Inflation Targeting:
Does It Improve Economic Performance?
Stephen M. Miller *
Department of Economics
University of Nevada, Las Vegas
Las Vegas, Nevada, U.S.A. 89154-6005
stephen.miller@unlv.edu
WenShwo Fang
Department of Economics
Feng Chia University
Taichung, Taiwan
wsfang@fcu.edu.tw
Ozkan Eren
Department of Economics
University of Nevada, Las Vegas
Las Vegas, Nevada, U.S.A. 89154-6005
ozkan.eren@unlv.edu
Abstract
The last two decades witnessed a dramatic transformation of how central banks operate. An
increasing number of central banks now use inflation targeting as their monetary policy control
mechanism. A series of papers attempt to measure the effectiveness of inflation targeting on
economic performance. The basic challenge in such tests is that inflation targeting appeared
during a time when inflation trended downward across nearly all countries – those that did and
did not adopt inflation targeting. This paper reviews the existing methods used to test for the
effectiveness of inflation targeting and compares the findings of these different methods for both
developed and developing countries. In general, inflation targeting does not affect economic
performance in developed countries but does exert a positive effect on economic performance in
developing countries. We conclude that the effectiveness of inflation targeting policy garners
little, or only transitory, support based on evidence from developed countries. Much more
support exists for developing countries.
Keywords:
inflation targeting, difference in differences, fixed and random effects,
treatment effects, developed and developing countries
JEL classification:
C52; E52, E58
* Corresponding author
1.
Introduction
The last two decades witnessed a dramatic transformation of how central banks operate. Rather
than keeping policy actions secret, the new view stresses openness and transparency as important
hallmarks of good central banking practice. 1 Moreover, an increasing number of central banks
now use inflation targeting (IT) as their monetary policy control mechanism. New Zealand led
the way by adopting this policy in 1990. At the beginning of 2012, 27 central banks used IT,
about two-thirds of them developing countries (see Table 1). Closely related research on central
banking falls into several categories, such as central bank independence, credibility,
accountability, transparency, and communication as well as the evaluation of monetary policy
strategies (e.g., Walsh, 1995a, b; Faust and Svensson, 2001; Issing, 2005; Fatás et al., 2007; Lin
and Ye, 2007; Acemoglu et al., 2008; Blinder et al., 2008; Gonçalves and Salles, 2008; Svensson,
2009; Brito and Bystedt, 2010, and Siklos, Bohl, and Wohar, 2010).
Bernanke et al. (1999), Truman (2003), Bernanke and Woodford (2005), and Mishkin and
Schmidt-Hebbel (2007a, b) provide detailed discussions of how central banks conduct IT and
how to improve the framework and institutions of monetary policymaking. Walsh (2009) surveys
recent evidence on the effects of IT on macroeconomic performance.
Measuring the effectiveness of IT requires a careful definition of the objectives of
monetary policy. A narrow view focuses exclusively on the inflation rate itself, which may reflect
the mandate given to the central bank by the government. Expanding this view a bit, we may
judge the central bank’s IT policy based on its ability to reduce inflation and its variability. More
broadly, the Federal Reserve, for example, operates under a dual mandate of stable prices as well
as high employment and stable growth. Thus, a broad view of IT effectiveness focuses on
1
The title of Grieder (1987) “The Secrets of the Temple” give a picture of mystery about central bank operations.
Current thinking largely overturns that view of a secretive central bank operation.
2
lowering inflation and its variability as well as raising output growth and lowering its variability.
Of course, with multiple goals, the monetary policy maker may face trade-offs between the
various goals when it implements its IT policy.
In addition to the goals of monetary policy, the real world of policy making requires that
we evaluate the attainment of goals over time. That is, macro and monetary economists have
long argued that monetary policy affects prices with a lag. In classic studies, Friedman (1961,
1972) and Friedman and Schwartz (1982) find that prices respond to monetary changes over a
long time period in the US and the UK. Friedman and Schwartz (1982, p. 412) report a long-run
one-for-one response of inflation to an increase in money growth, with most of the response
occurring within four years for both countries. Two decades later, Batini and Nelson (2002)
reaffirm these results, showing that it takes one to four years between changes in monetary
policy and the resulting change in inflation. This result persists despite changes in monetary
policy arrangements in the two countries.
Regarding the attainment of multiple goals, Svensson (1999, 2010) suggests flexible IT,
where the central bank strives not only to stabilize inflation around the inflation target but also to
stabilize the real economy. In the theoretical literature, the typical modeling strategy for optimal
monetary policy adopts a central bank objective function of a quadratic loss function, where the
two components of the loss function equal deviations of the inflation rate from its target as well
as deviations of real GDP from its target. Time plays a crucial role in a flexible IT regime.
Svensson (1997, 1999) demonstrates theoretically that when policy makers also target output
fluctuations, gradual adjustment of the intermediate inflation target to the long-run goal is
optimal. Bernanke et al. (1999) conclude that “output and employment remain concerns of
policy-makers after the switch to inflation targeting can be seen in the fact that all the targeting
3
countries have undertaken disinflation only gradually, to avoid putting undue pressure on the real
economy.” (p. 291). They describe a two-year lag between this monetary policy and its effect on
inflation as a common estimate (p. 320). The time necessary for the central bank to achieve its
inflation target may depend on the weight assigned to output stabilization. Smets (2003) shows
in the Euro area that when society puts equal weight on inflation stabilization and output gap
stabilization, the optimal policy horizon for maintaining an inflation target equals 4 years. An
increasing (decreasing) weight on output implies that the optimal policy horizon becomes longer
(shorter) and the central bank moves more (less) gradually. This issue becomes more
complicated in today’s worldwide economic recession, originating in the US subprime mortgage
market and the run up in energy and food prices. The IT countries cannot place too much
emphasis on inflation, potentially at the expense of economic recovery.
The effectiveness of IT policy garners little support based on evidence from developed
countries. Romer (2006) describes the view of inflation targeting due to Anna Schwartz as
“conservative window dressing.” That is, IT makes no difference in actual economic
performance between IT and non-inflation targeting (NIT) countries, where the NIT countries
commit, nonetheless, to reducing inflation. Those that disagree with this view argue that It
matters as it anchors inflation expectations, generating better economic performance. Much more
support exists for developing countries. One approach to the question employs panel data and
fixed-effect type estimation.
In an early analysis, Ball and Sheridan (2005) employ cross-section difference-indifferences ordinary least squares estimation to compare economic improvements in seven
OECD IT countries to thirteen OECD NIT countries. They discover that after countries adopt IT,
the level and variability of inflation and output growth of these countries do improve.
4
Non-targeting countries, however, also experience these improvements around the same time.
They argue that better economic performance reflects factors other than the monetary regime and
conclude that IT does not produce a major effect. They also note, however, that adopting IT does
not appear to affect the economy negatively. 2
Batini and Laxton (BL, 2007) and Gonçalves and Salles (GS, 2008) apply the
Ball-Sheridan method to test whether the adoption of IT affects inflation and output growth of
developing countries. They show that IT countries lower average inflation (BL and GS), inflation
volatility (BL), and output growth volatility (GS) relatively more than non-inflation targeting
(NIT) countries. A more recent paper by Brito and Bystedt (2010) uses panel-data difference and
system generalized method of moments (GMM) analysis for 46 developing countries, showing
that IT lowers inflation, but at the cost of lower output growth. Additionally, reductions of
inflation and output growth volatility prove insignificant. Their findings cast some doubt on the
effectiveness of IT for developing countries.
A second approach applies propensity score matching to estimate treatment effects (i.e.,
adoption of IT). Vega and Winkelried (2005) apply matching methods to evaluate treatment
effects of IT for 109 developed and developing countries jointly, and find that IT countries
reduce the level and volatility of inflation, when compared to NIT countries. Lin and Ye (2007)
use the propensity score matching method to evaluate the treatment effects, if any, from adopting
IT in seven industrial countries with a control set of 15 NIT industrial countries. They discover
no significant effects on inflation and its variability, supporting the “window-dressing” view of
IT. Walsh (2009) uses the same method and sample data as Lin and Ye (2007), finding that IT
does not significantly affect output growth or its variability. Lin and Ye (2009) extend their
2
Based on Markov-switching estimates, Dueker and Fischer (1996, 2006) provide comparative analysis for six
developed countries (Australia, Canada, Germany, New Zealand, the UK and the US) and conclude that no clear
evidence supports different inflation performance between IT and NIT countries.
5
propensity score matching model for developed countries to evaluate the IT treatment effect in a
sample of 52 developing countries and discover that IT significantly lowers inflation and its
variability.
The empirical findings produce mixed evidence in economic performance for developed
and developing countries. Different methodologies may lead to different findings that are
difficult to compare directly. The evidence does suggest that adopting IT produces little, if any,
effect in developed countries, but significant improvements in economic performance in
developing countries.
Since time lags in the effect of monetary policy generally imply different effects at
different times after policy adoption, ignoring such effects may contribute to the mixed
conclusions in the existing literature. Also, developed and developing countries may exhibit
different time profiles when adopting IT due to different fiscal, financial, and monetary
institutions with different inflation histories.
Fang, Miller, and Lee (2012) compare the effects of IT on inflation and output growth
and their volatilities in eight developed countries and 13 developing countries, addressing
short-run and long-run treatment effects. The propensity scoring exercise sheds light on different
effects of IT over time in developed and developing countries. First, they find that IT lowers
inflation rates for all IT countries. This effect decays and becomes insignificant in the short run
in developed countries, but remains significant in developing countries. Second, short-run costs
emerge in reduced output growth as well as increased inflation and output growth variabilities in
developed countries, but no such costs occur in developing countries. Third, while the short-run
costs disappear over time in developed countries, inflation, inflation variability, and output
growth variability remain lower from the short-run to the long-run in developing countries.
6
Further, output growth does not significantly change in the short- or long-run for developing
countries. In sum, developing countries gain more from IT policy than do developed countries.
Eren, Miller, and Fang (2012) apply the synthetic control method of Abadie and
Gardeazabal (2003) and Abadie et al. (2010) to the UK and Chile. The synthetic control method
generalizes the standard difference-in-differences (fixed-effect) method in that it allows
time-varying unobserved cofounder variables whereas the difference-in-differences method
imposes constancy on such cofounder variables. They find that IT did not effectively reduce
inflation in the UK but did in Chile.
The basic problem in assessing the effectiveness of IT that all methods try to address
involves the construction of a counterfactual that represents the outcomes in the IT adopting
country, if that country did not adopt IT. The fixed-effect methods implicitly assume that the
sample of countries not adopting IT stand as the counterfactual. The propensity score matching
and the synthetic control methods each try to construct the counterfactual based on a weighted
average of a few selected countries in the sample of those not adopting IT. The researcher selects
the countries to enter the weighted average based on a screening method that evaluates the
closeness of these NIT countries to the IT adopting countries over some economic criteria. The
synthetic control method generalizes the propensity score matching method in that it permits
time-varying unobserved confounded variables.
The prior discussion focuses on empirical attempts to measure the effectiveness of IT on
improving economic performance – lowering the average inflation rate, raising average rate of
output growth, and lowering inflation and/or output growth volatility. Another, related literature
examines the successfulness of IT in reducing inflation persistence. Benati (2008), Gerlach and
Tillmann (2012), Levin and Piger (2006), O’Reilly and Whelan (2005), Siklos (1999), and
7
Tillmann (2012) provide mixed evidence as to whether IT reduces the persistence in inflation
rates.
The rest of the paper is organized as follows. Section 2 reviews the literature on the
effectiveness of inflation targeting, describing the various methods as well as their pluses and
minuses. Section 3 concludes.
2.
Effectiveness of IT: Literature Review
Monetary economists typically model the policy process as follows. The central bank chooses its
operating targets to optimize its objective function subject to the macroeconomic model of the
economy. The central bank objective function usually includes the deviations of inflation and
output from certain target levels. Hence, they derive central bank reaction functions or monetary
policy rules that describe how central banks alter their policy in response to macroeconomic
changes. The most frequently used policy instrument is the short-term interest rate and, therefore,
“monetary policy rules” typically mean “interest rate rules” (Fourcans and Vranceanu, 2004).
Interest rate rules depend on the deviations of a set of macroeconomic variables, such as
the inflation rate and output, from their target values. Taylor (1994) shows that monetary policy
in the U.S. conforms to a simple monetary policy rule, in which the short-term interest rate
adjusts according to inflation and real output deviations from target inflation and potential output,
respectively, and closely follows observed movements in that interest rate. When the inflation
rate exceeds (falls below) its target, the rule recommends an increase (decrease) in the interest
rate. This term captures the goal of the central bank to achieve price stability. When the output
gap is positive (negative), the Taylor rule recommends an increase (decrease) in the interest rate.
Therefore, adjustments of the interest rate, vis-à-vis the output gap, reflect policy actions
designed to preempt an anticipated rise in inflation.
8
We can write the original form of the Taylor rule as follows:
it = π t + r + α (π t − π * ) + β xt ,
(1)
where it is the target nominal interest rate,
π t is the inflation rate, r is the long-run
equilibrium real interest rate, π * is the target for inflation, and xt is the output gap.
The adoption of an inflation targeting means setting π * as the target of monetary policy.
The Taylor rule represents a possible method for implementing monetary policy by controlling
the short-term interest rate to achieve the target inflation rate.
Over 15 years of empirical analysis of the effectiveness of inflation targeting does not yet
reveal a final, strong consensus. Researchers face a problem in trying to identify whether the
adoption of inflation targeting proves effective. To wit, the period of adoption of inflation
targeting, which begins in the early 1990s, also corresponds with the Great Moderation, whereby
inflation rates and their variability fall. Researchers implement various techniques to try and
evaluate the effectiveness of inflation targeting – difference-in-differences estimation, fixed- and
random-effects estimation, propensity score matching, synthetic control method, and so on. The
following subsections review the econometric approaches to the IT effectiveness question,
describe the pluses and minuses of the different methods, and discuss the general findings of the
analyses.
Difference in Differences
Ball and Sheridan (2004) employ difference-in-differences estimation, using a sample of
developed countries. Of the 20 countries considered, seven adopted IT between 1990 and 1995.
For countries that adopt IT, they calculate average values of the economic variables in the preand post-targeting periods. For the countries that do not adopt IT, they compute the average
values for hypothetical pre- and post-targeting periods. The break-point time between pre- and
9
post-targeting equals the average time of adoption of IT by those countries in the targeting group
of seven.
Ball and Sheridan (2004) perform two different sets of regressions. The first set adopts
the following specification:
xi , post − xi , pre =
α 0 + α1 Di + ei ,
(2)
where xi,post (xi,pre) equals country i’s value of x in the post- (pre-) targeting period and Di equals a
dummy variable that equals one if the country targets inflation and zero otherwise. They
experiment with different pre-targeting periods that began in 1960 (i.e., the long “pre” period)
and 1985 (i.e., the short “pre” period).
Ball and Sheridan (2004) note that the average inflation rate of IT countries exceeds that
of NIT countries during the pre-targeting period. Thus, those that adopt IT may experience a
reduction in inflation merely because they face more room to improve, a problem of regression
to the mean. To address this problem, they run a second set of regressions that introduces the
average inflation rate in the pre-targeting period. That is, they estimate the following
specification:
xi , post − xi , pre =
α 0 + α1 Di + α 2 xi , pre + ei .
(3)
They set the variable x to include the average inflation rate and its variability, the output
growth rate and its variability, the long-term interest rate, and the variability of the short-term
interest rate. Occasionally, a significant coefficient on the dummy variable implies that inflation
targeting improved economic performance, but such significant effects only appear in equation
(2) that does not control for regression to the mean. The coefficient of the dummy variable in
equation (3) does not test significant in any of the regressions examined. Ball and Sheridan (2004)
conclude that their analysis provides no support for IT improving economic performance
10
amongst their sample of developed countries.
Using the difference-in-differences method of Ball and Sheridan (2204), Batini and
Laxton (BL, 2007) and Gonçalves and Salles (GS, 2008) test whether the adoption of IT affects
inflation (BL and GS) and its variability (BL) as well as output growth volatility (GS) for slightly
different samples of NIT developing countries, respectively, against 13 IT countries. In other
words, Batini and Laxton (BL, 2007) and Gonçalves and Salles (GS, 2008) consider whether the
IT-ineffectiveness conclusion of Ball and Sheridan (2005) also holds for developing countries,
who experience more difficulties with inflation.
The BL (2007) sample of 42 countries runs from 1985 through 2004. The time break
point for the NIT countries occurs in 1999:Q4, using the average of the time breaks for the IT
countries that ran from 1997:Q2 to 2002:Q1. BL estimate a transposition of equation (3) that
adjusts for regression to the mean. The exact specification is as follows:
=
xi ,t φ[α T Di ,t + α N (1 − Di ,t )] + (1 − φ ) xi ,t −1 + ei ,
(4)
where φ equals the speed with which x returns to its group alpha. This specification provides
more intuitive insight as to the regression to the mean. To wit, when φ equals zero, x depends
only on its history, while when φ equals one, x reverts to its group alpha in one period. BL find
that adopting IT significantly reduces the average inflation rate and the volatility of inflation at
the 5-percent level. They do not find significant effects of adoption of IT on the volatility of the
growth rate of output, although the coefficient is negative. 3 BL do not test the effect of IT
adoption on the real output growth rate.
The GS (2008) sample of 36 countries runs from 1980 through 2005. The time break
point for the NIT countries occurs in 1998, using the average of the time breaks for the IT
3
BL report a significant reduction in the volatility of the output gap for IT countries.
11
countries that runs from 1991 to 2002. GS estimate equation (3) that adjusts for regression to the
mean and find that adopting IT significantly reduces the average inflation rate at the 10-percent
level, and significantly reduces the volatility of GDP growth volatility at the 5-percent level.
They do not find significant effects of adoption of IT on inflation volatility. GS also do not test
the growth rate of output.
The difference-in-differences method faces at least one problem in the current
implementation. The effectiveness of adopting IT theoretically requires a comparison to the
counterfactual of how the IT-adopting economy would have performed in the absence of IT. The
difference-in-differences method implicitly takes the actual performance of the average economy
from the list of NIT countries in the sample. Moreover, the necessity of dividing the NIT
countries experience into pre- and post-IT adoption, when they do not, in fact, adopt it creates
another arbitrary assumption that can affect the outcome. On the other hand, the
difference-in-differences method is a well-established procedure for analyzing the effect of a
treatment.
Fixed Effects
We can interpret the difference-in-differences method as panel data estimation, where the panel
includes many countries and two time periods – pre- and post-IT adoption. A few researchers
implemented panel evaluations of IT adoption. Wu (2004) and Willard (2006) estimate
differenced GMM estimations for developed countries. Brito and Bystedt (2010) consider a
sample of developing countries, using differenced and system GMM estimation.
Wu (2004) and Willard (2006) demonstrate that adopting IT does not produce any
significant improvement in economic performance. That is, the adoption of IT captures a
“window dressing effect.” Brito and Bystedt (2010) show that IT lowers inflation, but at the cost
12
of lower output growth. Additionally, neither the reduction of inflation volatility nor of output
growth volatility proves significant.
Wu (2004), Willard (2006) and Brito and Bystedt (2010) all include the lagged inflation
rate to capture the regression to the mean effect, which then allows for the adoption of IT as a
window dressing effect. Implicitly, these authors assume that the adoption of IT does not affect
the speed of regression to the mean. That is, as currently models, these authors only test for a
significant IT effect on the constant term in the regression. They could test for the effect of IT on
the slope coefficient of the lagged inflation effect.
Propensity Score Matching
Vega and Winkelried (2005), Lin and Ye (2007), and Lin and Ye (2009) use the propensity score
matching method to evaluate the treatment effects, if any, from adopting IT. Consider the average
treatment effect on the treated (ATT) of IT that depends on the following equation:
ATT =
E[Yi1 Di =
1] − E[Yi 0 Di =
1] ,
(5)
where Di = 1 (= 0) denotes the treatment (non-treatment) state when country i adopts (does
not adopt) IT. Thus, ( Yi1 Di = 1 ) equals the value of the outcome (e.g., the inflation rate) actually
observed in the targeting country and ( Yi 0 Di = 1 ) equals the counterfactual outcome that would
occur, if the targeting country does not adopt the policy. Two important issues arise in this
equation. First, we cannot observe the second term in the ATT. We do not know the inflation rate
of the targeting country, absent such a policy. Second, the first term assumes implicitly that once
the binary variable switches from 0 to 1, the inflation rate adjusts instantaneously. This
specification provides no room for a lag effect when implementing the targeting policy or for
differing magnitudes of effects over time.
The existing literature initially developed the propensity score matching methods to
13
address the first issue. 4 The matching method chooses a non-targeting control group of countries
to mimic a randomized experiment to reduce the bias in the estimation of the treatment effects
with observational data sets. Empirically, studies replace E[Yi 0 Di = 1] with E[Yi 0 Di = 0, X ] ,
which is observable. That is, under the conditional independence assumption, the ith country’s
outcome (the inflation rate, Yi1 or Yi 0 ) does not depend on the targeting policy chosen
conditional on a set of explanatory variables (X). Rosenbaum and Rubin (1983) propose probit
(or logit) models to estimate propensity scores, which measure the probabilities that countries i
and j adopt IT policy, given X, to match the targeting countries (i) and control countries (j). In the
selection process, the common support condition, P ( D = 1 X ) < 1 , holds to ensure that analogous
non-treatment units exist to compare with the treated ones.
Using propensity score matching, Vega and Winkelried (2005), Lin and Ye (2007, 2009),
and Walsh (2009) estimate the ATT of equation (5) as follows:
ATT =
1
N

∑ Y
i

i
−
∑ w( p , p
i
j∈S p
j

)Y j  ,

(6)
where pi and p j equal the propensity scores for observation i in the targeting group and j in
the control group, respectively. N is the number of observations of the targeting group in the
sample. S p is the region of common support. w( pi , p j ) is the weight given to observation j
when matched to observation i. Thus, this specification measures the long-run ATT during IT.
Lin and Ye (2007) use the propensity score matching method to evaluate the treatment
effects, if any, from adopting IT in seven industrial countries with a control set of 15 NIT
industrial countries. They show no significant effects on inflation and its variability, supporting
4
Caliendo and Kopeinig (2008) provide an excellent review and practical guide for implementing the matching
estimator.
14
the “window-dressing” view of IT. Walsh (2009) uses the same method and sample data as Lin
and Ye (2007), but extends the analysis to consider additional variables, finding that IT does not
significantly affect output growth or its variability. He also supports Lin and Ye’s findings that IT
does not significantly affect inflation and its variability. Lin and Ye (2009) extend their
propensity score matching model for developed countries to evaluate the IT treatment effect in a
sample of 52 developing countries and discover that IT significantly lowers inflation and its
variability.
Vega and Winkelried (2005) apply matching methods to evaluate treatment effects of IT
for 109 developed and developing countries jointly, and find that IT countries reduce the level
and volatility of inflation. They report results for the entire sample and the sample divided into
developed and developing countries. They also introduce a measure of regression to the mean to
accommodate the concern raised by Ball and Sheridan (2005). They conclude that the adoption
of IT does reduce inflation and its variability for the entire sample and for the two sub-samples of
developed and developing countries, although the effects for developing countries are larger.
Moreover, regression to the mean does exert an important effect, but their more inclusive sample
of IT and NIT countries still shows better performance by IT countries, even for their sub-sample
of developed countries.
The second issue motivates the estimation of the ATT immediately after the policy
adoption, using the strategy of De Loecker (2007). Fang, Miller, and Lee (2012) implement the
De Loecker (2007) method for IT. The implementation modifies equation (6) to determine the
individual value of ATT at every time τ as follows:
individual
ATT
=
τ

1 
∑ Yiτ − ∑ w( pi , p j )Y jτ  ,
j∈S p
Nτ i 

15
(7)
where τ = {1, 2, …, T} denotes the τth year after the policy adoption. 5 The technique also
estimates the cumulative ATT from 1 to τ as follows:
ATTtCumulative
=
 tt

∑  ∑ Yit − ∑ ∑ w( pi , p j )Y jt  .
=t 1 j∈S p

∑ N t i =t 1
1
t
(8)
t =1
where τ = {1, …, T}. This estimator provides a cumulative effect of IT in the short-run to the
long-run time frame.
Developed country findings
For the treatment effect on inflation, Fang, Miller, and Lee (2012) find a significant negative
effect emerges in the first year after adoption. The inflation gap shrinks in the second year and
widens in the third year, although none of these effects prove significant. The evidence from
matching suggests that IT lowers inflation in only the first year after the policy adoption. The
second, third, and other years exhibit negative, but insignificant effects. Significant cumulative
effects disappear after three years.
The treatment effect on inflation variability increases significantly in the first and second
years after adoption. In the third and the fourth years, the effect falls to small levels
quantitatively, nearly half negative, although none prove significant. Thus, no beneficial effect of
IT emerges for inflation rate variability. Significant cumulative effects disappear after four years.
Conceptually, under an IT framework, the central bank places increased weight on
inflation stabilization and reduced weight on real economic stability. Thus, a trade-off occurs
between the inflation and output growth rates, or the output cost of lowering inflation,
particularly, in the short-run. Hutchison and Walsh (1998) find that the short-run output cost of
5
The selection of the lag length seems somewhat arbitrary, since we do not know exactly the weight the targeting
countries put on inflation stabilization or other objectives. Smets (2003) shows that the optimal policy horizon
equals four years when inflation and output stabilization receive equal weights.
16
disinflation in New Zealand started to rise in the early 1990s around the time of the central bank
reform. Once the central bank’s disinflationary policy obtains credibility, however, it may receive
a credibility bonus that should reduce the output cost of lowering inflation. Gonçalves and
Carvalho (2009) show that inflation targeters suffer smaller output losses during disinflations
when compared to non-targeters. Benhabib and Spiegel (2009) provide long-term evidence of a
threshold effect for the relationship between inflation and output growth. That is, for inflation
rates under (above) 3.23%, the correlation between inflation and output growth is positive
(negative).
Fang, Miller, and Lee (2012) follow the same procedures to evaluate the treatment effect
on output growth. Targeters experience significantly lower output growth in the first year after
adoption. The credibility bonus emerges in the second year, where the negative output growth
rate falls, but the decrease is insignificant. The targeters enjoy higher but insignificant, output
growth in the third and fourth years after adoption. Significant cumulative effects disappear after
two years.
Conventional thinking of the Phillips-curve tradeoff between the inflation rate and the
output gap focuses on levels. Taylor (1994) argues that the policy tradeoff more appropriately
relates to a tradeoff between the variabilities of the output growth and inflation rates. Fuhrer
(1997) demonstrates that the short-run tradeoff between the inflation and output growth rates
implies a long-run tradeoff between their variabilities. The optimal monetary policy (that
minimizes variability of the central bank’s targets of the level of inflation and the level of real
output relative to potential) implies dramatic increases in the output growth rate variability, when
policy attempts to make the inflation rate variability too small. His empirical results suggest that
balanced responses to inflation and output are consistent with balanced preferences over inflation
17
and output variability. Cecchetti and Ehrmann (2002) observe that while the variability of
inflation falls more in the IT countries than in NIT countries, output variability falls far less in
the former than in the latter. When the IT countries increase their revealed aversion of inflation
variability, they suffer increases in output volatility. Erceg (2002) argues that IT reflects the
perceived monetary policy frontier of the economy, the policymaker’s tradeoff between the
volatilities of inflation and real activity. Adopting a narrow inflation target range may induce
considerable volatility in real activity. Arestis et al. (2002) report mixed evidence for individual
targeting countries. The adoption of inflation targets results in a more favorable monetary policy
tradeoff in New Zealand, the UK, and Sweden, meaning a substantial decrease in the output gap
volatility for a given inflation volatility. No change occurs in Canada, and a decrease in the
inflation rate variability accompanied by an increase in output gap volatility across Australia and
Finland. When these authors compare the ratio of output gap volatility to inflation volatility
between IT and NIT countries, the ratio in the NIT countries exceeds that in the IT countries.
Fang, Miller, and Lee (2012) find that the output growth variability increases sharply and
significantly in the first two years after IT begins. The variability becomes even larger at the end
of second year. In the third and fourth years, they find positive, but insignificant, effects.
Significant cumulative effects disappear after four years.
Thus, Taylor’s inflation and output growth volatility trade-off shifts toward a less
advantageous trade-off in the first and second years after the adoption of IT. That is, both
inflation and output growth variability increase in the short run with IT. On a cumulative basis,
that shift disappears in a significant sense in the following years.
Developing country findings
Fang, Miller, and Lee (2012) report that IT decreases the inflation rate significantly in all four
18
future periods considered for developing countries. While the effects from one year to the next
do not move monotonically, in each case the largest negative effect occurs in the fourth year.
Significant and large lagged effects of IT on inflation emerge in developing countries as
compared to the effects in developed countries. Moreover, significant cumulative effects remain
through the end of the sample, implying that IT countries experience lower inflation after
adopting IT.
In addition, they find that IT significantly decreases the inflation rate variability in the
four years after adoption. Again, long time-lag effects exist. Moreover, significant cumulative
effects remain through the end of the sample. Targeters, however, experience no significant
differences in the growth rate for the four future years considered. That is, no substantial cost of
output growth occurs in the process of disinflation, even for the cumulative effects. Finally, in
the first two years after adoption, IT adoption decreases output growth variability, although
generally insignificant. Then in the third and fourth years after adoption, IT significantly lowers
output growth variability and the largest effect occurs in the fourth year. In addition, significant
cumulative effects remain through the end of the sample. These findings suggest that no costs
and only benefits occur for developing countries who adopt IT. The benefits, however, occur
with a time lag.
Therefore, Taylor’s inflation and output growth volatility trade-off shifts toward a more
advantageous trade-off in the third and fourth years after the adoption of IT since both inflation
and output growth variabilities decrease with IT. This finding also emerges for the cumulative
effects through the end of the sample.
Synthetic Control
Abadie and Gardeazabal (2003) and Abadie et al. (2010) develop a data-driven, synthetic-control
19
procedure to consider the effects of policy intervention. That is, they consider case studies where
typical econometric techniques do not apply, because only one (or a few) unit(s) receive a
treatment. Their method constructs a weighted combination of untreated (control) units to
provide a good counterfactual for the treated unit(s). The more traditional methods of
difference-in-differences and propensity score matching models allow for the presence of
time-invariant unobserved confounder variables. Both methods eliminate these unobserved
confounder variables. The synthetic control method, however, permits time-varying unobserved
confounder variables.
Consider J+1 countries and assume that the intervention (e.g., adopting IT) only affects
one country after time t. Let YitN denote the observed outcome for any country i (i = 1, …, J+1)
at time t (t = 1, …, T) in the absence of the intervention and T0 denote the number of
pre-intervention periods such that 1 ≤ T0 < T . Let Y jtI denote the j-th country’s outcome after
the intervention of interest in periods T0 to T. That is, the intervention does not affect the counties
prior to the implementation period t ∈ {1, …, T0} and all i ∈ {1, …, J+1}. That is, YitN = YitI .
Further, we assume that the treatment does not affect the outcomes of the untreated countries.
Given the above, we can express the effect of the intervention on the treated country as
follows:
α=
YitI − YitN .
it
(9)
Borrowing the standard terminology from the treatment literature, we can write the observed
outcome for the treated country as follows:
I
Y=
YitN + α it Dit ,
it
(10)
where Dit is an indicator that takes on the value of one if country i at time t receives treatment;
20
zero, otherwise.
We enumerate the only country exposed continually to the policy interventions after T0
( 1 ≤ T0 < T ) as country one. Then, we can write that
=
1 if i 1 and t > T0 
Dit = 

 0 otherwise 
(11)
Prior to the intervention, each country experiences the outcome with no treatment. After
the intervention, country one now experiences the outcome with treatment while each of the
other countries still experiences the outcome with no treatment. We need to measure the
unobservable Y1tN to estimate α1t for t ∈ {T0+1, …, T} in equation (9). We do observe Y1tN for t
∈ {1, …, T0} as well as YitN for i = 2, …, J+1 and for t ∈ {1, …, T0}. Thus, we construct a
model that can explain Y1tN from the values of YitN and other explanatory variables. With that
model, we then can construct the counterfactual values for Y1tN for t ∈ {T0+1, …, T}.
Abadie et al. (2010) assume that a factor model explains Y1tN as follows:
ft + βt X i + γ t fi + ε it ,
Y1tN =
(12)
where φt denotes an unknown time-varying common factor with constant factor loadings, Xi
denotes a vector of observed covariates with corresponding time-varying βt vector of unknown
parameters, γt denotes a time-varying vector of unobserved common factors with the
corresponding vector fi of unknown factor loadings, and εit equal the unobserved transitory
shocks with mean zero.
Consider a vector of weights W = (w2, …, wJ+1) such that each weight is positive and the
weights sum to one. Each of the possible set of weights W represents a potential synthetic control.
That is, we apply the weights to the countries that do not experience an intervention during the
21
pre-intervention period such that
J +1
J +1
=j 2
=j 2
J +1
J +1
ft + βt ∑ w j X j + γ t ∑ w j f j + ∑ w jε jt .
∑ w jY jt =
(13)
=j 2=j 2
The optimal weights, if they exist, produce the following outcomes:
J +1
J +1
J +1
=j 2=j 2
=j 2
*
, and ∑ w*j X j X 1 ,
Y1T=
=
∑ w*jY j1 Y=
11 , ..., ∑ w jY jT0
0
(14)
where W * = ( w2* , ..., w*J +1 ) equal the optimal values. Abadie et al. (2010) show that under
reasonable conditions, the following result emerges;
J +1
Y1tN − ∑ w*jY jt ≈ 0 ,
(15)
j =2
which implies that we can approximate the outcome variable in the first country in the absence of
the treatment or intervention for any time t with a synthetic control country, a weighted average
of the control countries with the optimal weights defined by equation (14).
Therefore, we can approximate the estimate for the effect of intervention as follows:
J +1
αˆit = YitI − YitN = YitI − ∑ w*jY jt for t ∈ {T0 + 1, ..., T } .
(16)
j =2
Abadie et al. (2010) demonstrate that equation (15) holds exactly if and only if {Y11 , ..., Y1T0 , X 1' }
belongs to the convex hull of [{Y21 , ..., Y2T0 , X 2' }, ..., {YJ +11 , ..., YJ +1T0 , X J' +1} . Absent this
condition, the approximation in equation (15) still holds for optimal weights. Now, the issue is
how close an approximation is equation (15). The researcher, however, can calculate the size of
the discrepancy and determine whether the characteristics of the exposed country lead to a
sufficient match.
The more traditional methods of difference-in-differences and propensity score matching
models allow for the presence of time-invariant unobserved confounder variables. Both methods
22
eliminate these unobserved confounder variables. The synthetic control method, however,
permits time-varying unobserved confounder variables. The synthetic control approach leads to
the calculation of the following values:
J +1
J +1
*
f1 .
=
∑ w*j X j X=
1 , and ∑ w j f j
(17)
=j 2=j 2
Using these values will produce an unbiased estimate of Y1tN . But, we cannot implement the
synthetic control method in this way, since we cannot observe (f1, …, fJ+1). As Abadie et al.
(2010) show, equation (12) implies that a synthetic control can fit X1 and the pre-intervention
outcomes if and only if it fits X1 and f1. Thus, equation (17) holds.
Eren, Miller, and Fang (2012) use the synthetic control method to evaluate the
effectiveness of IT for one developed (e.g., the UK) and one developing (e.g., Chile) country.
They find little evidence that IT significantly reduced the inflation rate in the UK, but strong and
significant evidence that IT reduced the inflation rate in Chile.
3.
Conclusion
Several reasons may explain why studies find or do not find significant effects of IT on inflation
and its variability or output and its variability. First, and most obvious, the specific countries
examined influence the outcomes. On the one hand, Ball and Sheridan (2005), Lin and Ye (2007),
and Walsh (2009) show that the available evidence for a group of developed countries does not
support the view that adopting IT brings the inflation rate and its variability down or that it
affects the output growth rate and its variability. On the other hand, Vega and Winkelried (2005)
with a larger sample find significant effects of IT for a sample of developed and developing
countries both in a combined and in separate sub-samples, although the developed country effect
is smaller. Further, Gonçalves and Salles (2008) and Lin and Ye (2009) find that developing
countries can significantly lower both the inflation rate and its variability by adopting IT. These
23
differences may reflect different inflation performances and, therefore, the motivation to adopt
this monetary policy in developed and developing countries. Developed countries experience
much lower mean inflation rate (variability) than developing countries in the full sample and the
pre-IT period. Since the monetary authorities choose IT to maintain their already low inflation
rates or to converge to a lower rate, rather than to squeeze high inflation rates down, researchers
find no significant effects for developed countries. The OECD developed countries who adopt IT
generally match the inflation performance of other OECD developed countries. In contrast,
studies find significant negative effects for developing countries, where policy attempts to reduce
a high inflation rate to achieve real results of lower inflation. The developing country targeters
consistency experience better inflation performance than their non-targeter partners.
Second, different time horizons may lead to different findings. Lin and Ye (2007) and
Walsh (2009) use a long sample of data (15 years from 1985 to 1999) to evaluate IT and find no
significant effects for seven developed countries in the OECD. Lin and Ye (2009), on the other
hand, use a long sample of data (21 years from 1985 to 2005) and find significant effects for
thirteen developing countries. Fang, Miller, and Lee (2012) report similar findings over an even
longer period (23 years from 1985 to 2007) for both developed and developing countries.
Considering the lagged effects of this monetary policy, they show that for developed-country
targeters, the inflation gain, output growth loss, and the inflation and output growth rate
variabilities increase in the years after the adoption of IT. These three costs of disinflation
disappear over time, suggesting only short-lived effects of IT adoption. A different story emerges,
however, for developing countries. Now, they show that for developing-country targeters, the
inflation gain in the four-year short-run and the medium term occurs without output cost, and the
inflation and output growth rate variabilities decrease in the short-run as well as the medium
24
term, suggesting that the effects of IT depend on time. The reduction in the inflation rate and its
variability as well as the reduction in output growth rate variability also continue to exist in the
long run.
Third, and most important, the evidence of treatment effects needs careful interpretation,
when evaluating effectiveness of IT. The policy irrelevance or “window-dressing” view generally
comes from the long-run treatment effect for developed countries. Ball and Sheridan (2005) note
that the inflation rate declines in the inflation targeting era for both targeters and non-targeters.
Fang, Miller, and Lee (2012) report short-run and long-run treatment effects, suggesting that IT
does matter if targeters simply want to keep low the inflation rate as in other developed countries
in the OECD (e.g., the US and Japan). 6 IT lowers inflation rates immediately in the first few
years after the policy adoption. Targeters reach their target at the end of the second year and keep
inflation low to the end of our sample. In sum, developed countries can achieve their policy goal
in two years, which proves consistent with a one- to four-year lag in effect of monetary policy on
inflation reported in Friedman and Schwartz (1982), Bernanke et al. (1999), and Batini and
Nelson (2002). Inflation rates of targeters beyond the second year after the policy adoption prove
no different from those of non-targeters and, thus, no more treatment effects emerge. Moreover,
Fang, Miller, and Lee (2012) report that dynamic analyses for developed countries imply a lower
output growth and higher inflation and output growth variability trade off lower inflation in the
short-run, although not in the long-run.
In contrast, Fang, Miller, and Lee (2012) argue that developing countries experience
success of IT with significant intertemporal and cumulative treatment effects in the four-year
short-run as well as in the long-run. This success may reflect the following observations or views.
6
That is, the US and Japan did not need to adopt inflation targeting, since inflation did not seem a threat.
25
First, targeters who adopt IT want to substantially reduce inflation from a high to a low level.
The long-run treatment effect on the inflation rate shows this fact. These observations explain
why in previous studies, researchers find support for IT in developing countries. Second, as
suggested by Bernanke et al. (1999) and Svensson (1997, 1999), targeters choose to reach their
inflation targets gradually to avoid costs of disinflation. Thus, developing countries require a
much longer time (at least longer than the two years for developed countries) to fulfill this
monetary policy.
Subsequent analysis reveals that the policy-irrelevance conclusion does not prove robust
to short-run analyses for developed countries and more information emerges in the four
outcomes at different times for developing countries. This result demonstrates the
misspecification (misinterpretation or missed information) of the treatment estimates, if
researchers neglect the dynamic adjustment process of policy adoption.
Eren, Miller, and Fang (2012) use the synthetic control method of Abadie and
Gardeazabal (2003) and Abadie et al. (2010), finding that IT does not significantly reduce
inflation in the UK, but does significantly reduce inflation in Chile.
References:
Abadie, A., Diamond, A., and Hainmueller, J., 2010. Synthetic Control Methods for Comparative
Case Studies: Estimating the Effect of California's Tobacco Control Program. Journal of
the American Statistical Association 105, 493-505.
Abadie, A., and Gardeazabal, J., 2003. The Economic Costs of Conflict: A Case Study of the
Basque Country. American Economic Review 93, 113-132.
Acemoglu, D., Johnson, S., Querubin, P., and Robinson, J. A., (2008). When does policy reform
work? The case of central bank independence. Brookings Papers on Economic Activity,
351-417.
Arestis, P., Caporale, G. M., and Cipollini, A., (2002). Does inflation targeting affect the trade-off
between output gap and inflation variability. Manchester School 70(4), 528-545.
26
Ball, L., and Sheridan, N., (2005). Does inflation targeting matter? In B. S. Bernanke and M.
Woodford (eds.), The Inflation Targeting Debate. Chicago: University of Chicago Press,
249-276.
Batini, N., and Nelson, E., (2002). The lag from monetary policy actions to inflation: Friedman
revisited. International Finance 4, 381-400.
Batini, N., and Laxton, D., (2007). Under what conditions can inflation targeting be adopted?
The experience of emerging markets. In: F. Mishkin and K. Schmidt-Hebbel (eds.),
Monetary Policy under Inflation Targeting. Central Bank of Chile, Santiago, 1-38.
Benati, L., 2008. Investigating inflation persistence across monetary regimes. Quarterly Journal
of Economics 123, 1005–1060.
Benhabib, J, and Spiegel, M. M., (2009). Moderate inflation and the deflation-depression link.
Journal of Money, Credit and Banking 41, 787-798.
Bernanke, B. S., Laubach, T., Mishkin, F. S., and Posen, A., (1999). Inflation Targeting: Lessons
from the International Experience. Princeton University Press: Princeton.
Bernanke, B. S., and Woodford, M., (2005). The Inflation Targeting Debate. Chicago: University
of Chicago Press.
Blinder, A. S., Ehrmann, M., Fratzscher, M., Haan, J., and Jansen, D., (2008). Central bank
communication and monetary policy: A survey of theory and evidence. Journal of
Economic Literature 46(4), 910-945.
Brito, R. D., and Bystedt, B., (2010). Inflation targeting in emerging economies: Panel evidence.
Journal of Development Economics 91, 198-210.
Caliendo, M., and Kopeinig, S., (2008). Some practical guidance for the implementation of
propensity score matching. Journal of Economic Surveys 22, 31-72.
Cecchetti, S. C., and Ehrmann, M., (2002). Does inflation targeting increase output volatility? An
international comparison of policymakers’ preferences and outcomes, in N. Loayza and K.
Schmidt-Hebbel (eds.), Monetary Policy Rules and Transmissions, Santiago: Central
Bank of Chile, 247-274.
De Loecker, J., (2007). Do exports generate higher productivity? Evidence from Slovenia.
Journal of International Economics 73, 69-98.
Dueker, M. J., and Fischer, A. M., (1996). Do inflation targeters redefine central bank inflation
preferences? Results from an indicator model. in K. Alders, K. Koedijk, C. Kool, and C.
Winder (eds.), Monetary Policy in a Converging Europe. Dordrecht, Netherlands: Kluwer,
21-37.
27
Dueker, M. J., and Fischer, A. M., (2006). Do inflation targeters outperform non-targeters?
Federal Reserve Bank of St. Louis Review 88(5), 431-450.
Erceg, C. J., (2002). The choice of an inflation target range in a small open economy. American
Economic Review Papers and Proceedings 92(2), 85-89.
Eren, O., Miller, S. M., and Fang, W. S., (2012). Did inflation targeting work in the UK and
Chile? Case studies using the synthetic control method. Working Paper. Univeristy of
Nevada, Las Vegas.
Fang, W. S., Miller, S. M., and Lee, C. S., (2012). Short- and long-run differences in the
treatment effects of inflation targeting on developed and developing countries. Working
Paper, University of Nevada, Las Vegas.
Fatás, A., Mihov, I., and Rose, A. K., (2007). Quantitative goals for monetary policy. Journal of
Money, Credit and Banking 39, 1163-1176.
Faust, J., and Svensson, L. E. O., (2001). Transparency and credibility: Monetary policy with
unobserved goals. International Economic Review 42(2), 369-397.
Fourcans A. and Vranceanu R. (2004). The ECB interest rate rule under the Duisenberg
presidency. European Journal of Political Economy 20, 579-595.
Friedman, M., (1961). The lag in effect of monetary policy. Journal of Political Economy 69(5),
447-466.
Friedman, M., (1972). Have monetary policies failed? American Economic Review Papers and
Proceedings 62, 11-18.
Friedman, M., and Schwartz, A. J., (1982). Monetary Trends in the United States and the United
Kingdom: Their Relation to Income, Prices, and Interest Rates, 1867-1975. Chicago:
University of Chicago Press.
Fuhrer, J. C., (1997). Inflation/output variance trade-offs and optimal monetary policy. Journal of
Money, Credit and Banking 29(2), 214-234.
Gerlach, S., and Tillmann, P., 2012. Inflation targeting and inflation persistence in Asia–Pacific.
Journal of Asian Economics, in press.
Gonçalves, C. E. S., and Salles, J. M., (2008). Inflation targeting in emerging economies: What
do the data say? Journal of Development Economics 85, 312-318.
Gonçalves, C. E. S., and Carvalho, A., (2009). Inflation targeting matters: Evidence from OECD
economics’ sacrifice ratios. Journal of Money, Credit and Banking 41, 233-243.
28
Greider, W., (1987). Secrets of the Temple: How the Federal Reserve Runs the Country. New
York: Simon and Schuster, 1987.
Hutchison, M. M., and Walsh, C. E., (1998). The output-inflation tradeoff and central bank
reform: Evidence from New Zealand. Economic Journal 108, 703-725.
Issing, O., (2005). Communication, transparency, accountability: Monetary policy in the
twenty-first century. Federal Reserve Bank of St. Louis Review 87(2), 65-83.
Levin, A. T., and Piger, J. M. 2006. Is inflation persistence intrinsic in industrial economies?
Working Paper Series 334. European Central Bank.
Lin, S., and Ye, H., (2007). Does inflation targeting really make a difference? Evaluating the
treatment effect of inflation targeting in seven industrial countries. Journal of Monetary
Economics 54, 2521-2533.
Lin, S., and Ye, H., (2009). Does inflation targeting really make a difference in developing
countries? Journal of Development Economics 89, 118-123.
Mishkin, F. S., and Schmidt-Hebbel, K., (2007a). A decade of inflation targeting in the world:
What do we know and what do we need to know? In F. S. Mishkin (ed.) Monetary Policy
Strategy. The MIT Press, Cambridge, Massachusetts.
Mishkin, F. S., and Schmidt-Hebbel, K., (2007b). Monetary Policy under Inflation Targeting.
Banco Central de Chile, Santiago, Chile.
Neumann, M. J. M., and von Hagen, J., (2002). Does inflation targeting matter? Federal Reserve
Bank of St. Louis Review 84, 127-148.
O’Reilly, G., and Whelan, K., 2005. Has euro-area inflation persistence changed over time? The
Review of Economics and Statistics 87, 709–720.
Rosenbaum, P., and Rubin D., (1983). The central role of the propensity score in observational
studies for causal effects. Biomertrika 70, 41-50.
Siklos, P. L., 1999. Inflation-Target Design: Changing Inflation Performance and Persistence in
Industrial Countries. Federal Reserve Bank of St. Louis Review, March/April, 41-58.
Siklos, P. L., Bohl, M. T., and Wohar, M. E., 2010. Challenges in Central Banking: The Current
Institutional Environment and Forces Affecting Monetary Policy. Cambridge University
Press, New York.
Smets, F., (2003). Maintaining price stability: low long is the medium term? Journal of Monetary
Economics 50, 1293-1309.
29
Svensson, L. E. O., (1997). Inflation forecast targeting: Implementing and monitoring inflation
targets. European Economic Review 41, 1111-1146.
Svensson, L. E. O., (1999). Inflation targeting: Some extensions. Scandinavian Journal of
Economics 101, 337-361.
Svensson, L. E. O., (2009). Transparency under flexible inflation targeting: Experience and
challenges. Sveriges Riksbank Economic Review 1, 5-44.
Svensson, L. E. O., (2010). Inflation targeting. Forthcoming in B. M. Friedman and M.
Woodford (eds.) Handbook of Monetary Economics. North-Holland, Amsterdam.
Taylor, J. B., (1994). The inflation-output variability trade-off revisited. In Goals, Guidelines,
and Constrains Facing Monetary Policymakers, Federal Reserve Bank of Boston
Conference Series No. 38, 21-38.
Tillmann, P., 2012. Has inflation persistence changed under EMU? German Economic Review 13,
86–102.
Truman, E., (2003). Inflation targeting in the world economy. Washington, District of Columbia:
Institute for International Economics.
Vega, M., and Winkelried, D., (2005). Inflation targeting and inflation behavior: A successful
story? International Journal of Central Banking 1, 153-175.
Walsh, C. E., (1995a). Optimal contracts for independent central banker. American Economic
Review 85, 150-167.
Walsh, C. E., (1995b). Is New Zealand’s Reserve Bank Act of 1989 an optimal central bank
contract? Journal of Money, Credit and Banking 27, 1179-1191.
Walsh, C. E., (2009). Inflation targeting: What have we learned? International Finance 12,
195-233.
Willard, L., (2006). Does inflation targeting matter? A reassessment. Princeton University CEPS
Working Paper, vol. 120. February.
Wu T., (2004). Does inflation reduce inflation? An analysis for the OECD industrial countries.
Mimeo: Princeton University.
30
Table 1:
Inflation Targeting Countries: 2012
Adoption
Target 2012
Armenia
4% ±1.5%
Jan-06
New Zealand
1%–3%
Dec-89
Australia
2%–3%
Jun-93
Norway
2.50%
Mar-01
Brazil
4.5% ±2%
Jun-99
Peru
2% ±1%
Jan-02
Canada
2%(a)
Feb-91
Philippines
4.0% ±1%
Jan-02
Chile
3% ±1%
Sep-99
Poland
2.5% ±1%
1998
Colombia
2%–4%
Oct-99
Romania
3% ±1%
Aug-05
Czech Republic
2% ±1%
Dec-97
Serbia
4.0% ±1.5%
Jan-09
Ghana
8.7% ±2%
May-07
South Africa
3%–6%
Feb-00
Guatemala
4.5% ±1%
2005
South Korea
3% ±1%
Apr-98
Hungary
3%
Jun-01
Sweden
2%
1995
Iceland
2.50%
Mar-01
Thailand
3.0% ±1.5%(b)
May-00
4.5% ±1%
Jul-05
Turkey
5.0% ±2%
Jan-06
Israel
1%–3%
Jun-97
United Kingdom
2%
Oct-92
Mexico
3% ±1%
2001
Indonesia
Country
Year
Target 2012
Adoption
Country
Year
*
This year indicates when countries de facto adopted inflation targeting. Official adoption dates may vary.
(a)
(b)
Mid-point of 1%–3%
Target proposed by central bank at start of 2012, pending cabinet approval.
31