Austerity and the IMF - Kenneth Rogoff

8/3/2019 Austerity and the IMF - Kenneth Rogoff

1/28

Austerity and the IMF

Kenneth Rogof

T h e F i f t h A n n u a l

Richard H. Sabot Lecture

A p r i l 2 0 1 0

T h e C e n t e r f o r G l o b a l D e v e l o p m e n t


2/28


3/28

Te Richard H. Sabot Lecture Series

Te Richard H. Sabot Lecture is held annually to honor

the lie and work o Richard Dick Sabot, a respected

proessor, celebrated development economist, successul

Internet entrepreneur, and close riend o the Center

or Global Development (CGD) who died suddenly in

July 2005. As a ounding member o CGDs Board o

Directors, Dicks enthusiasm and intellect encouraged our

beginnings. His work as a scholar and as a development

practitioner helped to shape the Centers vision o independent research and

new ideas in the service o better development policies and practices.

Dick held a PhD in economics rom Oxord University; he was Proessor

o Economics at Williams College and taught previously at Yale University,Oxord University, and Columbia University. His contributions to the elds o

economics and international development were numerous, both in academia

and during ten years at the World Bank.

Te Sabot Lecture Series hosts each year a scholar-practitioner who has made

signicant contributions to international development, combining, as didDick, academic work with leadership in the policy community. We are grateul

to the Sabot amily and to CGD board member Bruns Grayson or the

support to launch the Richard H. Sabot Lecture Series.

Previous Lectures

2009 Kemal Dervi, Precautionary Resources and Long-erm Development

Finance.2008 Lord Nicholas Stern, owards a Global Deal on Climate Change.

2007 Ngozi Okonjo-Iweala, Corruption: Myths and Reality in aDeveloping Country Context.

2006 Lawrence H. Summers, Harnessing the Development Potential o

Emerging Market Reserves.


4/28

Kenneth Rogof

Kenneth Rogo, one o the worlds leading experts on

global economics, is the Tomas D. Cabot Proessor o

Public Policy and Proessor o Economics at Harvard

University. His treatise, Foundations o International

Macroeconomics, is the standard text used in graduate

courses throughout the world, and his monthly

syndicated column on global economics is published in

over 30 countries.

Rogo served as Economic Counsellor and Director o Research at the

International Monetary Fund rom August 2001 to September 2003 and,

beore that, as Director o the Center or International Development at

Harvard. He also holds the lie title o international grandmaster o chess.

He received a BA rom Yale University in 1975 and a PhD in economics rom

the Massachusetts Institute o echnology in 1980. His most recent book, with

co-author Carmen M. Reinhart, is Tis ime Is Diferent: Eight Centuries o

Financial Folly, an empirical investigation o past nancial crises.

In this years lecture, Te Perils o Financial Globalization and the IMF,

Proessor Rogo will discuss historic vulnerabilities o developing countries

to nancial crises, alternative approaches to mitigating risks and dealing with

crises in the uture, and the past and prospective roles o the IMF and the

G-20.


5/28

Kenneth Rogof| 1

Austerity and the IMF

It is a great honor to present the th annual Richard H. Sabot lecture at the

Center or Global Development. In this talk, I will take up a relatively narrow

but absolutely undamental question in the international monetary system,

particularly in developing countries: Is the International Monetary und (the

IMF) guilty o bringing excessive austerity to the countries that turn to it or

bailout unding? Should the IMF instead put much more weight on encourag-

ing countercyclical scal policy, as it does in rich countries? Extremely dicult

and complex issues underlie these seemingly straightorward questions. My

modest aim in this lecture is to help clariy the issues so as to promote rational

dialogue.

I will argue that the simplest and perhaps most cutting version o the IMF

austerity charges is simply conused. IMF loans typically relieve austerity; theydo not make it worse. IMF support helps a country engage in less procyclical

budget contraction than it might have been orced to do otherwise. Tat said,

the IMFs judgments in calibrating programs involve a huge range o subjec-

tive decisions about politics, psychology, and economics, judgments that are

dicult to get right consistently. oward the end o my remarks, I will argue

that in many respects, the greatest problem with IMF programs is not exces-sive austerity with debtors but excessive generosity toward creditors.

It is a ascinating time to think about the uture o the International Monetary

und in its eorts to help countries deal with the perils o nancial globaliza-

tion. Just a ew years ago, the IMF seemed on deaths doorstep, set to go into

prolonged hibernation i not extinction. Castrated by a ailure to adequately

explain its role in the Asian crisis, and made seemingly irrelevant by the long

s credit boom, the IMF ound itsel virtually beret o borrowers. Even

poor Arican countries shunned IMF money. oday, o course, the world has

been turned upside down. Te historic April meeting o the G- leaders

in London assigned the IMF a pivotal role in stemming the global nancial

crisis and pledged a massive quadrupling o its resources. G- presidents


6/28

2 | Fifth Annual Richard H. Sabot Lecture

urged the IMF to prevent the crisis rom inecting developing countries and

emerging markets. A clear subtext to their message was that world leaderswanted the IMF to be generous and orthcoming with its resources, lend-

ing now and asking questions later. At the same time, the IMF was charged

with taking a more orceul role in global macroeconomic surveillance, and a

supporting (albeit ill-dened) role in nancial market regulation. Now, with

the IMF returning to Europe ater a nearly three-decade hiatus (Portugal

had a program in the early s), an institution that had so recently been

dormant has come to take perhaps the largest role on the world stage it has

ever had.

Yet, even as the IMF is now able to be more generous with its expanded re-

sources, the challenge o persuading countries to reign in their budget decits

to restore condence in international markets is very much a central one.

Ater the nancial crisis, the IMF poured resources into Central and EasternEurope, even into governance-challenged economies such as Romania and the

Ukraine. For the moment, the IMF does not need to call in any o its loans,

but what will happen over the next couple years as the IMF is inevitably orced

to tighten the screws? Will the austerity charge again cripple its eorts to

restore stability? How well will Western Europe deal with the sustained budget

austerity resulting rom the crisis in the euro, particularly across the SouthernCone countries?

Beore turning to my main topic, I will digress briefy in the next two sections

to the role o the Center or Global Development, Richard Sabots vision in

helping ound it, and to my own journey rom macroeconomics to develop-

ment.

Richard Sabot and the Role of the Center for

Global Development

I did not have the privilege o knowing Richard Sabot, but his important

contributions to development research and policy, as well as his role as a

ounder o the Center or Global Development, are well known. Sabot


7/28

Kenneth Rogof| 3

pioneered the exploration o topics that remain at the heart o develop-

ment research today, including the importance o rural-to-urban migration(witness China today), the role o women in successul development, and

also the relation between inequality and growth. Dick developed many o

these ideas in his tenure at the World Bank, where he infuenced a genera-

tion o economists. O course, his vision helped create the Center or Global

Development.

Te Center has a critical role to play in the development policy research nexus.

While universities are a huge resource or producing both ideas and energetic

young students, the research bias at universities is weighted heavily toward

producing academic journal articles. Te greatest rewards in academia are or

infuencing methodology and technique rather than policy. Te World Bank

itsel, o course, produces a great deal o applied research. In addition to spon-

soring outside research, the bank produces a large number o its own studies.Indeed, our years ago, I served on a committee chaired by Angus Deaton and

including Abhijit Banerjee and Nora Lustig that reviewed a decade o World

Bank research. Drawing on meetings, interviews and reports rom more than

two dozen o the worlds leading development researchers, the Deaton com-

mittee produced a report (Deaton et al. 6) that encompassed both the state

o development research in general and the state o development research atthe World Bank in particular.

Te bank has, o course, produced a great deal o important research through-

out its history, and the decade under examination (ending in ) was no

exception. Nevertheless, I was impressed by how many important gaps and

question marks remained in the banks research program. Te bureaucracy also

aces challenges prioritizing resource allocation or creating and (even more

so) maintaining data sets, and in synthesizing diverse results rom around the

bank. With all due recognition o the banks accomplishments, the exercise

highlighted the importance o having independent think tanks such as CGD,

which have a critical mass o scholars and no political agenda to undermine

their credibility.


8/28


A Short Personal Digression on My Introduction to

DevelopmentI came to development economics rom a background as an international

macroeconomist. In the early s, I was working on topics such as empirical

exchange-rate modeling and orecasting. My paper on central bank inde-

pendence (Rogo ) oered a rationale or why countries might want to

delegate monetary policy to an independent central bank that is more ocused

on stabilizing infation than is society as a whole. In a world where most devel-

oping countries were attempting to x their exchange rates, exchange-rate ore-

casting was not a topic on the ront burner. As or central bank independence,

even i rigidity in exchange rates did not constrain the scope or independent

monetary policy, it would have been virtually impossible politically and

institutionally to imagine a meaningully independent central bank in most

developing countries.

How times have changed! Although o course there are diering degrees o

interpretation, central bank independence is now the norm throughout much

o the world. Imagine: India has (to some extent) a fexible exchange rate, and

Brazil has one o the most fexible exchange rates in the world. At the same

time, the central banks in both countries have developed into powerul inde-

pendent institutions. As developing economies have advanced, their macroeco-

nomic problems look increasingly like the ones that aced advanced countries a

ew decades ago.

My rst work that ocused on developing economies, however, consisted o a

series o joint papers with Jeremy Bulow on sovereign debt crises, written in

the second hal o the s. Bulow and I argued that the reigning approach

(in the academic literature) o Eaton and Gersovitz () was unsatisactoryboth theoretically and empirically. At the very least, Eaton and Gersovitzs

narrow ocus on exclusion rom capital markets as the penalty or deault was

elegant but ar too narrow. Later work has by and large supported the view

that, at the very least, one must work with a much broader notion o reputa-

tion and the penalty to deault, one that covers trade, banking, and possibly


9/28

Kenneth Rogof| 5

even national security issues. Perhaps even more importantly, the reputation

view obscures the undamental importance o legal and political institutions,making it virtually impossible to think about policy design. Why have a world

bankruptcy court, as some have proposed, i the whole underpinning or sov-

ereign debt repayment is reputational? (See Rogo and Zettelmeyer [] or

a survey o the literature on international bankruptcy mechanisms.)

Later, my stint as chie economist at the International Monetary Fund rom

to had an enormous infuence on my research. I became immersed

in a range o development issues, learning a great deal rom my colleagues in

the research department and throughout the International Monetary und as

well as rom economists and policymakers around the world. My paper with

Ayhan Kose, Eswar Prasad, and Shang-Jin Wei (Kose et al. 6) attempted to

take a detached view o the empirical literature on international capital market

integration and development, a topic about which there had been a high ratioo polemics to act. We ound the mass o literature to be surprisingly ambigu-

ous with respect to growth and macroeconomic volatility, which we attributed

in later work to the act that the indirect eects o capital market integra-

tionboth positive and negativeswamp the eects o the traditional capital

accumulation, consumption smoothing, and diversication eects emphasized

in the literature.

It was also during this period that I began a long ongoing collaboration with

Carmen Reinhart. Since leaving the IMF in , Carmen and I have been

doing research on the history o international nancial crises going back

almost years. Our recent book, Tis ime Is Dierent: Eight Centuries o

Financial Folly, shows the universality o nancial crises across time and space.

Te analysis o the book suggests that it was indeed olly in the mid-s to

predict the demise o the IMF as an organization that was no longer needed

because there would never again be a major sovereign debt crisis.

Tanks to the G-, the IMF has the capacity to hand out more and larger

loans than ever. o be eective, however, the IMF ultimately has to be able to


10/28


help countries live within their means, at least on a long-term basis. Its eorts

to help spendthrit governments reign in their budgets has oten led to thecharge that it mindlessly preaches austerity with little regard to social conse-

quences. It is this core topic that I wish to take up in todays lecture.

The IMF and the Austerity Charge

Perhaps the leading charge against the IMF over its long history is that it

invariably imposes harsh austerity conditions on hapless developing countries

in their time o greatest need. Over the past three decades the world press

has been replete with images o mass demonstrations against IMF programs,

with ood riots against price hikes and shortages being a particularly compel-

ling graphic. Although the IMF has been accused o many mistakes over the

years (most amously perhaps by Joseph Stiglitz []), the austerity charge is

the one that cuts most deeply at the IMFs moral authority and ultimately its

power and eectiveness. Is the austerity charge air?

Te short answer is that the IMF almost always makes austerity or a bankrupt

country lighter than it might otherwise have aced. Te supercial view that

the IMF creates needless austerity is just nave and wrong. However, even as

the IMF relieves austerity, a great deal o judgment is required in calibrating

how much and or how long the IMF should help, and under what condi-tions. Te deeper question o how the IMF should design and calibrate its

rescue programs to relieve austerity is a legitimate one over which views on

best practice are constantly evolving.

raditional anti-IMF rhetoric is so eective that most people are shocked

when they are told that IMF programs usually relieve austerity rather than

make it worse. How can that be, they will ask, given so many press articles

and so much political rhetoric reinorcing the populist view that the IMF puts

budget balance and the interests o oreign banks ahead o the welare o or-

dinary people? Straightening out the simplest misconceptions about the IMF

and austerity is an essential starting point to any rational conversation about

what the real issues are.


11/28

Kenneth Rogof| 7

Te basic point is that, in most cases, the International Monetary und is e-

ectively a lender o last resort. Countries typically go to the IMF when otherlenders have turned their backs, and there is nowhere else to turn. Unlike pri-

vate lenders, the und is a quasi-UN organization that draws on money rom

governments around the world rather than private lenders.

Te circumstances o IMF lending programs vary, but certainly the most

common include a government, ater many years o running large decits

borrowing heavily rom the private sector, nding itsel unable to roll over

its old debts as they mature (unable to nd new loans to cover old principal

payments as they come due). At the same time, o course, the government is

unable to nance ongoing budget decits. Tere are many nuances, depend-

ing or example on the extent o exposure to private sector oreign creditors,

but the twin decits o government budget decits and current account

decits is the canonical predicament. In the typical case, whether triggered bya shit in international capital markets or the domestic economy, the country

experiences a Dornbusch-Calvo sudden stop o nancing fows. Perhaps the

world price o the countrys primary commodity export has suddenly plum-

meted. Perhaps a political upheaval has caused creditors to doubt whether

the country will continue on its old growth path. Perhaps a shock to world

markets stemming rom a sudden monetary tightening or risk event aectingthe rich countries has caused a sudden retreat rom risky lending abroad. (For

example, the Volker monetary tightening in the early s, together with a

concomitant drop in global commodity prices, was a key trigger o the Latin

American debt crisis o the s. See Bulow and Rogo [] or Reinhart

and Rogo, [].)

One way or the other, the unding spigot is shut o and the country nds it-

sel in deep nancial trouble. Absent an international lender o last resort, not

only would a country likely be orced to deault on its existing debt, it would

be unable to continue sustaining trade balance (current account) decits o

any kind. A country cannot buy more than it sells unless someone is willing to

lend it money. Not every country on the brink o deault is running an ongo-


12/28


ing decit (sometimes it simply cannot renance its existing debt), but that is

the most common situation. Examples include Argentina in and mucho Eastern Europe in the run-up to the nancial crisis.

Here is the key point. In the ace o ongoing decits, a sudden stop in inter-

national lending implies that citizens would eel the eects o a tightening belt

even without an IMF program. And that does not include any penalties lend-

ers might impose or deaulting, such as making it dicult or exporters to

access the bank trade credits that are the lieblood o international trade. Tat

is precisely why, in the canonical case, a country turns to the IMF.

When the IMF comes in, it typically makes a country a bridge loan at an

interest rate ar below what it could imagine getting on international markets,

even prior to the crisis. Tat is, the IMF usually charges an interest rate only

slightly greater than the rate that triple-A borrowing countries such as theUnited States and Germany pay. In principle, the IMF bridge loan allows the

troubled debtor time to adjust to the shit in international lending, to sustain

essential services, and to restructure nances or an eventual return to interna-

tional borrowing. Again, one must underscore the point that the IMF does not

impose itsel on a country; it comes in only by invitation. A country opens its

doors to the IMF (admittedly usually quite reluctantly) precisely because thegovernment knows that an IMF bridge loan can help cushion austerity.

Tus, as I have argued in the past (e.g., Rogo , which restates my review

o Stiglitz ), the act that the IMF is oten ound near the scene o sharp

budget cutbacks does not mean that it is the primary cause. Doctors can be ound

treating plague victims; this does not mean doctors cause plagues. Fire engines are

ound near res; this does not mean re engines cause res. Te simplistic populist

notion that the IMF is the root cause o austerity in a crisis is utterly nave and

wrong, despite the huge success o many polemicists in propagating this view.

For the most part, nancial crises are caused by countries mismanaging their

own nances, at the very least taking undue risks during boom times (in the


13/28

Kenneth Rogof| 9

spirit o the title o my book with Reinhart). Whereas oreign lenders

are not necessarily blameless, in the rst instance it is the debtor countrys pastgovernments and perhaps current government that are responsible. O course,

there oten is a huge desire to pin the blame elsewhere. In act, one o the

IMFs primary roles in a crisis is oten to serve as a lightning rod or domestic

unrest, thereby giving the government cover to implement austerity measures

that would, in act, have been harsher in the IMFs absence.

The Difficult Judgments in Designing IMF Programs

to Relieve Austerity

Dispensing with the usual populist argument against the IMF, however, is only

a starting point or evaluating IMF programs. Te act that the IMF does not

make austerity worse by no means implies that it always does the best possible

job calibrating its relie programs. Tis is a much more complex issue involv-

ing a myriad o subtle judgments. I will attempt to give only a favor o thearguments here.

Te rst point is that the IMF is in the business o making bridge loans only,

usually until a country can return to international credit markets. Sometimes

getting repaid is not always a reasonable expectation when dealing with the

poorest countries in Arica and elsewhere. Te IMF has on occasion beenpersuaded to eectively write o part o its loan, but this applies to only a very

small raction o its portolio. For the most part, the IMF is a revolving credit

acility that relies on its de jure seniority in international lending markets to

get repaid in ull even when other creditors are orced to take haircuts on

their loans. Te World Bank, with its greater fexibility to make concessional

loans and ar larger sta to give basic development advice, generally takes the

lead on the poorest countries. Te IMFs meat and potatoes are richer middle-

income countries and sometimes upper-middle-income countries. Te IMF

needs to be repaid on these huge loans, which sometimes reach $ billion

or more. I countries consistently deaulted on the IMF, then the IMF itsel

would nd itsel unable to pay the creditor countries who lent it money. Te

IMF itsel would be bankrupt.


14/28

10 | Fifth Annual Richard H. Sabot Lectur e

Te length o IMF bridge loans is not set in stone. Te terms are to some

extent xed by IMF rules, but setting the length is the rst o many subtlejudgments that the IMF and its overseers must make. In the past, the und

would typically aim to be repaid within two or three years, sometimes longer

in special cases, oten issuing new loans to pay back the old ones. Ater the

recent nancial crisis, the G- leaders have arguably given the IMF the green

light to stretch out its bridge loans slightly urther (say to ve years) on the

grounds that the crisis will likely have long-lasting eects and that private

credit markets may not have strengthened suciently to replace IMF unding

within the usual time window.

Tere are, o course, no hard and ast rules about the traditional implementa-

tion period, or the extended one. Te und has learned over the years that a

country must be given a signicant period o time to adjust, but at the same

time there must be some pressure or sustained adjustment. As a matter opolitical expedience, backloading adjustment is risky because it makes it much

harder or a government to reach the nish line o a program. Ater all, the

ultimate objective is to bring a governments nances under sucient control,

with sucient credibility that private investors are willing to reenter. Also, o

course, the typical business cycle downturn ater a nancial crash lasts a little

over eighteen months and sometimes signicantly longer (Reinhart and Rogo). A country is much more likely to be able to return to private markets

ater its economy starts growing again, making debt sustainability calculations

less grim. Still, calibrating the adjustment period and the speed o adjustment

is a judgment call.

Te und must also decide how large its package needs to be. I the sum is too

large or the IMF alone, as it sometimes is (e.g., in the recent Greek nancial

crisis), the und oten works together with a consortium o wealthy indi-

vidual country lenders. ypically, at least one nancially strong country has a

pronounced interest in helping the debtor country. Tis was what happened,

or example, when the United States helped Mexico during its mid-s

nancial crisis.


15/28

Kenneth Rogof| 11

In determining the size and duration o a loan package, there are two urther

crucial issues. First, to what extent can the und persuade private creditorsto continue extending loans, thereby alleviating the its need to pay o the

countrys loans as they come due? Te question o private sector involvement

is a very contentious one. In principle, the IMF could bully private lenders

into signicant concessions because they might well be orced to make larger

ones without IMF involvement. Te und, however, also takes into account

potential costs to the countrys long-term credit standing as well as disruption

to international credit markets more generally. My own strong bias is that the

IMF gets ar too easily sucked in to bailing out private creditors, but we will

return to this issue later when we are ready to discuss moral hazard more ully.

Suce it to say that determining the extent o private sector involvement is a

key decision the lender o last resort invariably aces.

A second issue is how large a decit the country will be allowed to run and orhow long. Te IMF aims to relieve austerity and orestall procyclical scal poli-

cy, but by how much and or how long? Again, there are very subtle judgments

to be made about economics, politics, and market psychology. I adjustment is

very slow, will it be credible to markets, and will politicians be able to sustain

it? Te und knows that governments very oten backslide on programs, some-

times badly missing budget targets. I the planned trajectory is unambitiousto start with, what happens when the government underperorms? Critics

may rightly argue that in some cases the IMFs trajectory is too harsh, that a

soter, slower trajectory would have been just as successul in restoring market

credibility, perhaps more so i it had been politically easier to digest. But these

are enormously subtle judgments, oten very dicult to calibrate in the heat o

battle, or even ater the act.

In addition to politics, the und oten has to be concerned whether it is getting

reliable gures. In its everyday surveillance activities the und makes a great e-

ort to help countries, and in some cases to persuade them, to provide reliable

budget gures on an internationally comparable basis. Tis is an extremely

dicult task, even in the richest, most-developed countries. Budgets are both


16/28


very political and very technical. Governments have both incentive and means

to distort gures, not just or oreign investors but also or local consump-tion. For example, in many countries, the government eectively controls a

signicant share (up to percent or more) o the overall economy, through

quasi-government agencies, national utility companies, and others. Particularly

in the case o governments that are under pressure rom international mar-

kets, this can create enormous incentive to use delays in payments o various

orms to make the short-term budget picture seem better than it is. Tis might

involve delays o payments to national utilities or resources consumed or,

very requently, delays o payments to civil servants. Tis is part o the reason

Reinhart and I nd that debt explodes in the years ater a nancial crisis, as

these hidden debts come fying out o the woodwork. Some o these devices

are relatively easily detected even by the IMFs very small sta on the ground.

Other tricks involve deeper camoufage and are harder to uncover.

A classic device that remains very popular today even in rich countries is

to sell buildings and sign long-term leases to rent them back. Suppose, or

example, the postal service owns a building it uses. It will sell the building but

at the same time sign a ninety-nine-year lease to rent it back. On the books,

the sale will show a big gain, whereas the oset rom the rst years annual

rent will be relatively small. On paper, there is a big improvement in thebudget, but rom an economic perspective, there is very little or no change. A

related issue arises, o course, with privatization o large national companies.

Te government gets a lump sum payment in the short run but oten loses a

signicant stream o revenue rom the company or utility in the long run. Te

IMF oten encourages such sales in cases where it perceives the possibility o

large eciency gains, that is, i the private sector will run the company more

eciently. Te same gain appears on the governments balance sheets, but the

desired eciency gains are not always orthcoming, especially i there is no

eective regulation or antimonopoly legislation.

Te IMF must also judge the countrys political capacity to absorb austerity

measures. When countries experience sovereign debt and repayment prob-


17/28

Kenneth Rogof| 13

lems, social unrest and government turnover is requently the norm. Te und

must put together a program that will be credible with markets. One that isobviously politically unsustainable will not be credible. Again, a great deal o

judgment is involved.

ypically, a key element o persuading markets that a countrys nances are

sustainable involves demonstrating that its debt-to-income ratio, typically

soaring in the atermath o the nancial crisis, will eventually start alling.

Sustainability analysis is extremely tricky in IMF programs since, typically, the

programs call or the debt-to-income ratio to rise in the early years. Initially,

growth is usually slow or negative, the countrys exchange rate usually declines

(raising debt-to-income ratios to the extent debt is denominated in oreign

currency), and the und typically allows the country to run decits, oten

even on the primary decits, that is, decits excluding interest payments on

debt. Greeces recent EU/IMFdesigned austerity program, where debt beganat approximately percent o GDP and is expected to rise to percent

o GDP, is a case in point. Tat such programs are dicult to sustain is not

surprising. Reinhart, Rogo, and Savastano () and Reinhart and Rogo

() demonstrate the ragility o sustainability analysis.

Nevertheless, the central importance o a countrys debt-to-income trajectoryunderscores the importance o growth as a central goal o und programs. Aus-

terity helps the numerator o a countrys debt-to-income ratio but can impede

the growth o the denominator (income) as well. Again, complex judgments

are involved. Clearly, to the extent the und can help the debtor country insti-

tute supply side reorms (e.g., creating more eciency in its tax and pension

systems, reducing monopoly and corruption), it is more likely to succeed in

establishing a credible program. Supply side reorms are even more delicate po-

litically than austerity measures in many cases, and the und can only aord to

become involved in cases where there is no other approach. In the case o the

Asian nancial crises, the und was accused o having gone too ar

in speciying detailed supply-side reorms. In the recent cases in Eastern and

Central Europe, the und has tried to scale back conditionality on supply side


18/28


and structural reorms. Tis helps the und keep a low prole in the short run,

but o course may cause problems in the longer run, as supply side reorms areessential to achieving the growth necessary or long-term sustainability o these

countries massive decits. In the case o urkey in the s, the und seemed

successul in being a catalyst or supply-side reorms that the government

aimed to achieve without being accused to the same degree as in Asia o being

too involved with details.

I have ramed the entire debate as having to do with how much, and or how

long, the IMF can and should relieve austerity by its programs. For complete-

ness, it is worth noting that the debate is oten ramed in terms o whether the

IMF should endorse Keynesian countercyclical policy in the ace o the inevi-

table recession that accompanies a sovereign debt crisis (see Te Economist

). Tat is, should the country run countercyclical scal policy to ght the

eects o the recession; in other words, should it run large decits. Te readerwill already recognize that at one level this argument is absurd. I a country

were in a position to run large countercyclical decits, it would not be at the

IMFs doorstep. Countries come to the IMF precisely because there are no

other lenders. At a deeper level, there is a legitimate point that i the country

cuts expenditures too drastically too quickly, it risks dramatically deepening

its recession and destabilizing its economy. In evaluating the extent to whicha country should be lent money to run ongoing decits, the IMF can and

should take the eects on output into account. O course, there is great un-

certainty, with academic research on Keynesian scal multipliers by and large

yielding very mixed results. We will return to Keynesian countercyclical policy

in our discussion o the IMF policy towards austerity in richer countries.

In sum, although polemicists have been extremely eective in pinning the

austerity moniker on the IMF, the nave view that the und is the cause o aus-

terity is wrong. So too is the nave view that the IMF should be encouraging

Keynesian countercyclical scal policy in bankrupt countries. Tis is not to say

that the IMFs job o lessening austerity and rescuing countries rom having

to contract decits too quickly is easy, or that it always gets it right. Tere are


19/28

Kenneth Rogof| 15

indeed a myriad o dicult judgments that involve not only macroeconomics

but politics, sociology, and psychology. Tese judgments are dicult to boildown to a simple ormula, and it is perectly reasonable to second guess the

unds choices and calibrations, both or the sake o improving existing pro-

grams and or better designing uture ones. But the simplistic polemics about

excessive austerity and ailure to understand Keynesian countercyclical scal

policy are, by and large, vacuous.

Is the Real Problem with IMF Program Generosity

(with Creditors)?

In my mind, the big question about IMF programs is whether they are ar

too biased toward bailing out creditors, stoking moral hazard and sowing the

seeds o uture crises. Tat is, does the und design its programs with a view

to making sure a country is able to pay o its creditors, particularly oreign

creditors? One obvious interpretation o this bias, o course, is simply that theund places a heavy weight on the interests o creditors, perhaps because its

voting structure gives such large weight to the rich countries who, at least over

the past three to our decades, have by and large always been lenders rather

than borrowers. Foreign lenders, both bond and bank lenders, typically come

disproportionately rom rich countries as well.

Now, to some extent, the view that the und takes into account rich-country

creditors is correct. But the und would claim a deeper rationale or promoting

ull repayment o debts, based on the view that ensuring creditors are repaid is

ultimately in both the debtor countrys interest and in the broader interest o

the global nancial system. Both arguments have merit but do not necessarily

withstand a considered cost-benet analysis in many cases.

Why might it be in the countrys interest to repay its creditors? As Eaton

and Gersovitz () amously argued, one o the reasons a country pays its

debt is to preserve its reputation or repayment, thereby preserving uture

access to capital markets. As noted earlier, Bulow and Rogo () argue

that countries ace legal obstacles to trade nance and other problems i in


20/28


deault. In either case, countries suer important penalties to deault. But

as Bulow and Rogo show in their bargaining-theoretic ramework, and asGrossman and Van Huyck () show in a generalization o Eaton and

Gersovitzs reputation model, the act that the debtor may be pressed into

paying something does not mean it should always do everything possible to

pay in ull, regardless o consequences or stability and welare. In act, it can

be in creditors interests to recognize that when they do not have the leverage

to collect ull repayment, they might do better by accepting partial repay-

ment.

Both theory and the long history o sovereign debt crises beore the IMF

suggest that creditors let to bargain on their own without expectations o a

bailout might well allow a country in diculties to deault partially. Indeed, as

Ozler () and others have shown, countries oten return to bond markets

surprisingly quickly ater a partial deault, with only modestly higher risk pre-mia than they might have paid otherwise. Te notion that deault is necessarily

a disaster or a country is no more simple or straightorward than the austerity

argument we dispensed with earlier.

O course, having one country deault might have destabilizing eects more

broadly on the international monetary system and these eects must be takeninto consideration. But again, the risk o contagion is very dicult to assess;

the international monetary system has withstood literally hundreds o indi-

vidual country deaults in over the course o history, and it is hardly the case

that every country deault has proved systemic.

Perhaps the greatest argument against the unds strong bias towards bail-

ing out creditors is the moral hazard argument modeled in Bulow and

Rogo ( and ), a decade beore the amous Meltzer report (Melt-

zer et al. ) popularized the idea. I the und consistently helps bail out

creditors, they will not charge suciently high risk premia or risky loans.

Countries will borrow too much and get into trouble too oten. Bailouts

may help put a check on the allout rom current crises but lay the seeds


21/28

Kenneth Rogof| 17

o uture ones. In my view, the moral hazard argument is a very powerul

reason or caution in bailouts i not even or trying to end them entirely.Teoretical academic research certainly suggests that working out an inter-

national bankruptcy system could go a long way toward substituting or

the unds bailout mechanism (Rogo and Zettelmeyer ). Instead, the

current system eectively subsidizes debt at the expense o other orms o

investment such as equity.

I have argued this point or a very long time, prior to my days at the und

in the early s. For example, Bulow and Rogo () asked whether

policy and legal institutions are tilted altogether too much in avor o the

bond market over equity markets and oreign direct investment. (Tis

point is similar but distinct rom the writings o Bhagwati, Rodrik, and

Stiglitz, who all complain o short-term capital fows in general.) It is also

consistent with the empirical work o Peter Henry and others that equitymarket integration is ar more clearly benecial to growth than bond mar-

ket integration. In general, legal, tax and institutional orces all conspire to

avor debt lending.

In some sense, then, the problem is not so much that the und is uniormly

too stingy with countries, though the judgments inherent in virtually everyprogram can be argued on a case-by-case basis. Te problem is that IMF policy

is too generous toward creditors.

One nal point: the IMFs arrival means that bond holders are o the hook.

As Qian, Reinhart, and I document (Qian, Reinhart, and Rogo orthcom-

ing), there have been numerous instances where countries enter IMF programs

but end up deaulting anyway. Te most amous case is Argentina (),

but other recent examples include Indonesia, Uruguay, and the Dominican

Republic. Te IMF can oer country the opportunity and incentives to repay

its loans, but it is by no means always successul. Te endgame could be the

same or some o todays highly indebted European countries (See table , next

page).


22/28


The IMF and Countercyclical Pol icy in RichCountries

Lastly, we come to the unds approach to countercyclical scal policy in devel-

oped countries that have strong open access to international capital markets.

Here the unds advice is suitably more nuanced. During growth periods, the

und typically encourages countries to run smaller decits or even surpluses,

in anticipation o possibly needing to run countercyclical scal policy during

a downturn. Te und has a urther agenda, typically encouraging countries

to coordinate countercyclical policy during a recession; without coordination,

countries would not internalize the spillover eects o each others actions.

Arguably, the IMF overestimates the Keynesian benets o countercyclical

scal policy, rather than ignoring them as it is oten accused o in developing

countries. In addition, the und worries, o course, about excessive imbalances

in current accounts between countries, a topic Lawrence Summers took up inhis Sabot Lecture three years ago.

Conclusions

Over the years, the IMF has morphed many times, including to enlarge its

role since the recent nancial crisis. But the austerity tag has stuck throughout,

Table : IMF progr ams have ofte n not been enough

to keep count ries from ser ial de faultEx ample s of IMF progra ms preceding default

Year(s) of IMF program(s) Year(s) of Default(s)

Argentina

Indonesia , ,

Uruguay Dominican Republic

Venezuela

urkey

Source: Data rom Qian, Reinhart, and Rogof (orthcoming)


23/28

Kenneth Rogof| 19

sometimes undermining the unds infuence and eectiveness. I have argued

here that the most nave polemic argument that the und presses austerity isexactly thatnave. In act, IMF programs typically relieve austerity, although

calibrating and designing them is an extremely complex process that involves

a myriad o judgments about politics, credibility, and economics, judgments

that are o course air game or criticism and analysis.

I there is a systematic faw in IMF programs, it is not necessarily excessive

austerity policies toward debtors, but its generosity toward creditors. It argu-

ably overestimates the costs to a country o a negotiated deault and, more

importantly, underestimates the long-term moral hazard problems created by

recurrent bailouts. I the international nancial system o the st century is to

be less prone to crises, a good starting point would be to work toward insti-

tutional changes to redirect nancial fows toward equity and oreign direct

investment, away rom debt instruments. In this regard, an InternationalMonetary Fund, with its huge pool o lendable resources and its arguable bias

toward ensuring repayment, is problematic. From this perspective, it is the

unds generosity, not its austerity, which is ultimately the problem.


24/28


References

Bulow, Jeremy, and Kenneth Rogo. . Multilateral Negotiationsor Rescheduling Developing Country Debt: A Bargaining TeoreticFramework. International Monetary Fund Staf Papers(4): 644.

. . A Constant Reconstructing Model o Sovereign Debt.Journal o Political Economy:.

. . Cleaning Up Tird World Debt without Getting aken to theCleaners,Journal o Economic Perspectives4:4.

Deaton, Angus, et al. 6. An Evaluation o World Bank Research. Washington: World Bank. Available at www.princeton.edu/~deaton/downloads/An_Evaluation_o_World_Bank_Research_-.pd

Eaton, Jonathan, and Mark Gersovitz. . Debt with PotentialRepudiation, Teoretical and Empirical Analysis. Review o EconomicStudies4(): .

Te Economist. . A Stimulating Question: Can Emerging EconomiesNow Aord Counter-cyclical Policies? Te Economist, December ,, .

Grossman, Herschel, and John Van Huyck. . Sovereign Debt as aContingent Claim: Excusable Deault, Repudiation and Reputation,American Economic Review(): .

Henry, Peter. . Capital Account Liberalization: Teory, Evidence, andSpeculationJournal o Economic Literature4(4): .

Kose, Ayhan, et al. 6. Financial Globalization: A Reappraisal.International Monetary Fund Staf Papers6(): 6. (Earlier versionsappeared as IMF Working Paper WP/6/ [December 6]; andNBER Working Paper No. 44 [August 6].)

Meltzer, Alan, et al. . Report o the International Financial InstitutionAdvisory Commission. Available online at www.house.gov/jec/im/meltzer.pd.

Ozler, Sule (), Have Commercial Banks Ignored History?AmericanEconomic Review (June): 6.

Qian, Rong, Carmen Reinhart, and Kenneth Rogo. Forthcoming. OnGraduation rom Deault, Infation and Banking Crises: Elusive orIllusion? In NBER Macroeconomic Annual 2010. University o ChicagoPress.

Reinhart, Carmen M. and Kenneth S Rogo. . Tis ime Is Diferent:


25/28

Kenneth Rogof| 21

Eight Centuries o Financial Folly. Princeton University Press.

Reinhart, Carmen M., Kenneth S. Rogo, and Miguel A. Savastano. .Debt Intolerance. Brookings Papers on Economic Activity:4.Rogo, Kenneth. . Te Optimal Degree o Commitment to an

Intermediate Monetary arget. Quarterly Journal o Economics:6.

. . Te IMF Strikes Back. Foreign Policy4:46.Rogo, Kenneth, and Jeromin Zettelmeyer. . Bankruptcy Procedures or

Sovereigns: A History o Ideas, 6-. International Monetary Fund

Staf Papers4(): 4.Stiglitz, Joseph. . Globalization and its Discontents. New York: W.W.

Norton.


26/28


27/28


28/28

Independent research andpractical ideas for global prosperity

www.cgdev.org

1800 Massachusetts Ave NW

Documents

Austerity and the IMF - Kenneth Rogoff