inflation stabilization in chronic inflation countries: the ... · the capital account improves...
TRANSCRIPT
Munich Personal RePEc Archive
Inflation stabilization in chronic inflation
countries: The empirical evidence
Reinhart, Carmen and Vegh, Carlos
University of Maryland, College Park, Department of Economics
August 1994
Online at https://mpra.ub.uni-muenchen.de/13689/
MPRA Paper No. 13689, posted 02 Mar 2009 05:53 UTC
JORNADAS INTERNACIONALES DE ECONOMIA
trprogramas de estabiljzaci6n:
La experiencia reciente I';;fl Amenea Latina ft
BANCO CENTRAL DEL URUGUAY Hontevideo, 8 Y 9 de agosto de 1994
ESTABILIZACION EN PAISES DE INFLACION CRONICA:
La Evidencia Empirica
CARMEN REINI~T (Departamento de Investigaciones, FMI)
CARLOS VEGH (Departamento de Investigaciones, FMI)
- T
Outline
For discussion at the conference only
July 31, 1994
Inflation Stabilization in Chror:ic Inflation .countries:
The Empirical Evidence 11
by
Carmen M. Reinhart and Carlos A. Vegh
Research Department, IMF
I. Introduction
Chronic inflation is characterized by long periods (i.e., several
decades) of high (relative to industrial countries) and persistent ,·inflation
(Pazos, 1972). The phenomenon of chronic inflation emerged in several Latin
American countries after World War II, and has been a distinguishing feature
of the economic landscape of several countries ever since. After numerous
failed attempts, some countries have succeeded to reduce inflation close to
international levels (Argentina, Chile, Israel, and Mexico).
There is a rich history of stabilizatj~ons in chronic inflation
countries, which spans more than three decades. This provides a unique
opportunity to identify the main patterns of adjustment and examine
11 This outline has been prepared to be discussed at the conference on
"Inflation Stabilization: The Recent Experience in Latin America," organized
by the Central Bank of Uruguay, to be held on August 8-9, 1994. The views
expressed her~ are those of the authors arid do not necessarily represent
those of the IMF.
r
- 2 -
econometrically some of the main features of disinflation in chronic
inflation countries. Our study is based on 17 stabilization plans from 1964
to the present in 7 countries: Argentina, Brazil, the Dominican Republic,
Israel, Mexico, Peru, and Uruguay (see Table 1). Twelve of these programs
were based on the use of the exchange rat,~ as the nominal anchor, while 5
were based on the use of a monetary aggregate.
In addition to reviewing the main "stylized facts" for these 17
stabilization plans (in the spirit of Kigne1 and Liviatan (1992), Vegh
(1992), and Calvo and Vegh (1994)), we resort to different econometric
exercises and thus are able to provide sm~e rigorous econometric basis for
several key features of disinflation in cllronic inflation countries.
II. Empirical regularities
This section reviews the main empirieal regularities of exchange rate-
based and money-based programs. These are:
1. Inflation converge only slowly to the rate of growth of the nominal
anchor (Chart 1).
2. The real exchange rate appreciates in both exchange rate-based and
money-based programs (Chart 2).
3. In exchange rate-based programs, there is an initial boom in
economic activity (real GDP and consumption) followed by a later
contraction. In money-based programs, the recession occurs at the beginning
of the programs (Table 2).
4. The external accounts (trade and current account) worsen in
exchange rate-based stabilizations.
- 3 -
5. The capital account improves sub!;tantially in exchange rate-based
stabilizations (although external factors seem to play an important role),
see Table 3. A similar story applies to stock markets (Chart 4). The boom
in asset markets is not limited to the equity market, but is also evident in
real estate prices, which tend to rise throughout the program.
6. International reserves appear to increase substantially in both
exchange rate-based and money-based stabilization, with the resulting
accumulation being more pronounced in periods of heavy capital inflows
(Chart 3 and Table 3). The accumulation of reserves in money-based
programs suggests considerably intervention in foreign exchange rate
markets, which may reflect an unwillingness to let the real exchange rate
appreciate.
7. Real interest rates have increased in the initial stages of money
based-stabilization. In exchange rate-based stabilizations, real interest
rates have generally fallen in orthodox pl~ograms (although in Chile they
remained at very high levels) and increased in heterodox programs (Table 4).
8. The fiscal stance varies considel~ably across programs. In
successful programs, there is a substantial fiscal tightening. In other
cases, however, the fiscal outcome may simply reflect the pick-up in
economic activity and the resulting increase in tax revenues (Table 5), as
emphasized by Talvi (1994).
III. Inflation Stabilization and Economic Growth
Both theory and casual evidence suggest that exchange rate-based
-----_._-_ ..-_.- •.._.~~-----------------
- 4 -
stabilization programs have been characterized by a boom-bust pattern in
economic activity, particularly private consumption. Money-based plans are
associated with and initial recession followed by recovery in economic
activity. The evolution of real GDP growth and consumption growth shown in
Table 2 highlights the marked differences in the so-called "inflation-
output" tradeoff between the two types of plans. The more systematic
econometric treatment discussed below also lends support to the "recession
now versus recession later hypothesis".
A panel of seven countries which have collectively implemented
seventeen inflation stabilization plans o,rer the last thirty years was used
to test if mean rates of growth in real GDP are statistically different
during inflation stabilization plans. The sample is 1964-1993 and the
countries are: Argentina, Brazil, Chile, Dominican Republic, Mexico, Peru,
and Uruguay. The results are robust to variations in the estimation
technique; the fixed-effects estimates are presented in Table 6.
The main results are:
1. The mean rate of growth increases in the early stages of an
exchange rate-based plan: the estimated irLcrease is about 2.3 percent and
statistically significant.
2. During the last year of the program, a recession ensues. Note
that these estimates do not imply an output decline of 5.4 percent, but a
decline of 1.8 percent (5.4 below the mean growth rate of 3.6). The
difference in growth in the final year of the program is also statistically
significant.
- 5 -
3. The mean rate of real GDP growth falls by about 5.4 percent in
the early phases of money-based programs. There is also evidence of a later
rebound in economic activity, but (unlike the early recession) this
difference is not significant.
4. The cumulative effect on output of an exchange rate-based
stabilization plan varies considerably as the duration of the plan varies.
As Table 7 illustrates, given the magnitude of the late recession, the level
of output ends up being lower for plans lasting three years or less. Short-
lived plans are costly.
5. More generally, the evidence suggests that output is not
invariant to th~ policy regime, as suggested by some of the recent
endogenous growth literature.
IV. How effective are the nominal anchors?
The speed and magnitude of the decline in inflation depends crucially
on a multitude of institutional factors as well as on macroeconomic and
microeconomic policies. The analysis that follows does not attempt to trace
out the various transmission mechanisms through which the nominal anchor
(the exchange rate or money) affect prices. Rather, the emphasis is on
examining the "reduced-form" dynamic interaction between these two variables
during selected stabilization plans. The goal is to assess how quickly a
policy shock is transmitted to.inflation, and thus make statements about the
presence or absence of inflation inertia and the effectiveness of policy.
A total of 11 plans are examined (nine exchange-rate based and two
money-based) .. A bivariate vector autoregi:ession (VAR) using monthly
- 6 -
inflation rates and devaluation rates is estimated for the exchange rate-
based plans; money growth (Ml) and inflation define the system for the
money-based plans. The period of estimation covers only the duration of the
plan, hence it is free from the parameter inconstancy that is at the heart
of the Lucas critique. We evaluate, the effectiveness of policy at two
levels.
First, we determine if there is a systematic and significant causal
relationship between the nominal anchor and the target variable, inflation.
Significant causality is necessary but not sufficient for policy to be
effective, since the impact may be significant but quantitatively small.
"For instance, one would expect the pass-through from exchange rates to
prices to depend on the openness of the economy, the degree of indexation,
the extent of credibility, and so on. Secondly, to assess the issue of
quantitative importance as well as speed of adjustment we examine the impact
on inflation of a hypothetical tempo.rary s:hock to the nominal anchor.
The main results can be summarized as foll,ows:
1. In eight out of the nine exchange rate-based programs there is
a significant causality from the rate of devaluation to the rate inflation
(Table 8). The exception is the Argentine 1967 plan, where the fixity of
the rate resulted in serious collinearity problems. The recent
Convertibility plan also shows some of the same problems. As one would
expect with an exogenous policy instrument, the exchange rate is not
affected by past (or current) inflation with the exception of Israel. Money
also appears exogenous during the money-based plans considered. However,
significant causality from money growth to inflation was detected in the
- 7 -
Chilean 1975 plan. Thus, causal patterns appear to be constant across
countries and across time.
2. In six of the nine episodes examined, past inflation
significantly affected current inflation, indicating systematic inflation
inertia. The inertial behavior was most pronounced in Chile, Mexico, and
Israel.
3. Charts 5-8 plot the response of monthly inflation to a
reduction in the rate of devaluation (or money growth) at an annual rate of
20 percent lasting 12 months. Note that:
a) The smallest impact on inflal:ion is seen in the plans of the
1960s, where inflation falls by 13.6 perceIlt for Argentina and only 10
percent for Uruguay (this is the lowest response in our sample). It may
possibly be reflecting the relatively closed nature of the economy at that
time.
b) The maximum effectiveness is reached in the Tablita plans,
where the decline in inflation ranges from 113.5 percent for Argentina to 22
percent for Uruguay, with Chile in the middle. Further, the decline in
inflation for the cases of Argentina and Uruguay is fairly sudden. These
two observations possibly suggest that the m)sence of more significant
declines in inflation in these programs may l:lavehad more to do with "policy
inertia" rather than with "inflation inertia". The inflation inertia
argument appears to find more support in the Chilean data, as it takes a
year for inflation to decline to its minimum level.
c) The exchange rate pass-through appears to decline in the 1980s
and 1990s. For the four exchange rate-based programs, the maximum decline
----------------------------------------------------
- 8 -
in inflation in response to a twenty percent reduction in the rate of
devaluation ranges from 11 percent for Mexico to 15 percent for the
Argentine Convertibility plan. This result remains puzzling. Perhaps
greater indexation and/or lower credibility may account for this change.
The time (in months) it takes to reach the minimum ranges from 3 months for
Uruguay to 14 months for Israel.
d) With regard to the money-based plans, the response to the
shock for both Chile and the Dominican Republic is gradual, taking 12-15
months to reach its most pronounced effect~ The inertia is also evident in
the Chilean 1975 money-based plan, where It takes a year for inflation to
decline by about 14 percent. Possibly, the more delayed effects may reflect
the inherent difficulties in monetary con1:rol.
e) Based on three indicators of inertia (the significance of past
inflation, the response to a temporary policy shock, and the speed of
adjustment) for the more recent peri?d, I~:rael and Mexico and, to a lesser
degree Chile, stand out as the countries in which inflation inertia has been
most pervasive. On the other hand, in Argentina, the Dominican Republic and
to a lesser extent, Uruguay, inflation inertia does not appear to be that
much of a problem. The results also highlight that the interaction between
instrument and target, not only vary across countries (which is obvious),
but can change markedly across time.
Table 1. Major Inflation Stabilization Plans
Plan Pori 0 d
Exchange Rate-Based Stabilization Plans
Orthodox plans
Argentine tablita
Chilean tablita
Uruguayan tablita
Convertibility Plan (Argentina)
Uruguay 1991
1978:12 - 1981:2
1978:2 - 1982:6
1978: 10 - 1982: 10
1991:4 - present
1991: 1 - present
Heterodox plans
Argentina 1967
Austral (Argentina)
Brazil 1964
Cruzado (Brazil)
Israel 1985 11 .
Mexico 19871/
Uruguay 1968
1967:3 - 1970:10
1985:6 - 1986:9
1964:3 - 1968:8
1986:2 1986: 11
1985:7 - 1990:6
1987: 12 - 1992:2
1968:6 -1971:12
Money-Based Stabilization Plans
BONEX (Argentina)
Collor (Brazil)
Chile 1975
Dominican Republic
Peru
1989:12 - 1991:1
1990:3 -1991:1
1975:4 - 1977: 12
1990:8 - present
1990:8 - present
1 /These a re success fu l p rog ram s, hence the te rm Ina l da te has been a rb itra rily
se t to five ye a rs .
Table 2. Inflation Stabilization:Real GDP
( •.nou- .•.l rAt. of growth, in peroent)
Proc; r.a.Dl
thr ••
yeAr. Year
belor. Firat. Second Third proqr •.•
ergST"" Iter Year refT ended
Exchange Rate-Based Plans
Argentina 1967Argen tine Tablita
Argen tina (Austral)Arg en tin e Coo vertib ilityBrazil 1964
Brazil (Cruzado)O1ilean Tablita
Israel 1985MeXico 1987
Uruguay 1968Uruguayan Tablita
Uruguay 1991Average
6.5-4.0
-0.4
-2.7
Money-based Plans
Argentina (Bono:)Brazil (Collor)
O1ile 1975Dominican Rep.1blic 1990Peru 1990Average
-1.8
24-1.9
4.6-3.9
-0.1
3.30.2201.1
0.23.50.70.9
0.1-4.4
-12.9-5.4-4.4-5.4
0.87.1
7.38.9
26
7.68.2
4.01.21.1
6.20.04.8
8.7
23.1
8.3
3.63.3
6.36.05.37.1
3.30.7
4.9 26-5.7
26
Growth and Consumption CyclesReal Private Consumption
CAnna •. l rat. of qro'ttth, in percent)
'1'h r ••
Progra.JIl,
,. ••. ra r •• r
before Fir.t Second Third proqra.m
pro9£tm YeAr rear year ended
Exchange Rate- Based Plans
Argen tina 1967Argentine TabIita
Argentina (Austral)Argentine Coovertibility
Brazil 1964
Brazil (Cruzado)O1ilean Tablita
Israel 1985Mexico 1987
Uruguay 1968Uruguayan TablitaUruguay 1991Average
6.8-4.21.2
-2.1
3.6
281.0
0.60.3
0.50.2
-5.3
0.5
3.5
-0.9
261.7
5.0
3.5 10.8
3.67.8 -14.16.1 1.3
4.4 264.8 -1.0
20 -9.65.04.6 -1.2
7.7-2.8
25
Money-based Plans
0.9
9.9
Argentina (Bono:)Brazil (Collor)O1ile 1975Dominican Rep.1blic 1990Peru 1990Average
-1.2
-0.5
-6.3
-0.31.5
-1.45.4
Hot •• J Dot_ lndi.c~t. the dAtA Are not .,..i.l.bl •• D•• h •• J.ndJ.oAt. d.t. do not .pp1 r.
r11••• Are .ucce •• ful p rog rA II., h'.nee the eer1!in.l d,.t. b,.. b •• n .rbitr.ri.ly •• t to fi,... y •• r •.
O n. r •• r b .fo r ••
loure • ., Buf",a" and z,.J.d.r",a" 0 ")1 , Fav.ro and s.".J.on 0 ")1 , /Ugu.l and z,J..,J.otan {l"'I,
z,u.tJ.g 0 "21 , x.d.lro. 11 ")/, ral.,J., { l"41 , Vlano (1 "01 , Int.rnatJ.onal FlnanoJ.al ItatJ..tJ.ca IIXFI,
world S.'nJc t.bl •• , Fund~ci.on H.dit.zr~n ••. , nAtio'n.l .ourc •• , .nd Fund .e.ff •• e1".tt •••
2614.4
7.915.73.36.4
7.514.81.8
8.2
9-1.5
7.5
-1.8
-2.5
-11.4
-12.9-15.3
-8.8
10.10.7
6.59
6.3
6.4
510.46.4
0.37.5
10.86.2
4.05.6
6.4 4.4
-3.6
0.7
4.0
4.3 10.2
-0.96.8 -12.1
4.3 0
5.7 4.9
24 -9.7
13
5.9 -0.5
3.9
16
-1.1
10.0
:pro9r~l'll
T a b le 3 . T h e E x te r n a l A c c o u n t s
Current account balance excluding offk:ial transfers
T h r c o
yeAr. Year
b e fo r e r ir .t S e o o n d T h ir d p r o 9 r a .m
p r o q r am Y .a .r r ." r r ." r e n d e d
Exchange Rate- Based Plans
Argentina 1967Argentine TablitaArgentina (Austral)Argentine ConvertibilityBrazil 1964
Brazil (Cruzado)Chilean Tablita
Israe11985Mexico 1987
Uruguay 1968Uruguayan Tablita
Uruguay 1991Average
1.4
0.5-2.1-1.4
-1.2
-2.3
-2.7-3.3
0.71.7
-3.7
0.5-1.3
Money- based Plans
Argentina (Bonex)
Brazil (Coilor)Chile 1975
Dominican Republic 1990Peru 1990Average
-2.4
-0.3-1.8
-5.9-4.1
-2.9
1.0
-0.4
-2.7-1.5
0.1
-2.1
-7.1
4.7-2.2
1.4
-5.2
0.7-2.1
1.3
-0.7
-5.8
-3.7
-4.3
-2.6
-0.3
-2.3
-3.70.4
-5.7
5.5-2.9
0.9-7.5-1.8
-3.3
1.5
-3.3
-5.1
-2.3
-1.2
-3.3-0.1
-7.1
-2.5
-3.2
-2.7-4.5
1.1
-3.2
-2.6
-5.2
-3.9
-0.7
-2.8
-3.9
-0.8
-0.5-9.5
-7.0
-2.8
-5.2
-4.7
-0.3-4.1
-2.2
Capital balance including errors and omissions(a.. a peroent of aOP)
proqra.JIl
T h r o .
y e a r . ~ear
b e fo r e T ir .t S .c o n d T h ir d p r o q r a J l
p r o g r a .m Y .a r Y e a r Y e a r e n d e d
Exchange Rate- Based Plans
Argentina 1967Argentine TablitaArgentina (Austral)Argentine ConvertibilityBrazil 1964
Brazil (Cruzado)Chilean Tablita
Israel 1985Mexico 1987
Uruguay 1968Uruguayan TablitaUruguay 1991Average
-1.0
-2.9
-6.0
-0.1
1.0
-0.7
3.1
-3.5-1.3
Money- based Plans
Argentina (Bonex)Brazil (Coli or)Chile 1975Dominican Republic 1990Peru 1990Average
-2.3
-2.8-1.3
-0.0-5.4
-2.3
Hot •• ' Dot. iDdloAt. tb. dAtA ~r. n o t AVA1JAbJ •• D •• h •• indicAt. d.tA do not Apply.
l/~h •• e Are .uoo ••• lul progrA~.J h.noe th. t.r.ln.J d.te b •• been .rbitrArily .et to lip. y •• r •.
$ourc •• , tnt~rnAtion.l FinAnoi.l 6tAti.tlo., And world Econnomic outlookJ tHF.
4.6-1.6
-2.8
-6.4
22.4
-3.4
12.0-2.82.4
-1.8
-2.0-1.0
-13.6
-4.2
-4.5
-4.4
21.2
11.8
-7.2
4.3
3.813.0
7.0
7.9
10.2
3.4
- 3.0
2.6
-0.2
-6.8
-3.2
14.6
24.0
-6.0
7.0
5.4 15.2
18.4
7.4
3.6
-8.0
1.1
-1.6
12.0
5.2
Table 4. Real Interest Rates i.l Selected Stabilization Programs.
Real interest rates b
(in percent per year)
Four
Quarters First Last
before Four Four
Program Quarters Quarters
Programs Period a ~verage) (average) (average)
Exchange Rate- Based
Argentina 1967 1967-1970
Brazil 1964 1964-1968
Uruguay 1968 1969-1971
Argentine tablita 1979-1981 0.7 -2.8 5.9
Chilean tablita 1978-1982 70.9 43.0 46.4
Uruguayan tablita 1979-1982 18.2 -7.2 24.9
Austral (Argentina) C 1986 20.0 48.0 -7.5
Cruzado (Brazil) C 1986 -4.5 8.5 -9.5
Israel 1985 d e 1986-1990 -2.0 21.2 11.0
Mexico 1987d 1988-1992 -2.9 292 2.0
Convertibility (Arg.) f 1991-1992 38.1 -2.0 4.0
Money-based
Chile 19759 1975-1977 127.2 58.0
Bonex (Argentina) 1990 -7.4 112.7
Collar (Brazil) h 1990 -8.1 -2.4
Dominican Rep. 1990 f i 1990-1992 15.1 13.7
Peru 1990 f 1990-1992 --17.3 235.0 48.1
.• Calendaryears during which the program was taken to be in effect for the purpOses
ofcurrentaccountfigure~ .
b Quarterly real lending rates unless otherwise indicated. Periods specified in
Tables 1 and 4 apply. Dots indicate data are not available. Dashes indicate data do not apply.
C Real interest rates are reported for two-quarter periods, and exclude the initial
price shock.
d Duration of program has been arbitrarily set to five years ..
•• Real interest rate before the program refers to two quarters before.
f Program in progress.
g Annualreal interest rates.
/, Monthly averages of overnight interest rates on government securities. Real interest
rate after the program refers to first three quarters.
i Real interest rates for 1991.3 and 1991.4. Before January 1991, interest rates were
subje<:t to controls.
Sources: Bali no (1991), Barkai (1990), Bufman and Leider.nan (1993),
Castro and Rond (1991), Cukierman (1988), Kiguel and Liviatan (1989),
Perez-Campaneroand Leone (1991), International Enand:d Statistics (IMF),
and national sources.
Table 5. Public Sector Balance in Se!ec,ed Stabilization Program$ 1/
(as percentofGCP)
Three
Yean
before
Program Fint Second Third Fourth Fifth
Program$ Period 11 (average) Year Year Year Year Year
Exchange Rate-Ba$ed
Argentina 1967 1967-70 -4.1 -2.0 -2.0 -1.5 -1.6 -4.1
Brazil 196431 1964-68 -4.0 -3.2 -1.6 -1.1 -1.7 -1.2
Uruguay 1968 1%9-71 -1.8 -2.5 -1.3 -5.8 -2.6
Argentine tablita 1979-81 -5.9 -4.7 -6.1 -8.5 -5.3
Chilean tabIita 1978-81 0.8 1.5 4.8 5.4 0.3 -4.0
Uruguayan tablita 1979-82 -1.4 (1.0 -0.3 -1.5 -9.1 -3.9
Austral (Argentina) 1986 -7.1 -2.0 -4.3
Cruzado (Bram) 1986 -3.9 -3.6 -5.5
Israel 19854/ 1986-90 -4.9 ~.3 0.0 -0.3 -4.0 -2.7
Mexican 198741 51 1988-92 -0.3 -4.5 -1.7 2.6 2.3 3.3
<Anvertibility (Arg.) 61 1991-93 -6.0 -0.9 0.6
Uruguay 1991 1991-93 -4.8 -0.9 -0.0 -1.6
Money-based
Chile 1975 1975-77 -20.1 -2.0 3.9 0.4 1.5
Bonex (Argen tina) 1990 -9.3 -2.3 -1.2
<Allor (Brazil) 51 1990 -5.8 1.4 1.5
Dominiean Rep. 199061 1991-92 -5.9 (1.1 1.6
Peru 19906/ 1991-92 -7.5 -2.6
Souces : B u tm an and Le lde rm an (1993), C orbo and So Ilm ana (1,~1), 01 Te lla and D om busch (1989),
K lgue landLM a tan (1989 ). Lem grw e r(1977 ). M ede Iros (1993), R W lO S (1986), Wana (1990 ),
na tiona l sou rces , and Fw d s ta ff es tim a les .
1 / O Ve ra ll ba lance o f the non -financ Ia l pub lIc sec to r, u r"le ss o thuvJse Ind Ica ted . A m Inus s Ign
Ind ica tes a de fic it.
21 C a lenda r yea rs duJ ing v .h lch p rog ram w as In e ffed . If a p rogm m sta rted ea rty
In the yea r. tha t yea r Is a lso Induded . R gu res repo rted up to one yea r a fte r
the p rog ram ended . D o ts Ind ica te da ta a re no t avaH ab le .
3 / Fede ra l govem m eri .
.4/ D ura tion o f the p rog ram has been a rb ltra rftiy se t to five yea rs .
Table 8. Causality and Exchanpe Rate Regimes
Plan and Sample Period
Inflation EqualJon
F-statistic Probability value
EXchange Rate EqualJon
F-statistic Probability value
Exchange Rate-Based Stabilization Plans
Argentina
1967 Plan: 1965:1 - 1972 :6 "'v
Inflation 1.257 (0.287) 1.039 (0.407)
Exchange Rate 0.980 (0.444) 0.380 (0.889)
Tablita: 1978:12 - 1981:2
Inflation 0.772 (0.926) 0.255 (0.975)
Exchange Rate 3.398 (0.051) 0.026 (0.399)
Convertibility Plan: 1991:4 - 1994 :2v
Inflation 8.021 (0.001) 0.547 (0.654)
Exchange Rate 2.160 (0.115) 4.761 (0.008)
Chile
Tablita: 1978:2 - 1982:5
Inllation 26.283 (0.000) 1.480 (0.232)
Exchange Rate 4.765 (0.006) 3.706 (0.018)
Israel
1985:7 -1994:2
Inflation 10.509 (0.000) 6.193 (0.000)
Exchange Rate 2.655 (0.067) 3.449 (0.011)
Mexico
1987:12 - 1994:2
Inflation 6.988 (0.000) 1.324 (0.274)
Exchange Rate 4.533 (0.006) 4.544 (0.006)
~1968 Plan: 1966:1 -1973 :12 11
Inflation 1.565 (0.203) 1.579 (0.200)
Exchange Rate 5.635 (0.001) 2.337 (0.079)
Tabrrta: 1978:10 -1982:10
Inllation 7.264 (0.000) 0.723 (0.634)Exchange Rate 2.278 (0.057) 44.479 (0.000)
1991:1-1994:2
Inllation 0.568 (0.641) 2.139 (0.115)Exchange Rate 2_349 (0.092) 1.250 (0.309)
Memo item:
Chile
Real Exchange Rate Target; 1985:7 -1991:12
Inllation 1.968
Exchange Rate 0.248
Foreign Inllation 1.054
(0.127)
(0.862)
(0.374)
3.880
0.632
2.266
(0.013)
(0.597)
(0.088)
Money-Based Stabilization Plans
Chile
1975:4 -1977:12
Inllation
Money (M1)
28.036
2.793
(0.000)
(0.060)
0.515
1.334
(0.676)
(0.284)
Dominican Republic
1990:8 - present
. Inflation
Money (Ml)
0.367
0.843
(0.587)
(0.938)
7.0661.158
(0.000)
(0.371)
Notes: The optimal lag length was chosen according to the Schwarz: cJiteria Details are given in the appendix.
!/]hese plans fixed the exchange rate, hence for the duration of the plan it mimics the constant temJ of the regression
and a serious co/linearity problem emerges. To attemt to capture (; period were the exchange rate eJdJibitssome
variability,we broadened the sample adding about two years of obmvations at both ends. This was not
feasible for the case of the convertibility plan, since the period imm<1diatelyduring preceeding it a money-bBsed
plan was In effect (sea Table 1)
'11-1lentho sample Is extended further back (to 1960:1), there is a ~ignificant causal relationship from the exchange
Table 6. Inflation Stabilization and Growth:
Evidence from Panel Data
(1964 - 1993)
Dependent
Variable:
Constant Stabilization Plans Dum mies
(M aan) Exchange rate
E a,r1y Lata
Money
Early Late
Annual Growth Rate
in Real G DP
3.645
(0.518)
2.331
(1.037)
-5.454
(1.631)
-5.403
(1.931 )
1.822
(2.909)
N o tes : The coun trie s inc luded in · the sam p le a re A rgen tina . B raz il. C h ile . D om in ican R epub lic .
M ex ico . P e ru , and U ruguay . hence the re a re 210 obse rva tions . S tanda rd e rro rs a re in
pa ren theses . The random e ffec ts es tim a tes a re repo rted .
Table 7. Output Effects of Exchange Rate-Based
Stabilization Plans as the Duratic,o' of the Plan Varies
Level of Real G DP at the end of year:
Beginning
of th e
End
of th e
program 2 3 4 5 pro gra m
100
100
100
100
100
102.331
102.331
102.331
102.331
102.331
104.716
104.716
104.716
104.716
96.7498
107.157
107.157
107.157
99.0051
109.655
109.655
101.312
112.211
103.674
106.091
N o tes : These ca lcu la tions re fle c t the d iffe rence in ou tpu t due to the s tab iliza tion p lan and
a re based on the es tim a tes repo rted In Tab le 6 .
CHA ..R~T 1
Exchange Rate-Based StabilizationOrthodox Programs 12-Month Inflation Rates
ARGENTINE TAB UTA CHILEAN TABUTADecember 1978 February 1978
180 60
170
160
150 ~O140
130 W
120
110W
100 lO90
BOO o .~4 8 12 16 20 24 0 4 8 12 16 20 24 28 32 36 40 44 48 52
Months after beginning oC stabilization Months after beginning of stabilization
URUGUAY 1968 URUGUAYAN TABLITAJune 1968 October 1978
200 "0
180 m160
~O140
120({)
100 ~Of80'0
60::0
40
20 :10
°0 36]0 -
12 16 20 24 28 32 36 40 444 8 12 16 20 24 28 32 40 0 4 8
Months after beginning oC stabilization Months after beginning oC stabilization
ARGENTINE CONVERTIBILITYApril 1991
URUGUAYAN PLANJanuary 1991
300
4 8 U ~ ~ ~ nMonths af!er beginniot: of stabilization
1'0 --
1:0 \
1;0
110
](0
~O
10
,0
(0
~O
400 7 ~V n· ~
Months aCter ber:innint: of stabilization
Notes: The date for the beginning of the plan is giver above each graph. The 12-month inflation
. rates are plotted for the duration of the progr am.
CHART
Exchange Rate-Based StabilizationHeterodox Programs: 12-Month Inflation Rates
AUSTRAL PLAN
June 1985
1200
1000
800
600
400
200
o 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Months after ~inning of slabilization
CRUZADOFebruary 1986
250
200
150
100
50 01 2 3 4 567 8 9
Months after beginning of slabilization
ISRAELI PLAN MEXICAN PLANJuly 1985 Deccmbc:r 1987
450 180
400 160
350 140
300 120
250 100
200 80
150 60
100 40
50 20
°0 7 14 21 280
35 42 49 56 0 7 14 21 28 35 42 49 56
Months after beginning of stabilization Months after beginning of stabilization
Notes: The date for the beginning of the plan is given abov(: each graph. The 12-monlh inflation
rates are plotted for the duration of the program. F<'r the ongoing plans (Le. the Israeli
and Mexican plans) an arbitrary cut off of five years .vas chosen as the "end" of the plan.
CHART
Money-Based Stabilization12-Month Inflatior Rates
BONEX (ARGENTINA) COLLOR (BRAZIL)December 1989 March 1990
25 7000
206000
~5000.•
-a15c
'" 4000.•'"0
10E,3000
5 2000
010
1000(J 1 2 3 4 5 6 7 8 9 100 2 4 6 8 12
Months after beginning of stabilization Months after beginning of stabilization
DOMINICAN REPUBLIC PERUAugust 1990 August 1990
120 14
100 12 \
~ 10.•-ac 8~'"0 6
40 t:4
20 2
°0 5 10 15 2(J 25 30 35 00 5 10 15 2(J 25 30 35 40
Months after beginning of stabilization Months after N:ginning of stabilization
Notes: The date for the beginning of the plan is given abo're each graph. The I2-month inflation
rates are ploUed for the duration of the program.
CHART 1
Exchange Rate- Based ~tabilizationOrthodox Programs: Real Effecti rc Exchange Rates
ARGENllNE TABUT A
December 1978=100
Cf-IILEAN TABLITA
February 1978=100
4 8 12 16 20 24
Months after beginning or stabilization
110
105/100
95
90
85
80
75
70
650 4 3 12 16 20 24 2S 32 36 40 44 48 52
Me nlhs after beginning or stabilization
100
95
90
85
SO
75
70
65
60
55
500
URUGUAYANTABLITA
October 1978=100
AP GENTINE CO NVERTIB rrnYApril 1991=100
110 100
95
4 8 12 16 20 24 2S 32
Me nths arter beginning or stabilization
90
85
60 80
500 4 8 12 16 20 24 2S 32 36 40 44 48
Months after beginning or stabilization
URUGUAYAN PLlli
January 1991 =100
100
95
90
85
80
75
700 7 14 2t 28
Months .ftef beginning of sub liution
Notes: A decrease in the index indicates real appreciation. Tlle date for the beginning of
the plan is given below each graph. The real exchange ~ates are plotted for the
duration' of the program.
CHARTz
Exchange Rate-Based StabilizationHctcrodox Programs: Real Effcci ivc Exchange Rates
AUSTRAL PLAN CRUZADO
June 1985=100 February 1986 = 100
104 112
110
r100
108
98106
96
10494
92 102
901 2 3 4 5 6 7 8 9 1011 121314 15
1000 12" 3 4 5 6 7 8 90
Months after beginning or stabilization Months after beginning of stabilization
ISRAELI PLAN
July 1985=100
104
102
100
98
96
94
92
90
88
860 7 14 21 28 35 42 49 56
Months after beginning or stabilization
MEXICAN PlAN
December 1987=100
100
95
90
85
80
75
·70
65
600 7 14 21 28 35 42 49 56
Months after beginning of stabilization
Notes: A decrease in the index indicates real appreciatillO. The date for the beginning of
the plan is given above each graph. The real excl:1ange rates are plotted for the
duration of the program. For the ongoing plans ~i.e. the Israeli and Mexican plans)
an arbitrary cut off of five years was chosen as th.: "end" of the plan.
CHART 2
Money-Based Stabili~:ation ProgramsReal Effective Exchange Rates
BONEX (ARGENTINA) COLLaR (BRAZIL)December 1989=100 March 1990=100
130 lGO
120
no 150
100 14090
80 130
70
60120
50 llO40
306 8 10 12
1000 1 2 3 4 5 6 7 8 90 2 4
Months a(ter beginning o( wbilizalion Months after beginning o( stabilization
DOMINICAN REPUBLIC PERUAugust 1990=100 August 1990=100
100 140
98
1~ r%
94 120
92 11090
88 100
86 90
84
8280
800 5 10 15 20 25 30 35 40
700 5 10 15 20 25 30 35 40
Months a(ter beginning o( stabilization Months after beginning o( stabilization
Notes: A decrease in the index indicates real appreciati on. The date for the beginning of
the plan is given above each graph. The real excllange rates are ploued for the
duration of the program.
CHAF~T J,
Exchange R a te -B a s e d S ta b i l iz a t io n
Orthodox Programs: Tola! Reserves Minus Gold
ARGENTINA (1967)March 1967= 100
450
400
4 8 U M m M m ~ ~M on th s a C ta - b e g in n in g o f s ta b i l iz a t io n
URUGUAY (1968)June 1968=100
CHILEAN TABLITAFebruary 1918=100
8 U U m M m II ~ ~ « • IIMonths .C,er beginning of stabtlization
ARGENTINE CONVERTIBILITYApril 1991=100
U 16 m 24 m 32
M on th s a f t e r ~gillnin&oC s 1 3 b t l iz .3 t . io a
BRAZIL (1964)March 1964= 100
550
500
450
400
350
300
250200
150100 •..•..
~O 4 8 U U m M m II ~ ~ « • IIM on th s a f t e r b e g in n in g o f s ta b i l iz a t io n
ARGENTINE TABLITADecember 1918=100
200
180
160
1~
1m
100 ..-
80
600 8 U M m M
M on th s a f t e r b e g in n in g o f s ta b i l iz a t io n
URUGUAYANTABLITAJanuary 1918=100
1~
1m
100 V80
60
~
mo 4 8 U H m ~ m II ~ ~
Months aCta beginning of st.abtlization
URUGUAYAN PLANJanuary 1991=100
180
160
1~
1m
14 21 mMonth •• C'e< beginning of stabilization
35
Notes: The date for the beginning of the plan is given abov( each graph. Total Reserves Minus
Gold data is plotted for the duration of the program.
----------------------------------------
.'
CHAR,T'~,3
Exchange Rate-Based StabilizationHeterodox Programs: Total Reserves Minus Gold
AUSTRAL PLAN CRUZADOJune 1985=100 February 1986=100
260 11
240 105
220 100
200 95
180 90
160 85
140 80
120 75
100o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
700 1 2 3 4 5 6 7 S 9
Months after beginning of stabilization Months after beginning of stabi1iz:ation
ISRAELI PLAN MEXICAN PLAN
July 1985=100 December 1987=100
260 1
240 140220
200120
180 tOO
160 80
14060
120
100 40
80 2000 7 14 21 28 35 42 49 56 7 14 21 28 35 42 49 56
Months after beginning of stabilization Months after beginning of stabilization
Notes: The date for the beginning of the plan is given aX>ve each graph. The Total Reserves
Minus Gold data is plotted for the duration of tlIe program. For the ongoing plans
(i.e. the Israeli and Mexican plans) an arbitrary CJt off of five years was chosen as
the "end" of the plan.
CHARrrl3
Money-Based Stabilization ProgramsTotal Reserves Minus Gold
CHILEAN PLANApril 1975= 100
700
600
500
400
300
200
100
°02468WUMMWWll~~W~TIMonths after beginning of stabilization
coLLaR (BRAZIL)March 1990=100
170
160
150
140
1~
120
110
1001 2 3 4 5 6 7 8 9
Months after beginning of stabilization
500
450
400
350
300
250
200
150
100 0
BONEX (ARGENTINA)December 1989=100
350
250
30(i
2O<J
150
100
50 0 2 4 6 8 10 12
Months after beginning of stabilization
DOMINICAN REPUBLICAugust 1990=100
450
400
350
300
250
200
150-
100-
50-
o·o 5 10 15 20 25 30 35 40
Months after beginning of stabilization
PERUAugust 1990=100
5 10 15 20 'B' 30' 35
Months after beginning of stabilization
,I
40
Notes: The ~ate for the beginning of the plan is given above each graph. The Total Reserves
Minus Gold data is plotted for the duration of the program.
CHART', 4
Exchange Rate-Based S ta b i l iz a t io n
Orthodox Programs: Stock Market Price Indices
ARGENTINE TABLITADec<:mbcr 1978=100
700
600
500
400
300
200
<4 & 12 16 20 24
Months after beginning of 6tabilization
CHILEAN TABLITAFebruuy 1978 = 100
350
300
250
200
150
l00 ....•~
500 7"7'12 16 20 24 28 32 36 .w « 4S 52
Months after beginning of ltabilization
ARGENTINE C O N V E R T IB IL IT Y
April 1991=100
350
300
250
200
150
100
50 0 4 8 12 16 '"~ 28 32
Months after beginning <:>fstabilization
Notes: The date for the beginning of the plan is given above each graph. Ibe stock market price
indi~ are plotted for the duration of the program.
CHART 4
Exchange Rate-Based StabilizationHeterodox Programs: .
AUSlRAL PLAN CRUZADOJune 1985=100 Febru~ry 1986=100
350 200
190
300 1&0
170
250 160
150
200 140
no150 120
110
1000 1 234 5 6 7 8 9 10 11 1213 14 15
1001 2 3 4 5 6 7 8 9
Months after beginning of stabilization Moc,ths after beginning of stabiliutioo
MEXICAN PLANDecember 1987=100
1100
1000
900
SOO
700
600
500
400
300
200
100 0 4 8 12 16 20 24 28 32 36 40 « 48 52 56
Months after beginning of stabili~tion
Notes: The date for the beginning of the plan is given above each graph. The stock man::et price
indi~ are plottod for the duration of the program. For ahe ongoillg plans an arbitrary cut off
of five years was chosen as the "end" of the plan.
CHART 4
Money- Based StabilizationStock Market Price Indices
BONEX (ARGENTINA) COLLaR (BRAZIL)December 1989=100 Much 1990=100
100 1~
15090
1<W
to 1~
12070
110
~ 100
9050
to
<W0 4 IS 10 12
704 IS
Month& after bePnnin& or nabiliu.tion Month& after bePnnin& of mbiliu.tion
PERU (1990 PLAN)August 19'90 = 100
lto
170
1~
150
1<W
1~
120
110100)
90
to 0 4 ,-. 12 1 IS 20 24 21~ n 36 <W 44
Month& after be&i=in& or nabifuation
Notes: The date foc the beginning of the plan is given above each graph. The stock market price
indices an: plotted for the duration of the proram.
Chart 5. Exchango Ratos and Inflation: Programs of tho 1960s·
Effect of Temporary 20% Reduction in tho Rate of Devaluation
Percent
Ar entina 1967 Plan2
o-2
-4
-6
-6
-10
-12
-14
·16
-16
-20
Sample period:
1965:1-1972:6
o 4 10 22 246 12 14 16 202 6 16
Number of Mon:ths
Percent
2
o
-2
-4
-6
-B
-10
-12
-14
-16
-16
-20
Uru uay 1968 Plan
Sample period:
1966:1 -1973:12
o 2 4 6 6 10 12 '14 16 18 2~ 22 24 26 28 30 32 34 36
Number of Months
Notes: Temporary refers to a chl'lnge lasting 12 months. Results of the estimated
equations aro summarized in Tables 8-9.
Sources: International Financial Statistics, IMF and the authors.
Chart 7. Exchange Rates and Inflation: Programs of the 19808 and 19908
Effect of Temporary 20% Reduction In the Rate of Devaluation
Percent Percent
Israeli Plan Mexican Plan4
2
o·2-4
-$
.a·10
·12
·14
·16
·18
·20o
Sampleperiod:
1985:7 ·1994:2
Sample period:
1987:12 ·1994:2
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 o 2 4 6 8 10 12 14 16 18 20 22 24
Number of Months Number of Months
P e r c e n t Percent
Argentine Convertlbillt Plan Uru 1991 Plan
Sampleperiod:
1991:4 ·1994:2
Sampie period:
1991:1 ·1994:2
o 2 4 6 8 10 12 14 16 18 20 22 24 o 2 4 6 8 10 12 14 16 18 20 22 24
Number of Months Number of Months
N o tes : Tem po ra ry re fe rs to a change las tIng 12 m on ths . R esu lts o f the es tim a ted
equa tions a re sum m arized In Tab les 8 ·9 .
S ou rces : In te rna tIona l F inanc Ia l S ta tis tics , IM F and the au tho rs .
Chart 8_ Monoy and Inflation
Effoct of Tomporary 20% Reduction in tho Rate of Growth of Ml
Percent
2
o
-2
--4
-{l
-{l
-10
-12
-14
-16
-18
-20
Sample period:
1975:4 -1977:12
o 2 10 12 14 16 18 20 22 24 26 2S6 e
Number" of Monlhs
"2o-2-4
-{l
-{l
-10
-12
-14
-16
-18
-20
Sam~pcriod:
1990:8 - 1993:10
Number of Months
Notes: Temporary refers to a change/asting 12 months. Results of the estimated
equations are summarized in Tables 8-9.
Sources: International Financial Statistics, /MF slid the authors.