Fiscal policy in emerging markets: procyclicality and graduation.
Vegh, Carlos A.
Five key questions have guided my research on fiscal policy in
emerging markets: (1)
1. How is fiscal policy conducted in emerging markets compared to
industrial countries?
2. Why has fiscal policy often been procyclical in emerging
markets?
3. Are there developing countries that have
"graduated"--that is, switched from being procyclical to
countercyclical?
4. Has fiscal policy been an effective countercyclical tool?
5. Is the recent experience of some eurozone countries reminiscent
of past fiscal behavior in emerging markets?
This summary describes the main findings that have resulted from
this research agenda. In pursuing these issues, I have been very
fortunate to work with many talented co-authors, whose many
contributions will hopefully become clear below.
Fiscal Policy in Emerging Countries: When It Rains, It Pours
Figure 1, on the next page, shows the correlation between the
cyclical components of real GDP and government spending for 96 countries
(21 industrial and 75 developing) for the period 1960-2014. (2)
Industrial countries are denoted by gray bars while blue bars represent
emerging countries.
The visual impression is striking: With only two exceptions, Greece
and Portugal, all grey bars lie to the left of the graph, indicating a
negative correlation and hence countercyclical government spending in
industrial countries, while 81 percent of blue bars lie to the right of
the graph, indicating a positive correlation and hence procyclical
government spending in developing countries. In fact, the average
correlation for industrial countries is -0.23, compared to 0.21 for
developing countries. Both estimates are significantly different from
zero at the one percent level.
Although much less documented --mainly because data on tax rates
are much harder to come by--the same is true of tax policy. Based on a
novel annual dataset that comprises value-added, corporate, and personal
income taxes for 62 countries (20 industrial and 42 developing) for the
period 1960-2013, Guillermo Vuletin and I have concluded that tax policy
has been acyclical in industrial countries and mostly procyclical in
developing economies. (3) By procyclical tax policy, we mean that the
correlation between the cyclical components of tax rates and GDP is
negative; that is, it reinforces the business cycle.
The evidence thus strongly suggests that, unlike industrial
countries, developing countries have historically pursued procyclical
fiscal policy both on the spending and the revenue side. During bad
times, with capital flowing out and the economy mired in recession,
policymakers have often compounded the problem by contracting fiscal
policy.
Why has Fiscal Policy been Procyclical in Emerging Markets?
A natural question is why policymakers in developing countries
exacerbate already pronounced boom-bust cycles by pursuing procyclical
fiscal policy. This has been a puzzle in search of an explanation. The
two most convincing explanations are arguably that they have limited
access to international credit markets in bad times, and that political
incentives and institutional weaknesses tend to encourage
"excessive" public spending in good times. (4)
[FIGURE 1 OMITTED]
These two channels have in fact reinforced one another in bringing
about procyclical fiscal policy. Emerging countries' inability to
borrow in bad times--often in conjunction with calls for "fiscal
consolidation" from international creditors and organizations--has
typically left them with little choice but to cut spending and raise
taxes in the midst of severe recessions.
This situation has only been made worse by the tendency to save
little, if any, during temporary booms fueled by surges in commodity
prices and capital inflows. Time and again, policymakers have insisted
that good times were here to stay and spent accordingly. Spending
proceeds that are temporary in nature as though they were permanent
naturally forces governments to contract spending and raise taxes in bad
times to satisfy the intertemporal budget constraint (or, alternatively,
default). Put differently, the textbook recommendation of saving on
sunny days for rainy days has been seldom, if ever, followed in emerging
markets.
[FIGURE 2 OMITTED]
Graduation
Fortunately, fiscal policy is not an immutable phenomenon and
changes in market access and domestic financial institutions have
enabled many developing countries over the last 15 years to switch from
being procyclical to acyclical or even counter-cyclical, a phenomenon
dubbed "graduation" in my work with Jeffrey Frankel and
Vuletin. (5)
To see how fiscal policy cyclicality has evolved over time, Figure
2 shows, for each of the 96 countries in Figure 1, the correlation
between real government spending and real GDP for the periods 1960-1999
and 2000-2014. (6) By so doing, the plot is divided into four quadrants:
Established graduates (bottom-left): countries that have always
been countercyclical. Not surprisingly, 74 percent of these countries
are industrial, including the United States and the United Kingdom.
Still in school (top-right): countries that were originally
procyclical and continue to be so. Not surprisingly, 95 percent of these
countries are developing. A notable country in this group is Greece,
which in fact has become much more procyclical since the year 2000, with
the correlation increasing from 0.09 to 0.76.
Back to school (top-left): countries that were countercyclical but
then turned procyclical.
Recent graduates (bottom-right): countries that used to be
procyclical but have become countercyclical over the last 15 years.
Twenty one out of the 24 graduating countries (88 percent) are
developing countries. The overall graduation rate for developing
countries is 34 percent. As a result, the proportion of developing
countries that are procyclical has fallen from 81 percent to 65 percent.
The poster-boy of the graduation movement has clearly been Chile.
Between the two periods, Chile's correlation switched from 0.25 to
-0.68. In fact, Chile's fiscal stimulus package of close to three
percent of GDP in response to the global financial crisis of 2008-2009
was among the largest in the developing world.
[FIGURE 3 OMITTED]
The key to Chile's graduation was the adoption in the year
2001 of a fiscal rule that requires the government to run a structural
balanced budget. (7) The structural balance is computed by adjusting the
actual balance for the effects on tax revenues of deviations of actual
output from trend output and of deviations of copper prices from their
long-run value. These trends are based on forecasts produced by an
independent group of experts. By construction, a zero structural balance
forces the fiscal authority to save in good times and allows it to spend
in bad times.
Needless to say, fiscal rules are not a panacea and even Chile
broke its own rule in 2009 when, as a result of the stimulus package in
response to the global financial crisis, it ran a structural deficit of
1.2 percent. But clearly fiscal rules can be helpful as a guide to sound
fiscal policy and, when based on the structural fiscal balance, in
drawing the market's attention to the need to adjust for the
business cycle when evaluating current fiscal policy.
Even more important perhaps is the overall improvement in the
quality of fiscal institutions, including transparent budgetary
procedures, fiscal accountability, and broad agreement on fiscal
priorities. A structural fiscal rule a la Chile should be viewed as an
improvement in fiscal institutions. In fact, the empirical evidence
clearly suggests that improvements in the quality of institutions lead
to more countercyclical fiscal policy. (8)
How Effective is Counter-Cyclical Fiscal Policy?
We have established that about a third of developing countries have
graduated. This has brought the number of developing countries that have
pursued countercyclical fiscal policies over the last 15 years to 35
percent from just 19 percent in the period 1960-1999. The next question,
then, is: How effective has countercyclical fiscal policy been?
The size of the fiscal multipliers has, of course, been a perennial
question for the United States and, to a lesser extent, other industrial
countries. Until quite recently, however, the evidence for emerging
countries had, at best, been scant, due to lack of reliable quarterly
data. Estimates based on annual data are dubious simply because the main
identification mechanism--the BlanchardPerotti assumption that
government spending can react to GDP with only one period lag--strains
credibility when applied to annual data.
Ethan Ilzetzki, Enrique Mendoza, and I put together a novel
quarterly dataset for government spending for 44 countries (20
industrial and 24 developing) from the first quarter of 1960 to the
fourth quarter of 2007. Often "quarterly data" is simply
interpolated from annual data, so we went to great lengths to ensure
that only data originally collected on a quarterly basis was included.
(9) Perhaps our most important finding is that the size of fiscal
multipliers seems to depend critically on country characteristics such
as exchange rate regime and level of debt. In particular --as
illustrated in Figure 3--we find that the fiscal multiplier is
relatively large in economies operating under fixed (or, more generally,
predetermined) exchange rates, but is indistinguishable from zero under
flexible exchange rates. We also show that, on impact, the fiscal
multiplier is zero in economies with debt exceeding 60 percent of GDP,
presumably reflecting the belief in global capital markets that any
fiscal expansion is simply unsustainable.
As Alan Auerbach and Yuriy Gorodnichenko have shown for OECD
countries, another critical determinant of the size of the fiscal
multiplier is the stage of the business cycle, with the fiscal
multiplier being larger in recessions than in booms. (10) Moreover, in a
study on OECD countries, Daniel Riera-Crichton, Vuletin, and I have
shown that whether government spending is increasing or decreasing
matters as well. We find that the linear (or single) multiplier after
four six-month semesters is 0.40, rises to 1.25 if computed for
recessions (in line with Auerbach and Gorodnichenko), and to 2.3 when
computed for recessions and government spending going up. Intuitively,
the bias arises because government spending has a larger effect on
output when it increases than when it decreases and, even in OECD
countries, there are many instances in which government spending falls
in recessions, which biases downward the "true" multiplier.
(11)
Since emerging markets in particular face repeated crises, it is
also important to look at fiscal policy responses during crises--as
opposed to cyclical characteristics over the regular business cycle--and
see how they have evolved. Vuletin and I have looked at fiscal policy in
the midst of crises for seven Latin American countries accounting for
more than 90 percent of the region's GDP over the last 40 years and
concluded that countries such as Chile and Mexico have been able to
switch from procyclical to countercyclical fiscal policy responses. (12)
But the picture is uneven, as countries like Argentina, Uruguay, and
Venezuela continue to show a pronounced tendency to contract government
spending sharply in recessions.
Eurozone: The New Latin America?
Vuletin and I further show that the fiscal policy response in
recent recessions in the eurozone (still ongoing, of course, for
countries such as Greece) has been eerily reminiscent of the pervasive
response in Latin America several decades ago. Figure 4 shows the
correlation between the cyclical components of government spending and
GDP from the beginning of the recession to the first quarter of 2013 for
10 eurozone countries. We see that four countries (Greece, Ireland,
Italy, and Portugal) have been procyclical, with Greece, not
surprisingly, the most procyclical of all. (13) We further show that
contractionary fiscal policy during bad times extended the duration of
the recession, intensified the fall in GDP, and worsened social
indicators.
Final Remarks
We should note, in closing, that monetary policy has not escaped
the procyclical trap. In fact, over the period 1960-2009, about 40
percent of developing countries pursued procyclical monetary policy.
(14) When the sample is divided before and after the year 2000, about 35
percent of developing countries are found to have graduated to
countercyclical monetary policy.
The source of procyclicality in monetary policy is the need, in the
minds of many policymakers in emerging markets, to defend the domestic
currency in bad times by raising interest rates. Policymakers often
fear, with some justification, that sudden currency depreciation will
increase inflation, exacerbate capital flight, and render
dollar-denominated debt of both public and private agents more onerous.
But whatever the merits, defending the currency in bad times imparts an
unavoidable procyclicality to monetary policy.
In sum, while progress has been made in the conduct of
macroeconomic policies in emerging markets, many continue to pursue
procyclical monetary and fiscal policies. By aggravating already
volatile boom-bust cycles, such policies have negative effects on output
and social indicators. From a macroeconomic point of view, this is
arguably the main challenge faced by developing countries as another
cycle of capital outflows and low commodity prices works its way
through. Further research in this area may help in identifying factors
that may enable more developing countries to adopt countercyclical
macroeconomic policies.
Carlos A. Vegh is a research associate at the NBER, a non-resident
senior fellow at the Brookings Institution, and the Fred H. Sanderson
Professor of International Economics at Johns Hopkins University, where
he is jointly appointed in the School of Advanced International Studies
in Wishington, D.C., and the department of economics of the Zanvyl
Krieger School of Arts and Sciences in Baltimore.
Vegh's research focuses mainly on monetary and fiscal policy
in developing countries. He received his Ph.D. in economics from the
University of Chicago in 1987 and spent his early career in the
International Monetary Fund Research Department. From 1995 to 2013, he
was a tenured professor first at the University of California, Los
Angeles, where he also was vice-chair for undergraduate studies, then at
the University of Maryland.
He has been co-editor of the Journal of International Economics and
the Journal of Development Economics and recently became lead editor of
Economia, the journal of the Latin American and Caribbean Economic
Association. He has co-edited a volume in honor of Guillermo Calvo and
published a graduate textbook on open economy macroeconomics for
developing countries. He has been a consultant to the IMF, World Bank,
Inter-American Development Bank, and many central banks around the
world.
Vegh lives in Bethesda, MD, with his wife, Ratna Sahay, and
stepdaughter, Maansi.
(1) I use the terms "emerging markets" and
"developing countries" interchangeably, since the prototypical
developing country that I have in mind has fairly standard fiscal
institutions and is reasonably integrated into world capital markets.
Return to text
(2) Updated from C. Reinhart, G. Kaminsky, and C. Vegh, "When
It Rains, It Pours: Procyclical Capital Flows and Macroeconomic
Policies" NBER Working Paper No. 10780, September 2004, and in M.
Gertler and K. Rogoff, eds., NBER Macroeconomics Annual 2004, Cambridge,
Massachusetts: MIT Press, pp. 11-53. Return to text
(3) C. Vegh and G. Vuletin, "How Is Tax Policy Conducted over
the Business Cycle?" NBER Working Paper No. 17753, January 2012,
and American Economic Journal: Economic Policy, Vol. 7(3), pp. 327-70.
Return to text
(4) On the former explanation, see A. Riascos and C. Vegh,
"Procyclical Government Spending in Developing Countries: The Role
of Capital Market Imperfections" mimeo, University of California,
Los Angeles, 2003; G. Cuadra, J. Sanchez, and H. Sapriza, "Fiscal
Policy and Default Risk in Emerging Markets" Review of Economic
Dynamics, 13(2), 2010, pp. 452-69, and S. Bauducco and F. Caprioli,
"Optimal Fiscal Policy in a Small Open Economy with Limited
Commitment" Journal of International Economics, 93(2), 2014, pp.
302-15. On the latter explanation, see A. Tornell and P Lane, "The
Voracity Effect" American Economic Review, 89(1), 1999, pp. 22-46,
E. Talvi and C. Vegh, "Tax Base Variability and Procyclicality of
Fiscal Policy" NBER Working Paper No. 7499, January 2000, and
Journal of Development Economics, 78(1), 2005, pp. 156-90, and A.
Alesina and G. Tabellini, "Why is Fiscal Policy Often
Procyclical?" NBER Working Paper No. 11600, September 2005, and
Journal of the European Economic Association, 6(5), 2008, pp. 1006-36.
Return to text
(5) J. Frankel, C. Vegh, and G. Vuletin, "On Graduation from
Fiscal Procyclicality," NBER Working Paper No. 17619, November
2011, and Journal of Development Economics, 100(1), 2013, pp. 32-47.
Return to text
(6) While 2000 is certainly an arbitrary point to break the sample,
minor variations do not make a difference. Return to text
(7) For a detailed discussion of the Chilean case, see J. Frankel,
"A Solution to Fiscal Procyclicality: The Structural Budget
Institutions Pioneered by Chile" NBER Working Paper No. 16945,
April 2011, and Journal Economia Chilena, Central Bank of Chile, 14(2),
2011, pp. 39-78. Return to text
(8) J. Frankel, C. Vegh, and G. Vuletin, "On Graduation from
Fiscal Procyclicality" NBER Working Paper No. 17619, November 2011,
and Journal of Development Economics, 100(1), 2013, pp. 32-47. Return to
text
(9) E. Ilzetzki, E. Mendoza, and C. Vegh, "How Big (Small?)
Are Fiscal Multipliers?" NBER Working Paper No. 16479, October
2010, and Journal of Monetary Economics, 60(2), 2013, pp. 239-54. Return
to text
(10) A. Auerbach and Y. Gorodnichenko, "Fiscal Multipliers in
Recession and Expansion" NBER Working Paper No. 17447, September
2011, and in A. Alesina and F. Giavazzi, eds., Fiscal Policy after the
Financial Crisis, Chicago, Illinois: University of Chicago Press, 2013,
pp. 63-98, Return to text
(11) D. Riera-Crichton, C. Vegh, and G. Vuletin, "Procyclical
and Countercyclical Fiscal Multipliers: Evidence from OECD
Countries" NBER Working Paper No. 20533, September 2014, and
Journal of International Money and Finance, Vol. 52(C), 2015, pp. 15-31.
Return to text
(12) C. Vegh and G. Vuletin, "The Road to Redemption: Policy
Response to Crises in Latin America" NBER Working Paper No. 20675,
November 2014, and IMF Economic Review, Vol. 62(4), 2014, pp. 526-68.
Return to text
(13) Recall from Figure 1 that, historically, Greece and Portugal
have been the only two industrial countries with procyclical government
spending. Further, Figure 2 shows these two countries as "still in
school" Return to text
(14) C. Vegh and G. Vuletin, "Overcoming the Fear of Free
Falling: Monetary Policy Graduation in Emerging Markets" NBER
Working Paper no. 18175, June 2012, and in D. Evanoff, C. Holthausen, G.
C. Kaufman, and M. Kremer, eds., The Role of Central Banks in Financial
Stability: How Has It Changed? World Scientific Studies in International
Economics, Book 30, Singapore: World Scientific Publishing, 2013).
Return to text
Figure 4
EUROZONE: COUNTRY CYLICALITY OF FICAL POLICY DURING LAST RECESSION
Greece .74
Portugal .42
Ireland .28
Italy .12
France -.29
Spain -.40
Austria -.47
Netherlands -.76
Belgium -.77
Germany -.90
Source: C. Vergh and G. Vulentin, NBER Working Paper No. 19828
Note: Table made from bar graph.