Printer Friendly

Credibility of inflation targets in Poland.


One of the principal elements of inflation targeting is a public announcement of numerical targets in order to reduce policy uncertainty and guide expectations. The stronger the impact of policy announcements on expectations, the easier it is for policymakers to control inflation and to accommodate shocks without jeopardising the objective of price stability. The concept of credibility, which seems particularly relevant in this context, is the marginal credibility, defined as policymakers' ability to influence public expectations by means of policy announcements (Cukierman and Meltzer, 1986). (1)

This paper empirically investigates the marginal credibility of inflation targets in Poland. It proposes a new methodology to estimate credibility by collecting three pieces of information: survey data on inflationary expectations, forecasts generated from past macroeconomic data, and central bank's inflation targets. The aim is to statistically assess the relative distance between the targets and survey-based expectations on the one hand, and between these expectations and data-based forecasts on the other. The closer the expectations are to the targets relative to the data-based forecasts, the higher is the marginal credibility of the target (the public attaches greater weight to policy announcements than to past developments). It is optimal for private agents to use both past data and policy announcements in forecasting inflation, as parameters of a central bank's loss function change over time in an unobservable way and policy announcements partly reveal these changes (Cukierman and Meltzer, 1986). (2)

The paper starts by discussing existing empirical literature on the marginal credibility of policy announcements. The next section presents a short description of inflation targeting in Poland. The subsequent section gives a more detailed description of the proposed methodology. Estimation results follow, and some technical details of the estimation strategy are relegated into the Appendix.


The issue of marginal credibility attracted relatively little attention in the context of inflation targeting, but was taken up empirically in earlier literature on the credibility of policy announcements. (3) Weber (1991) estimated the marginal credibility of monetary policy targets by regressing actual policy outcomes (a proxy for expectations) on targets and expected outcomes conditional on past data, with the coefficient of the target interpreted as a measure of marginal credibility. The main drawbacks of this methodology are that estimates are constant over time, the set of information used by the private sector for data-based forecasts is restrictive (past inflation), and the proxy for expectations is imprecise.

In an early application to inflation targeting, Leiderman (1995) analysed the marginal credibility of targets in Israel by comparing expectations derived from financial instruments shortly before and after the announcement of a new target. The targets seemed to affect inflationary expectations. The proposed method, however, does not allow for statistical testing and requires high-frequency series on inflation expectations, which are not typically available.

The analysis of the effect of policy announcements on expectations in Johnson (2002, 2003) is the closest to the methodology presented in this paper. Johnson compares survey-based inflation expectations with data-based forecasts before and after the introduction of inflation targeting in five early targeters. Survey-based expectations are significantly lower than the data-based forecasts in the inflation-targeting period, leading to a conclusion that the announcement of targets reduced inflationary expectations. The main drawbacks of Johnson's (2002, 2003) methodology are that it does not quantify the impact of targets on expectations.


Adopting a full-fledged inflation targeting regime in Poland was a gradual process, resembling the approach taken by other emerging economies, most notably Chile and Israel.

Macro-stabilisation in the 1990s relied on exchange rate, but inflation projections announced by the National Bank of Poland constituted an important element of the policy package. These annual projections were widely publicised, and can be regarded as precursors to inflation targets. They were consistent with the rate of crawl (introduced in 1991) and later with the rate of depreciation of the central parity in the target zone (introduced in 1995). This early form of inflation targeting was remarkably close to the experience of Chile and Israel, where inflation targets co-existed with exchange rate pegs. First announcements of inflation targets in Israel were similarly regarded as forecasts consistent with a peg (Leiderman, 1995).

The projection announcements implied a smooth disinflation path, with occasional adjustments after large deviations of actual inflation from the targets. Figure 1 illustrates the strategy by extrapolating annual inflation targets to the next year. (4) In most cases, extrapolated targets are in line with subsequently announced annual targets. Monetary authorities seemed to target a certain speed of disinflation for several consecutive years: if a deviation from a given path had been encountered, there was an attempt to correct it in the following year.


The first multi-period disinflation path ran from 1993 to 1995, followed by a flatter path from 1996 to 1998. All targets in the first sub-period were missed (see Table 1). Overshooting in 1993 was possibly regarded as due to a temporary shock, and subsequent targets implied a continuation of the same disinflation path. However, inflation proved to be persistent, and both 1994 and 1995 targets--consistent with the same disinflation trajectory--were missed. This led to a revision of the disinflation strategy: targets from 1996 to 1998 implied a slower disinflation and proved easier to achieve.

The newly formed Monetary Policy Council (MPC) officially adopted inflation targeting as a monetary policy strategy in 1999. The MPC initially continued announcing annual inflation targets--now official and expressed as a band. Initially, the targeted speed of disinflation was in line with the 1996-1998 path (the record-low inflation in the first months of 1999 even led to a slight downward revision of the 1999 target). The targets were overshot in both 1999 and 2000 and the two subsequent failures led to an upward shift in the targeted disinflation path for 2001 and 2002. However, earlier monetary policy tightening and external shocks led in turn to a massive undershooting in 2001 and forced the MPC to revise downward the 2002 target in mid-year. The 2003 target was unchanged from 2002, and a slightly lower level of 2.5 [+ or -] 1% has been chosen as a constant target since 2004. (5) The constant end-year target was continuously missed, although deviations were symmetric and inflation remained close to the target.


The paper collects three pieces of information to estimate marginal credibility: inflation forecasts generated from past data, targets and survey-based expectations. Forecasts from past data are generated from an unconstrained VAR model ('data-based model'). Central bank targets are regarded as an alternative model ('target model'), with appropriate distributions around point targets to account for uncertainty. Survey-based expectations are assumed to be a weighted average of forecasts from the data-based model and the target model. The estimated weight attached to the target model is interpreted as the marginal credibility of policymakers' announcements. The proposed estimation method utilises all information from individual survey observations and entire predictive densities from the data-based and target models. For intuition, Figure 2 shows the predictive densities and the distribution of survey expectations for December-to-December inflation in January 1998. The closer the distribution of survey observations is to the predictive density from the target model, the higher is the estimated credibility of the target.


The analysis uses a survey of professional forecasters conducted by Reuters as a measure of inflationary expectations. (6) The survey covers between 10 and 27 financial institutions from October 1994 to December 2007 and, for each month, reports expected December-to-December inflation at the end of the current year. The paper analyses observations after January 1996, when the survey also started collecting data on expected year-on-year inflation 12 months ahead or, after November 2000, 11 months ahead. (7)

Central bank targets are regarded as an alternative model for forecasting inflation (the target model). Inflation targets are December-to-December inflation projections published in the budget law before 1998 and MPC targets since 1999. Since policymakers cannot control inflation perfectly, the predictive density from the target model is assumed to have a normal distribution with means equal to point projections before 1998 and to middle points of the band since 1999. Standard deviation of this distribution is assumed to be an average percentage deviation of boundaries of the band from middle points for 1999-2007.

Projections based on past economic developments are generated from an unconstrained VAR (the data-based model). The basic system has four variables: CPI (in logs), end-period exchange rate of PLN to the average of DM and USD (in logs), 1-month WIBOR (inter-bank rate) and the rate of unemployment. The model is estimated on 12-month differences using Bayesian methods as in Kadiyala and Karlsson (1997). (8) The basic specification imposes relatively tight Minessota-type priors on parameters, found to reduce the average inflation forecast RMSE compared to a non-Bayesian VAR. In order to check for robustness, priors are relaxed in an alternative specification. In all models, the estimation sample changes, mimicking the information set available to the private sector. Since the CPI data are released with a lag of 2 months, in any given month forecasts are generated from models estimated on a sample ending 2 months earlier. Since the first observation from the Reuters survey used in the paper is January 1996, the first estimation period for the VAR runs from January 1994 to November 1995. The structure of the model remains unchanged throughout the analysis. In the baseline specification, learning about the structure of the economy is approximated by increasing the estimation sample with no discounting of earlier observations. Given apparent policy regime changes discussed above, we also estimate the model using a 5-year rolling window sample for robustness check.

Forecasters are assumed to aggregate information from the two predictive densities by averaging point forecasts from them. Points from the target density are chosen randomly and independently. The choice from the data-based density is restricted to the region between the 15th and 70th percentile. This assumption implies that diffuseness of forecasters' predictive density is correlated, but not fully reflected in the degree of disagreement among forecasters. Zarnowitz and Lambros (1987) show that professional forecasters may have loss functions penalising certain ranges of predictive densities, reflecting either their strategic behaviour or loss functions of their customers, but the degree of disagreement is nevertheless positively correlated with uncertainty surrounding the forecasts. (9) An alternative specification for checking the robustness of results assumes that forecasters choose all points from the data-based density.

Point predictions of inflation are a weighted average of point forecasts from the data-based model and the target-model. Weights attached to the target-model forecast are interpreted as a marginal credibility and are estimated using Bayesian numerical methods. The posterior distribution of weights [phi]([omega]|[inf.sup.e.sub.survey]) is proportional to


where [[omega].sub.t] is the weight attached to inflation target at time t, [phi]([[omega].sub.t]) is a prior distribution of [[omega].sub.t], and f([]|[[omega].sub.t]) is the posterior predictive density of inflation conditional on [[omega].sub.t] evaluated at forecaster i's survey expectations []. Since it is assumed that survey expectations are drawn independently from the same posterior predictive density, the likelihood function of [inf.sup.survey.sub.t] = ([inf.sup.survey.sub.1t],..., [inf.sup.survey.sub.nt])' is a product of predictive densities evaluated at individual expectations [[PI].sup.n.sub.i = 1]f([]|[[omega].sub.t]). The posterior density of (or is proportional to the product of the likelihood function and the prior density for the weights. The posterior predictive density of inflation conditional on [[omega].sub.t] (denoted by f([inf.sup.e]|[[omega].sub.t])) is simulated by drawing from [[omega].sub.t.sup.*][] ([inf.sup.e]) + [(1 - [[omega].sub.t]).sup.*] [f.sub.VAR]([inf.sup.e]), where []([inf.sup.e]) is the predictive density of inflation from the target model and [f.sub.VAR]([inf.sup.e]) is the predictive density of inflation from the data-based model, and the resulting density is evaluated at []. Finally, the mode of [[omega].sub.t] is estimated by searching for a maximum of the posterior density kernel [kappa]([[omega].sub.t]|[inf.sup.survey.sub.t]). Percentiles of the distribution for [[omega].sub.t] are obtained by deterministic integration (using the trapezoidal rule) of the posterior density. A non-informative prior for [[omega].sub.t] is used so that the posterior density kernel is equal to the likelihood function treated as a function of [[omega].sub.t].

The marginal credibility is estimated for end-year targets and for 12-month-ahead implicit disinflation paths derived in the previous section. Since the paper uses monthly data (available annual series are too short), forecast horizons change in the first approach. Given the 2-month lag in data availability, the longest forecast horizon is 13 months (December-to-December inflation predicted using data until November of the previous year), and the shortest horizon is 2 months (December-to-December inflation predicted using data until November of the same year). (10) The second approach allows for a more meaningful analysis of month-by-month changes in credibility at a constant, 12-month-ahead forecast horizon. (11)

The weights are estimated independently for each period by comparing predictive densities with the distribution of survey projections. For this reason, neither changes in the forecast horizon nor changes in the inflation targeting regime pose a problem for the proposed estimation method (the latter under the assumption that professional forecasters continue using the same data-based model for projection purposes in both regimes).


We start from a graphical analysis of survey-based inflationary expectations, data-based projections, and inflation targets for both December-to-December and year-on-year inflation. Figures 3 and 4 show modes of the respective densities.

The introduction of official inflation targeting appears to have boosted the impact of targets on inflation expectations. With the time-varying forecast horizon, the impact of targets becomes progressively weaker throughout each year (as monetary policy has a limited influence in the near term), but in the sample this effect varies in speed and strength (Figure 3). Survey-based December-to-December expectations are closer to data-based projections than to targets for most of 1996 and 1997. Actual inflation, though, was close to the 1997 target, and the announcement effect seems stronger in 1998. Credibility appears to further improve in 1999, the first full year after the official introduction of inflation targeting; despite inflation acceleration projected from the data-based model, the mode of survey expectations is fixed at the middle point of the target in the first quarter of the year. In 2000, the impact of the target on expectations seems similarly strong in the first half of the year. However, after a large deviation from the target in both 1999 and 2000, forecasters seem to turn to data-based projections faster in 2001 and in subsequent years. Only starting in the first half of 2004 and after 2005 the impact of targets seems stronger and more lasting again, as data-based low inflation projections seem mostly discounted by forecasters.



The effect of targets on expectations over the constant year-ahead horizon follows a similar pattern (Figure 4). The impact seems to be weak at the beginning of the sample, enhanced by the official inflation targeting in 1999 and in 2000, and weakened after deviations from official targets in 2001. The distance between expectations and the target diminishes in 2002, although the mid-year revision of the target in this year brought it to the level consistent with the data-based forecast, which makes identifying the influence of the target and past inflation developments difficult. (12) The distance remains small between 2003 and 2007, when--with a temporary exception in 2004--VAR-generated low inflation projections are largely dismissed by forecasters.

Econometric analysis confirms that credibility of December-to-December targets was boosted by official inflation targeting, but offers more insight into the evolution of weights attached to targets. Figure 5 reports the estimated weights. As expected, the effect of the target tends to fall towards the end of each year, when inflation is clearly outside the control of monetary authorities. Taking into account this seasonal pattern, the credibility of the target seems to be relatively low in 1996 and 1997 and higher in 1998. The official introduction of inflation targeting, combined with hitting the target in 1998, resulted in the record-high credibility of the 1999 target. The weight attached to the target was high in the first quarter of this year despite large overshooting predicted by the data-based model. The weight was still high in 2000, although the falloff during the year was faster than in 1999. Large deviations in 1999 and in 2000 hindered credibility, reducing the weight between 2001 and 2003 (a spike in mid-2002 is due to bringing the target in line with inflationary developments mid-year). The impact of the target on expectations was again strong in the first half of 2004 and at the beginning of 2005, reflecting the discounting of low-inflation VAR forecasts. The estimated credibility at the beginning of 2006 is, however, lower, which is likely the effect of target undershooting in 2005. Credibility improves again in 2007.

The introduction of official inflation targeting had an even more pronounced effect on expectations over the 1-year-ahead horizon. Figure 6 indicates that estimated weights attached to inflation targets are on average higher in the official targeting period. The weights, however, drop after large deviations from the targets in 1999 and 2000. They recover in mid-2002, although the interpretation of the estimates in this period is complicated because of the revision of the target as discussed above. The weights have remained at a relatively high level since 2003, but with some fluctuations. Credibility temporarily dropped in 2004, when inflation accelerated and breached the December-to-December target. A more pronounced decline can be observed from mid-2005 to mid-2006, when both the actual inflation and data-based projections remained well below the MPC constant target.



In order to check the sensitivity of results to different specifications of the model used for generating predictive densities, the estimation is repeated with predictive densities from unrestricted VAR, VAR with weaker priors, and VAR estimated using the 5-year rolling window sample. The results are similar to the baseline. As an illustration, Figure 7 report results for 1-year-ahead expectations. There are two important differences from the baseline specification: the model with the unrestricted predictive density generates a higher credibility of the implicit targets throughout the sample, and the rolling window estimation indicates lower credibility from mid-2005 to end-2006. The first result is due to a much wider range of unrestricted predictive density, less likely to represent data-based projections by professional forecasters. The second result reflects a better forecasting performance of the rolling window model in the last years of the sample.


One of the potential pitfalls of the proposed methodology is identification problems when data-based forecasts converge to targets. Obviously, a successful inflation targeting framework may modify dynamic interactions between macroeconomic variables and the data-based VAR model will adapt to these changes. In the limiting case of credible strict inflation targeting with a constant low inflation target, the VAR forecast will eventually replicate the target, eliminating the difference between the data- and target-based forecasts. However, the adaptation of the VAR parameters will not be immediate, mimicking the learning process of economic agents after the switch to inflation targeting. In transition economies such as Poland, the learning process about the new policy regime seems lengthy due to protracted high inflation history and uncertainty surrounding policymakers' commitment to price stability. Another limiting case is when targets are adjusted to reflect inflation data-based projections. Such a case arises in mid-2002, hindering the interpretation of the estimated weights in this period.


The credibility of inflation targets in Poland seems to have been high, particularly after the formal introduction of inflation targeting. The proposed Bayesian approach makes efficient use of available information to provide an intuitive, time-varying measure of the credibility of the target. Although the paper does not attempt to fully explain determinants of credibility, its evolution seems to depend on the following:

* Institutional reforms: In 1999 the National Bank of Poland officially adopted inflation targeting as a monetary policy framework. The influence of the target on expectations seems to be particularly strong in 1999 and 2000, the first 2 years of official inflation targeting. The impact of targets on expectations is visible both for December-to-December and 1-year-ahead expectations.

* Past deviations from the target: Small deviations boost reputation and lead private agents to revise their mechanism of forming expectation by increasing the weight attached to the target, and vice versa. In the case of Poland, a negative reputation effect can be observed in 2001, after large deviations in 1999 and 2000. The impact of past deviations seems equally pronounced for December-to-December and 1-year-ahead expectations.

* Forecasting ability of data-based models: The increase in estimated weights attached to targets in 1999 and 2000 may reflect worsening of the forecasting ability of the VAR model. The weaker performance may in turn be a result of a gradual adjustment of model parameters after policy changes in 1999. While this may be regarded as a shortcoming of the model applied in the paper, it is likely that the policy change genuinely weakened the data-based forecasting ability of professional forecasters, who temporarily relied more on policy announcements.

Although high credibility of the targets can reduce the cost of disinflation, there are some potential dangers of the high marginal credibility for policy making. Official press releases of the MPC give considerable attention to the evolution of professional forecasters' inflationary expectations, suggesting that policymakers' decisions are to some extent influenced by this information. This is no different from the practice of advanced inflation targeters, such as United Kingdom and New Zealand, where professional forecasters' inflationary expectations take an important place in inflation reports. However, Bernanke and Woodford (1997) show that if monetary policy relies solely on inflationary expectations of outsiders, and if these expectations are in turn based only on announced inflation targets, the resulting indeterminacy may potentially lead to undesirable economic equilibria. In this context, Bernanke and Woodford (1997) stress the importance of structural models in monetary policy making. Structural econometric modelling is difficult in the transition environment, making the use of survey-based expectations more tempting for the central bank. It is, however, indispensable in any serious implementation of the inflation targeting regime.



Following Kadiyala and Karlsson (1997), the VAR system has the following form:

[y.sub.t] = [z.sub.t][GAMMA] + [u.sub.t]

where [z.sub.t] = {[x.sub.t], [y.sub.t-1],..., [y.sub.t-p]}, with [y.sub.t] a 1 x m vector of endogenous variables and [x.sub.t] a 1 x q vector of exogenous variables at time t. [GAMMA] is k(= q - pm) x m matrix of parameters. Stacking vectors [y.sub.t], [z.sub.t] and [u.sub.t] for t = 1,..., T into Y, Z and U gives a multivariate regression model:

Y = Z[GAMMA] + U

Allowing the subscript to denote the ith column vector, the equation for a variable i is [y.sub.i], = Z[[gamma].sub.i] + [u.sub.i]. By stacking the columns of Y, [GAMMA] and U into y, [gamma], u, it can be written as

y = (I [cross product] Z)[gamma] + u

Assuming that u ~ N(0, [PSI] [cross product] I) and using conjugate (Normal-Wishart) priors,


posterior distributions are given by

[gamma]|[PSI], y ~ N([bar.[gamma]], [PSI] [cross product] [bar.[OMEGA]]), [PSI]|y ~ iW([bar.[PSI]], T + [alpha])


In the paper, prior means make each endogenous variable an AR1 process with autoregressive parameter [alpha]:

[] = [alpha][y.sub.i,t-1] + []

Other parameters of the Normal-Wishart prior are set so that elements of [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] coincide with elements of the variance-covariance matrix specified for parameters in equation i:


where k is the lag length and [[sigma].sub.i] is a scale factor accounting for different variability in endogenous variables, set to the residual standard error from p-lag univariate autoregression for variable i.

Priors for the baseline forecasting model used in the paper are [[pi].sub.1] = 0.1, [[pi].sub.2] = 10e5 and [alpha] = 0.95, resembling Minnesota priors, but assuming stationary and persistent processes for all series. An alternative specification makes this prior close to diffuse by setting [[pi].sub.1] = 10e5.


Bernanke, BS, Laubach, T, Mishkin, FS and Posen, AS. 1999: Inflation targeting: Lessons from the international experience. Princeton University Press: Princeton.

Bernanke, BS and Woodford, M. 1997: Inflation forecasts and monetary policy. NBER Working Paper 6157. National Bureau of Economic Research.

Carroll, C. 2003: Macroeconomic expectations of households and professional forecasters. Quarterly Journal of Economics 118(1): 269-298.

Cukierman, A and Meltzer, AH. 1986: The credibility of monetary announcements. In: Neumann, MJM (ed). Monetary Policy and Uncertainty. Nomos Verlagsgesellschaft: Baden-Baden.

Faust, J and Svensson, LEO. 2001: Transparency and credibility: Monetary policy with unobservable goals. International Economic Review 42(2): 369-397.

Johnson, D. 2002: The effect of inflation targeting on the behavior of expected inflation: Evidence from an 11 country panel. Journal of Monetary Economics 49(8): 1521-1538.

Johnson, D. 2003: The effect of inflation targets on the level of expected inflation in five countries. The Review of Economics and Statistics 85(4): 1076-1081.

Kadiyala, KR and Karlsson, S. 1997: Numerical methods for estimation and inference in Bayesian VAR-models. Journal of Applied Econometrics 12: 99-132.

Leiderman, L. 1995: Inflation targeting in Israel. In: Leiderman, L and Svensson, LEO (eds). Inflation Targets. Centre for Economic Policy Research: London.

Ottaviani, M and Sorensen, PN. 2001: The strategy of professional forecasting Discussion Papers 01-09. Institute of Economics, University of Copenhagen.

Walsh, CE. 1999: Announcements, inflation targeting and central bank incentives. Economica 66: 255-269.

Weber, AA. 1991: Reputation and credibility in the European Monetary System. Economic Policy 6: 57-102.

Zarnowitz, V and Lambros, LA. 1987: Consensus and uncertainty in economic prediction. Journal of Political Economy 95: 591-621. Symposium Paper

(1) More recently, the concept of marginal credibility was investigated by Walsh (1999).

(2) The announcements are informative even if the central bank deliberately attempts to manipulate private sector expectations.

(3) Several studies empirically analyse average credibility, defined by Cukierman and Meltzer (1986) as a distance between targets and expectations, and further analysed in Faust and Svensson (2001). See Bernanke et al. (1999) for examples. Most of the empirical work finds a slow convergence of inflationary expectations to the targets.

(4) The extrapolation procedure assumes an asymptotic decline in the targeted inflation to a level consistent with price stability. The disinflation path is approximated by an AR1 process. Two data points needed to calibrate parameters of this process are given by the initial inflation and the terminal 'price-stability' rate, assumed at 2.5%. In all the calculations we use information sets currently available to economic agents. Targets were usually announced in September economic agents knew July inflation rates at the time of announcements.

(5) 2002-2003 targets implied temporary increases in inflation from prevailing levels to above the assumed 'price-stability' rate. The applied extrapolation procedure does not allow for such increases. Hence the line for 2002-2003 is at the assumed 'price-stability' level of 2.5%.

(6) Inflationary expectations of other groups in the society are not analysed. Households' expectations may rely on the views of professional forecasts (Carroll, 2003); hence the effect of the target on these expectations may be delayed and indirect.

(7) Since the CPI data are released with a lag, 12-month-ahead forecasts were effectively 13 months ahead starting from the previous month. After November 2000 the respondents were asked about 12-month-ahead forecasts starting from the previous month. The number of respondents reporting projections for December-to-December inflation gradually declined, potentially affecting the precision of estimates.

(8) Estimating the model on levels with seasonal dummies produces similar results, but generates higher average RMSE for inflation forecasts. Stationarity of the variables is not tested, as we do not impose co-integration restrictions (short samples), and Bayesian predictive densities are valid irrespective of stationarity properties of the data.

(9) See Ottaviani and Sorensen (2001) and references cited there for the analysis of strategic behaviour in forecasting.

(10) Changes in the forecast horizon do not pose a problem for the applied estimation method, as it is based on a comparison of predictive densities, which are available for all forecast horizons.

(11) After November 2000 the paper analyses the credibility of 11-month-ahead implicit targets to be consistent with changes in the Reuters survey (see footnote 7). The standard deviation of the distribution around the target in the second approach is assumed to be the same as for end-of-year targets.

(12) The target was similarly adjusted at the beginning of 1999 to reflect low inflation in the first months of this year. Data-based forecast, however, remains significantly higher in this period, limiting the identification problem.


Fiscal Affairs Department, International Monetary Fund, 700 19th Street NW, HQ2-06-779, Washington, DC 20431, USA. E-mail:
Table 1: Year-on-year inflation targets and outcomes

 Target Inflation

1993 32.2 37.7
1994 23.0 29.4
1995 17.0 22.0
1996 17.0 18.7
1997 13.0 13.2
1998 9.5 8.5
1999 8.0-8.5
1999 (April revision) 6.6-7.8 9.8
2000 5.4-6.8 8.6
2001 6.0-8.0 3.6
2002 4.0-6.0
2002 (July revision) 2.0-4.0 0.7
2003 2.0-4.0 1.7
2004 1.5-3.0 4.4
2005 1.5-3.0 0.7
2006 1.5-3.0 1.4
2007 1.5-3.0 4.0

Source: National Bank of Poland.
COPYRIGHT 2008 Association for Comparative Economic Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Symposium Paper
Author:Maliszewski, Wojciech S.
Publication:Comparative Economic Studies
Article Type:Report
Geographic Code:4EXPO
Date:Sep 1, 2008
Previous Article:Inflation targeting, policy rates and exchange rate volatility: evidence from Turkey.
Next Article:From rhetorics to practice in monetary policy: a Romanian perspective.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters