the net sum of the balance of direct investment, portfolio investment and other investment transactions. Net errors and omissions reflect statistical inconsistencies in the recording of entries and are included so that all debit and credit entries in the balance of payments statement sum to zero. By construction (of net errors and omissions), the overall balance of payments is equal to net changes in “reserves and related items,” the sum of transactions in reserve assets, exceptional financing, and use of Fund credit and loans.
Many studies have looked at the effect of IMF programs on both the overall BOP and the current account component of the BOP. The IMF mandate to address BOP problems has been clear throughout its history. The Articles of Agreement are explicit that IMF lending should go to countries experiencing BOP problems. The deficit country is taking in more imports or fixed assets or finance than it is generating through exports – the immediate purpose of an IMF arrangement is to provide a loan so that foreign debts can continue to be serviced and necessary imports can be purchased. The loan is intended to soften the blow as adjustments are made as the demand for imports and foreign financing is cut. Demand can be addressed in many ways: devaluation, where the demand for imports is cut by effectively raising their domestic price; the reduction of money supply by raising interest rates or limiting credit creation; and fiscal austerity, where governments reduce consumption both by raising taxes and by spending less. Yet it is not obvious that the IMF program will help. If governments fail to comply with IMF policy conditions, or if the IMF policies are not sufficient, BOP problems can persist. Indeed, if the IMF program causes a drastic contraction of the economy, it is possible for the BOP situation to worsen.
What have studies found? The broad consensus is that the IMF has had success in addressing balance of payments problems. For example, in an early study conducted by economist and professor of Latino Studies at the University of California, Santa Cruz, Manuel Pastor, IMF programs were found to have a positive, statistically significant effect on the BOP, using a before–after methodology analyzing Latin American countries during 1965–81. Another early study using a before–after approach to study Latin America, by Tony Killick, along with colleagues Moazzam Malik and Marcus Manuel, also found a statistically significant positive effect of IMF programs on the BOP. […]
What is the effect of IMF programs on economic development? For some, this is the most important question. Sustainable economic development and prosperity address many of the other economic problems discussed above. An economy that is growing can avoid or afford to sustain BOP and fiscal deficits, and can afford to maintain some degree of inflation. Economic development is also associated with numerous important indicators of quality of life for people. Some argue, however, that economic growth is not and should not be a goal of the IMF. They point out that the original purpose of the IMF was to address balance of payments problems and that the focus on economic growth is something that developed over time. The claim is that the IMF was never intended to promote economic growth.
Yet, this is not completely true. The Articles of Agreement call upon the IMF to provide members “with opportunity to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity” [emphasis added]. This certainly indicates that the IMF should at least not hurt prospects for economic growth.
The Report of the Executive Directors for the First Annual Meeting of the Board of Governors in 1946 was even more explicit:
The function of the Fund is to aid members in maintaining arrangements that promote the balanced expansion of international trade and investment and in this way contribute to the maintenance of high levels of employment and real income.
Right from the first meeting of the governing body of the IMF, high levels of employment and income were central.
Even though the IMF shifted its focus from the industrialized world to the developing world, the importance of promoting national prosperity remained. In fact, the IMF has become increasingly concerned with promoting growth and addressing poverty over time. As Michel Camdessus, the IMF Managing Director from 1987 to 2000, described,
Our primary objective is growth … It is toward growth that our programs and their conditionality are aimed. It is with a view toward growth that we carry out our special responsibility of helping to correct balance of payments disequilibria and, more generally, to eliminate obstructive macroeconomic imbalances. When I refer to growth, I mean high‐quality growth, not … growth for the privileged few, leaving the poor with nothing but empty promises.
Managing Director Horst Köhler, who took the helm at the IMF after Camdessus, emphasized the importance of promoting world financial stability, but he also echoed the views of his predecessor, contending that “the IMF should strive to promote non‐inflationary economic growth that benefits all people of the world.”
How effective has the IMF been at promoting economic growth? Not very. Not only is evidence of growth promotion weak, recent studies even show that IMF programs have a significant negative effect on economic growth. Early studies consistently showed no statistically significant effect. Out of nine before–after studies from 1978 to 1995, covering different countries, regions, and time spans, only one reported a significant positive effect. Four of the others reported no effect; two reported a statistically insignificant negative effect; and one reported an insignificant positive effect. Using with–without comparisons, results were similar – some show insignificant positive effects, others insignificant negative effects, still others show no effect at all, but none of them show a statistically significant effect.
With more sophisticated methodology, new results emerged. Khan’s 1990 study, which addressed nonrandom selection, showed a significant negative effect on growth in the short run, with the adverse effects on growth diminishing thereafter. In his study published in 1994, Conway built upon this result using an advanced technique to control for nonrandom selection on observed variables. He showed that IMF programs have an initial significant negative effect on growth, but a significant positive effect within three years. The take‐away point of Conway’s study is that IMF programs start out badly but end well.
The Conway study had a profound impact. The result made a lot of sense. As IMF economists Nadeem Ul Haque and Mohsin Khan reported in 1998: “In the case of growth, the consensus seems to be that output will be depressed in the short run as the demand‐reducing elements of the policy package dominate. Over time the structural reform elements of the program start to take effect and growth begins to rise.” A subsequent study by IMF economists Louis Dicks‐Mireaux, Mauro Mecagni, and Susan Schadler provided further evidence, showing that ESAF [Enhanced Structural Adjustment Facility] programs from the 1986–91 period appeared to have a statistically significant positive effect on output growth. This study used an advanced methodology to deal with the selection problem. It then went further, however, by testing some of the statistical assumptions underlying the model. They found that many of the assumptions were dubious, and this caused them to raise doubts about the reliability of the statistical findings.
Then a series of studies found a statistically significant negative effect on growth, using similarly advanced statistical techniques. The 2000 study by political scientist Adam Przeworski and me controlled for nonrandom selection on unobserved variables like “political will” and “trust.” The analysis on 79 countries from 1971 to 1990 showed a statistically significant negative effect on annual output growth of about 1.5 percent. Similar results were obtained on a larger sample including 135 countries from 1951 to 1990. No evidence of a long run positive effect was found.
In their 2003 study of Latin America, economists Michael Hutchison of University of California, Santa Cruz, and Ilan Noy of University of Hawaii show that IMF programs have a negative effect on economic growth. In fact, they show that the effect is worse for countries that “successfully” complete programs. This raises an important point [...]: even – indeed, especially – countries that complete IMF programs experience lower growth.
In their study, published in 2005, economists Robert Barro and Jong‐Wha Lee also found disappointing results. Using an instrumental variable approach to address the selection problem, they found that IMF programs have a negative effect in the short run that is not statistically significant, and a strong statistically significant negative effect on economic growth in the long run. This result runs directly counter to the consensus described by Haque and Khan in 1998. Finally, in a study published in 2006 that also uses an instrumental variables approach to the selection problem, economist Axel Dreher further confirms that IMF programs lower growth – his results also deal with compliance [...]. Dreher finds that compliance somewhat mitigates this effect, but even for countries that comply the effect is negative.
So, the newly emerging consensus is that IMF programs hurt economic growth. The initial contractionary effect of IMF programs is really not surprising. Some economists at the IMF have been quite forthright about why. IMF economist Vito Tanzi, for example, has argued that IMF programs induce governments to save on public investment, with nefarious consequences for growth. IMF economists Mario Blejer and Adrienne Cheasty point out that the high real interest rates induce good firms to shut down along with bad ones, which can also hurt growth. Plus, there is the straightforward effect of IMF austerity cutting demand, which drives down economic growth. […]
It is notable that studies of the effect of IMF programs on BOP, budget deficits, inflation, and growth reach different conclusions, depending on the methodology and data employed. This is not so with respect to income distribution. There have been three studies using three different methodologies and three different data sets. All come to the same conclusion: typically, IMF programs exacerbate income inequality.
Pastor conducted the first study in 1987, using the before–after approach to analyze labor’s share of income in Latin America during 1965–81. His conclusion was strong: “The single most consistent effect the IMF seems to have is the redistribution of income away from workers.” Pastor’s study was path breaking, but the early study was limited by the methodology, which did not account for nonrandom selection, and because it looked only at Latin America. These limitations were addressed by a young scholar at Harvard University, Gopal Garuda, who published in 2000 his study of the effect of IMF programs on overall income distribution. Garuda looked at a standard index of overall income inequality called the “Gini coefficient.” He addressed the selection problem by estimating the propensity of countries to participate in IMF programs [...]. Then he compared countries with and without IMF programs that had similar circumstances or “propensities” to participate in IMF programs. One interesting new finding Garuda discovered is that when countries unlikely to participate in IMF programs do participate, income inequality does not increase. However, for countries that are likely to participate, IMF programs exacerbate income inequality. The Garuda study was limited by the small amount of data on Gini coefficients that are available – this is why he incorporated a selection model within a with–without framework. He did not have enough data to employ a standard selection model.
In a study I published in 2002, the limited data problem was resolved by just looking at the manufacturing sector of the economy. The data on labor’s share of earnings from manufacturing are available for 2,095 observations of 110 countries from 1961 to 1993. With these data, a fully parameterized selection model is possible. The result of the study confirmed the two it built upon: IMF programs increase income inequality. […]
Evaluating the effects of IMF programs is analogous to evaluating the effects of medical treatments. If one were to compare the health of people undergoing medical treatment to people not, one might come to the quick conclusion that medical treatments hurt patients, because they are much less healthy than the rest of the population. This is obviously because people only go to the doctor when they are sick. Yet, some medical treatments have been found to be helpful, while others are benign or even malignant. Before coming to such conclusions, one must address the selection problem – under what circumstances is treatment applied?
Researchers have addressed the selection problem when analyzing the effects of IMF programs in various ways with increasing degrees of sophistication. Nevertheless, the conclusions in the literature are tentative. With each generation of studies come new and often contradictory findings.
According to the most recent studies and reviews, the IMF seems to be most effective in addressing balance of payments problems. It is less effective in addressing inflation. And recent studies show pernicious effects on economic growth. IMF programs exacerbate income inequality according to all studies that look directly at this question. In the area of social spending, the most recent study shows that spending on health and education may increase in dictatorships, where little is spent to begin with, but IMF programs make democracies that participate in IMF programs look more like dictatorships when it comes to spending on the poor.