## _wp11276

## Source details

**Canonical URL:** [_wp11276](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11276.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11276.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11276.pdf.json)

---

### Major findings and empirical conclusions
- Inflation targeting (IT) appears to be associated with lower inflation and lower inflation volatility.
- Limited evidence of an adverse impact of IT on output; some specifications show smaller increases or negative effects on real GDP growth.
- Empirical approach constrained by the small number of LICs that have adopted IT and the recent timing of their adoption within the sample period (sample runs up to 2008).
- Baseline quantitative summary: "1.4 percentage points in the baseline case. The same results hold in all our robustness checks."

### Scope, sample definition, and empirical strategy
- LIC definition: countries that were PRGT-eligible up to 2008 (includes Albania, Angola, Azerbaijan, India, Pakistan, and Sri Lanka).
- Official LIC inflation targeters (IMF MCM classification): Armenia, Ghana, Albania.
- Empirical strategy:
  - Case studies of LICs that adopted or are moving toward IT.
  - Comparative analysis of emerging-economy ITers and LICs (selecting "mature stabilizers" to reduce heterogeneity).
  - Descriptive statistics and regression/panel methods, including Difference–in–Difference and panel analysis (D-GMM and S-GMM variants).

### Key case studies and institutional notes
- Ghana
  - Formal adoption: May 2007 (preceded by informal practice since 2002).
  - Target: 12-month headline CPI change (includes energy and utility prices); medium-term goal 5 percent +/- 1 per-cent.
  - Performance notes:
    - Inflation fell from over 30 percent (early 2000s) to near 10 percent by mid-2006.
    - Headline inflation rose through 2007–08, stabilized in the 20 percent range in early 2009, and was 8.6 percent by December 2010 (actual).
    - Monetary policy responses were sometimes inadequate; policy interest rates were negative in real terms during 2008–09.
    - Inflation expectations not well anchored; growth mostly above SSA average since 2002.
  - Policy reaction emphasis: reduce variability of output and interest rates rather than hit annual inflation targets at all cost.
- Armenia
  - Official adoption: 2006; switched from a monetary anchor.
  - Target: 4 percent +/-1.5 percent (headline CPI annual rate as published by CBA); target reset annually in agreement with government.
  - Institutional strengthening: CBA legal independence enhanced; direct financing of government prohibited; accountability to parliament; no explicit sanctions for missing target.
  - Performance notes:
    - Growth averaged around 5 percent (1996–2000), accelerating to 12 percent (2001–05); inflation averaged 5.6 percent and 3.3 percent in those periods respectively.
    - Since 2006 inflation remained within the CBA band except in 2008; growth double-digit in 2006–07, ~7 percent in 2008, and -4 percent during the global crisis.
    - Monetary transmission mechanism remains relatively weak due to limited financial market depth.
- Albania
  - Inflation target for 2009-2011: 3.0 percent with a symmetrical tolerance band of +/- 1 percentage points (measured by annual CPI change).
  - Historical performance: largely maintained headline inflation within 3+/-1 percent over the past decade; during the crisis inflation rose above 4 percent due to depreciation pass-through and higher electricity prices.
  - BoA actions: enhance communication, expand information set, improve inflation forecasting methodologies, strengthen central bank independence.
  - Growth averaged around 6 percent over the past five years when IT framework was implemented.
- Moldova and Georgia
  - Moldova: shifted to IT at beginning of 2010; 2010 target 5 percent +/- 1.5 percent; central bank independence provisions included; historical challenges include weak transmission, low monetization, imperfect money and FX markets, and dollarization.
  - Georgia: in process of moving to formal IT; strengthened monetary instruments and analytics; constraints include extensive dollarization and volatile inflation due to food weight in CPI.

### Key elements of an inflation-targeting framework (Box 1, verbatim structure)
- An explicit central bank mandate to pursue price stability as the primary objective of monetary policy and a high degree of operational autonomy;
- An explicit quantitative target for inflation;
- An information inclusive strategy in which many variables, not just monetary aggregates and the exchange rate, are used to inform policy decisions;
- Central bank accountability for performance in achieving the inflation objective, mainly through high-transparency requirements for policy strategy and implementation;
- A policy approach based on a forward-looking assessment of inflation pressures, taking into account a wide array of information.

### Stated benefits of IT (verbatim labels)
- (i) countries’ efforts to disinflate;
- (ii) better anchor inflation expectations;
- (iii) allow the floating exchange rate regimes accompanying the IT framework in becoming efficient shock-absorbers;
- (iv) reduce exchange rate pass-through;
- (v) improve communication and transparency;
- (vi) clearly assign institutional responsibilities for inflation control.

### Preconditions and constraints for effective IT in LICs (Box 3 and related discussion)
- Typical preconditions (four broad categories):
  - Institutional independence (absence of fiscal obligation, operational independence, inflation-focused mandate, favorable fiscal balance, low public debt, central bank independence).
  - Well-developed technical infrastructure (forecasting, modeling, data).
  - Economic structure conducive to IT (prices deregulated; low sensitivity to commodity prices and exchange rates; low dollarization).
  - A healthy financial system (sound banking system; developed capital markets).
- Common impediments in LICs:
  - Weak monetary transmission mechanism (weak institutional frameworks, reduced securities markets, imperfect banking competition, high bank lending costs).
  - Fiscal dominance and reliance on seigniorage; shallow capital and FX markets.
  - Banking fragility and legacy financial repression.
  - Conflicting objectives (explicit or implicit exchange rate aims) that reduce IT effectiveness.
- Empirical observation: no inflation targeter had all preconditions fully in place before adopting IT.

### Empirical methods and samples (Difference–in–Difference and Panel)
- Baseline DID period: 1990-2008; hyperinflation exclusion rule: exclude country series with any year > 50 percent inflation.
- Control groups:
  - LIC control group: 29 "mature stabilizers" plus near-misses.
  - IT group (emerging-economy comparator): 10 countries (adopted IT prior to 2006).
  - Combined control group: LICs plus 18 non-IT emerging markets (regional comparability).
- Timing partition: pre/post split uses average IT adoption date, year 2002.
- Panel analysis: country sample of 10 emerging market IT countries and 29 LICs (N=39), sample period 1990–2008, yearly data summarized into three-year intervals. Inflation transformed as y_{i,t} = 100·log(1+Y_{i,t}/100). High inflation dummy = 1 if transformed inflation > 40 percent.

### Difference–in–Difference key quantitative results (selected coefficients preserved)
- Inflation (LIC control group, IT Dummy across models 1–5): -1.96*, -2.06*, -2.21**, -2.33**, -1.62*.
- Inflation Volatility (LIC control group, IT Dummy models 1–5): -1.41**, -1.54**, -1.64***, -1.49**, -1.47***.
- Inflation (Combined control group, IT Dummy models 1–5): -2.56**, -2.49*, -3.07***, -2.61**, -3.24***.
- Growth (LIC control group, IT Dummy models 1–5): -1.35**, -0.81, -1.55*, -1.66***, -2.10***.
- Growth (Combined control group, IT Dummy models 1–5): -0.19, -0.13, -0.16, -0.44, -0.67 (generally not significant).
- Note: Initial conditions coefficients (pre-period levels) are highly significant in many specifications (e.g., -0.73*** to -0.97*** across tables), indicating reversion-to-the-mean effects.

### Panel estimation key quantitative outcomes (selected coefficients preserved)
- Log Inflation (representative IT dummy coefficients across estimators): -2.11***, -1.70***, -1.98, -1.91, -1.75**, -3.13***.
- Log Inflation Volatility (representative IT dummy coefficients): -2.20***, -1.02***, 3.28*, 7.61, -1.02***, -1.25*** (note estimator sensitivity).
- Log Growth (representative IT dummy coefficients): -0.39, -1.22***, -0.33, -3.80, -1.35***, -2.03**.
- Log Growth Volatility (representative IT dummy coefficients): -0.73***, -0.47**, -0.03, 1.73, -0.56**, -0.59*.
- High Inflation Dummy (inflation equations): very large and significant coefficients (e.g., 33.06***, 29.98***, 30.08*** across estimators).

### Robustness, interpretation, and caveats
- IT associated reductions in inflation and inflation volatility are robust across diff-in-diff and panel GMM methods, though magnitudes and significance vary by specification and sample.
- Negative association with real GDP growth appears in some specifications (especially DID with LIC controls) but is not uniformly robust and may reflect LIC characteristics rather than causal IT effects.
- Results sensitive to sample composition (e.g., excluding currency-union members increases estimated IT effect: Model 4 shows -2.33 percentage points) and to treatment of high-inflation country series.
- Short post-adoption windows, coincident structural reforms, and potential endogeneity of IT adoption timing limit causal interpretation; panel GMM used to address endogeneity but estimator sensitivity remains.

### Policy-relevant implications and practical recommendations preserved from the source
- Preconditions are important but not strictly required for adoption; many IT adopters did not fully satisfy preconditions prior to adoption.
- Enhancements for IT effectiveness emphasized in country cases:
  - Strengthen central bank independence and legal protections (recapitalization rights, protection from external pressure, term lengths exceeding election cycles, conflict of interest prohibitions, disclosure requirements).
  - Improve communication policy and transparency (publication of decisions and minutes; regular reporting to parliament).
  - Expand information sets and improve inflation forecasting methodologies.
  - Develop financial markets and deepen monetary instruments to strengthen monetary transmission.
  - Avoid fiscal dominance: prohibit direct financing of government and build fiscal credibility.
- Practical caution: IT may need to be adapted to country circumstances (e.g., attention to monetary aggregates or the exchange rate when they contain valuable information for monetary transmission).

*Source: _wp11276*

### References .............................................................................................................

### _wp11276 - References

### Major findings and empirical conclusions
- Inflation targeting (IT) appears to be associated with lower inflation and lower inflation volatility.
- There is limited evidence of an adverse impact of IT on output.
- The paper’s empirical approach is constrained by the small number of LICs that have adopted IT and the recent timing of their adoption within the sample period (sample runs up to 2008).

### Scope, sample definition, and methodology
- LICs are defined as those countries that were PRGT-eligible up to 2008; therefore Albania, Angola, Azerbaijan, India, Pakistan, and Sri Lanka (which graduated from the PRGT list in 2010) are classified as LICs in this paper.
- Only three LICs are officially inflation targeters according to the IMF classification: Armenia, Ghana, and Albania.
- The empirical strategy:
  - Present case studies of LICs that have adopted IT or are on the way to adopting IT.
  - Compare macroeconomic outcomes of emerging-economy ITers with LICs, selecting in the sample only the emerging LICs identified as mature stabilizers to reduce heterogeneity.
  - Use descriptive statistics and regression/ panel methods (including Difference–in–Difference and panel analysis — see tables listed) to evaluate impacts.

### Context and historical notes on adoption
- First central bank to adopt IT: New Zealand in December 1989.
- Most recent adopter cited: Serbia in 2009.
- Several countries adopted IT after abandoning exchange rate pegs in crises (examples noted: Brazil and Indonesia).

### Quantitative descriptive comparisons preserved from the source
- Advanced economies — median inflation:
  - Five years prior to IT: ranged between about 2 to 14 percent.
  - Five years after IT: fell to the range of 1 to 5 percent.
- Emerging economies — median inflation:
  - Five years prior to IT: ranged between 4 to over 25 percent.
  - Five years after IT: dropped to 2 to 8 percent.
- The paper notes that the decline in volatility in advanced economies may partly reflect the “Great Moderation” era.

### Key elements of an inflation-targeting framework (Box 1, verbatim structure and content)
- An explicit central bank mandate to pursue price stability as the primary objective of monetary policy and a high degree of operational autonomy;
- An explicit quantitative target for inflation;
- An information inclusive strategy in which many variables, not just monetary aggregates and the exchange rate, are used to inform policy decisions;
- Central bank accountability for performance in achieving the inflation objective, mainly through high-transparency requirements for policy strategy and implementation;
- A policy approach based on a forward-looking assessment of inflation pressures, taking into account a wide array of information.

### Stated benefits of IT reported in the literature (verbatim list labels preserved)
- (i) countries’ efforts to disinflate;
- (ii) better anchor inflation expectations;
- (iii) allow the floating exchange rate regimes accompanying the IT framework in becoming efficient shock-absorbers;
- (iv) reduce exchange rate pass-through;
- (v) improve communication and transparency;
- (vi) clearly assign institutional responsibilities for inflation control.

### Conditions and cautions for IT effectiveness in LICs
- The IT framework is effective in countries that have:
  - (a) a well functioning monetary transmission mechanism;
  - (b) a degree of independence of monetary policy;
  - (c) absence of commitment to a particular level for the exchange rate;
  - (d) a certain amount of fiscal credibility.
- The paper cautions that several economic and structural problems in LICs may weaken the effectiveness of the IT framework.

### Paper structure (as presented)
- Section I: Introduction (poses the main question regarding IT’s impact on LIC macroeconomic performance).
- Section II: Global emergence of IT.
- Section III: Case studies of LICs that adopted or are adopting IT.
- Section IV: Literature review on IT impacts in advanced and emerging economies.
- Section V: Methodologies and empirical results for LICs.
- Section VI: Challenges LICs face in adopting an effective IT framework.
- Section VII: Summary and conclusions.

*Source: _wp11276 - References*

### Box 2. Why Inflation Targeting?

### Box 2. Why Inflation Targeting?

### Proponents' arguments
- Inflation targeting can help build credibility and anchor inflation expectations more rapidly and durably.
- Inflation targeting makes it clear that low inflation is the primary goal of monetary policy and it involves greater transparency to compensate for the greater operational freedom that inflation targeting offers.
- Inflation targets are intrinsically clearer and more easily observable and understandable than other targets since they typically do not change over time and are controllable by monetary means.
- Inflation targeting grants more flexibility: the target on inflation is typically interpreted as a medium-term goal, implying central banks pursue the inflation target over a certain horizon by focusing on keeping inflation expectations at target.
- Short-term deviations of inflation from target are acceptable and do not necessarily translate into losses in credibility.
- Inflation targeting involves a lower economic cost in the face of monetary policy failures compared with some alternative monetary commitments (like exchange rate pegs), whose policy failures can involve massive reserve losses, high inflation, financial and banking crises, and possibly debt defaults.
- The output costs of a failure to meet the inflation target are limited to temporarily higher-than-target inflation and temporarily slower growth, as interest rates are raised to bring inflation back to target.

### Opponents' arguments
- Inflation targeting offers too little discretion and so it unnecessarily restrains growth: it is argued to be too confining in terms of an ex ante commitment to a particular inflation number and a particular horizon over which to return inflation to target.
- Inflation targeting cannot anchor expectations because it offers too much discretion: some argue it cannot help build credibility in countries that lack it, because it offers excessive discretion over how and when to bring inflation back to target and because targets can be changed.
- Inflation targeting implies high exchange rate volatility: it is often believed that elevating price stability to the primary goal requires a benign neglect of the exchange rate, potentially increasing exchange rate volatility and harming growth.
- Inflation targeting cannot work in countries that do not meet a stringent set of “preconditions”, making the framework unsuitable for the majority of emerging market economies. Preconditions often considered essential include: the technical capability of the central bank in implementing inflation targeting, absence of fiscal dominance, financial market soundness, and an efficient institutional setup to support and motivate the commitment to low inflation.

### Emergence of IT in low-income countries (overview)
- According to the IMF’s Monetary and Capital Market Department (MCM) classification there are only three LICs worldwide that have adopted IT– Ghana, Armenia7 and Albania.
- Given that all of these countries have adopted IT recently, it would be premature to draw generalized conclusions about the performance of IT in LICs.

### A. Ghana — adoption and performance
- Adoption:
  - In May 2007, the Bank of Ghana (BoG) formally announced its adoption of formal inflation targeting.
  - BoG had been informally pursuing an inflation targeting regime for the previous few years.
  - The framework targets the 12-month change in the headline Consumer Price Index (CPI) which includes energy and utility prices; BoG also monitors other core inflation measures.
  - BoG had been building institutional, analytical and communication elements of the framework since 2002.
  - By 2007, key arrangements in place included central bank policy independence, instrument independence, bi-monthly meetings of a Monetary Policy Committee (MPC), and generally good transparency.
  - Further progress was needed on financial sector development, communications, and understanding of the monetary transmission mechanism.
- Performance and key statistics:
  - After several years of preparation and a successful disinflation strategy, inflation had been brought down from over 30 percent in the early 2000s to near 10 percent by mid-2006.
  - The Bank of Ghana set a medium-term goal of 5 percent inflation within a band of +/- 1 per-cent, along with some intermediate inflation-reduction targets.
  - Global food and fuel price shocks pushed inflation up within a few months of formally switching to IT.
  - Domestic factors and expansionary fiscal policy contributed to acceleration in inflation during 2008–2009.
  - Monetary policy was not adequately tightened during the period of rising inflation; policy interest rates were negative in real terms.
  - Headline inflation diverged from announced targets.
  - Inflation rose through 2007–08 and stabilized in the 20 percent range in early 2009 in part due to the slowdown of the economy.
  - With moderation of domestic demand and fiscal tightening, inflation was brought down to single digits by mid-2010 and to 8.6 percent by December 2010 (actual).
  - Inflation expectations are not well anchored in Ghana, reflecting the history of high and volatile inflation; anchoring expectations at a low level would require maintaining low inflation over an extended period.
  - Growth has been on an upward trend and mostly above the SSA average since 2002 when Ghana informally adopted IT.
- Policy reaction:
  - In responding to shocks, BoG sought to reduce variability of output and interest rates rather than hit pre-announced annual inflation targets at all cost.

### B. Armenia — adoption and performance
- Adoption:
  - Armenia officially adopted inflation targeting in 2006 by shifting from a monetary anchor.
  - Price stability is the primary objective, while other targets are subordinate.
  - The Central Bank of Armenia (CBA) has a clear quantified inflation target of 4 percent +/-1.5 percent.
  - The inflation target is measured by the annual rate of change in the headline Consumer Price Index, as calculated and published by the CBA.
  - The target is reset each year in agreement with the government in the context of the annual budget and has been revised upward on two occasions.
  - Under the IT framework, CBA independence was considerably strengthened (legal stipulations on independence, right to recapitalization, protection from external pressure, term of office exceeds election cycle, conflict of interest prohibition, fit and proper practice, disclosure requirements).
  - There is no fiscal dominance and direct financing of government is prohibited.
  - The CBA is accountable to parliament through regular reporting requirements; decisions and minutes of the CBA Board meeting are regularly published.
  - In case of failure to meet the target, there are no explicit sanctions.
  - The monetary transmission mechanism has remained relatively weak owing to the limited depth of financial markets.
- Performance and key statistics:
  - Prior to shifting to IT, prudent fiscal and monetary policies, large external inflows, and ongoing structural reforms contributed to double-digit growth in a low-inflation environment.
  - Following the sharp contraction of the early 90s, economic growth averaged around 5 percent for the period 1996–2000, before accelerating to 12 percent during 2001–05, while inflation averaged

5.6 percent and

3.3 percent, respectively.
  - Compared to the Commonwealth of Independent states (CIS), Armenia had higher growth and much lower inflation rates.
  - Since 2006, inflation remained within the CBA band, except in 2008 where it exceeded the band largely due to the global spike in food and fuel prices.
  - Growth remained in double digit during 2006-07, before dropping to about 7 percent in 2008, and turning negative (-4 percent) during the global crisis.

### C. Other low-income countries
- There is great heterogeneity among other LICs that are in different stages of adopting IT.
- Albania is the latest addition to the IT family.
- Several countries (such as Moldova, Georgia, etc.) could be described as informal inflation targeters, as their policy frameworks are geared toward price stability but inflation-targeting infrastructure is not fully in place.

*Source: World Economic Outlook, September 2005, International Monetary Fund.*

### 2004. Implementation was effective in 2009. The inflation target for the 2009-2011 period is

### _wp11276 - 2004. Implementation was effective in 2009. The inflation target for the 2009-2011 period is

### Albania: inflation-targeting adoption and recent performance
- Inflation target for the 2009-2011 period is set at 3.0 percent, with a symmetrical tolerance band of +/- 1 percentage points.
- The inflation target will be measured by the annual rate of change in the Consumer Price Index.
- The inflation target is applicable throughout the 2009-2011 period, implying that the actual inflation rates may temporarily deviate from target; such deviations are acceptable given external shocks or unforeseen circumstances beyond central bank control (BoA view).
- BoA actions to enhance the IT framework:
  - Making its communication policy more effective.
  - Expanding its information set.
  - Improving its inflation forecasting methodologies.
  - Significant progress in central bank independence.
- Performance observations:
  - Over the past decade, Albania has largely maintained headline inflation within the BOA 3+/-1 percent target range.
  - During the crisis, inflation rose above 4 percent mainly due to depreciation pass-through and higher electricity prices.
  - Administrative price increases over the next two years may drive headline inflation temporarily above the 3±1 percent target band in the near term.
  - Underlying inflation is expected to remain under control, and annual inflation is also expected to remain within the band.
  - Since 1992 the Albanian economy experienced high growth rates; growth averaged around 6 percent over the past five years when Albania implemented its inflation targeting framework.

### Moldova and Georgia: IT adoption status and institutional features
- Moldova:
  - Shifted at the beginning of 2010 to an inflation targeting framework.
  - The target for 2010 is 5 percent +/- 1.5 percent.
  - The inflation target will be measured by the annual rate of change in the headline CPI.
  - Under the new IT framework, the central bank will have full independence; the law provides: (i) the right to recapitalize; (ii) protection from external pressure; (iii) term of office exceeds election cycle; (iv) prohibition of conflict of interest; (v) fit and proper practice; and (vi) disclosure requirements.
  - No fiscal dominance; direct financing of government is prohibited.
  - Historical challenges: high inflation (except in 2009), weak monetary transmission, low monetization and financial development, imperfect money and FX markets, and dollarization.
- Georgia:
  - In the process of moving to a formal inflation targeting regime.
  - National bank strengthened monetary policy instruments and internal analytical capacity; greater traction on market interest rates.
  - Remaining constraints: extensive dollarization, volatile inflation due to large weight of food prices in the CPI.

### Literature survey on the outcomes of inflation targeting (IT)
- Central question: How have ITers performed compared to non-ITers and has IT made a significant difference?
- Main approaches in the literature:
  - Difference-in-difference.
  - Propensity score matching.
  - Panel estimations including generalized methods of moments (GMM) — Difference-GMM (D-GMM) and System-GMM (S-GMM).
- Performance measures used:
  - Inflation dynamics: average inflation rates, inflation volatility, inflation persistence.
  - Real economy effects: real GDP growth and growth volatility.
- Summary of empirical findings:
  - No study finds macroeconomic performance deteriorated after IT introduction.
  - Advanced economies: mostly small, mostly insignificant effects of IT on inflation, inflation variability, output growth, output variability, and long-term interest rates; some evidence that IT anchors long-run inflation expectations.
  - Emerging market economies: majority of studies find IT significantly reduces average inflation; evidence on inflation volatility and real economy effects is mixed and heterogeneous across countries.
  - Crisis-period evidence: Filho (2010) finds IT countries lowered nominal policy rates by more and achieved larger real rate loosening during the crisis; Roger (2010) finds IT countries may be less adversely affected by the financial crisis.

### Assessment of IT impact on low-income countries (LICs): empirical strategy
- Challenge: small number of LIC ITers and short IT duration; study uses inflation-targeting emerging countries as an approximation to infer potential LIC effects.
- Group construction and samples:
  - IT group: 10 countries (emerging economies that adopted IT prior to 2006).
  - LIC control group: 29 countries classified as “mature stabilizers” (based on 8 macroeconomic criteria) plus some that narrowly missed one criterion.
  - Non-IT emerging market control group: 18 countries (regional comparability with IT group, excluding European emerging markets).
- Data and indicator choices:
  - Four indicators: Inflation, inflation volatility, real GDP growth, growth volatility.
  - Annual inflation and real GDP growth from IMF’s World Economic Outlook (WEO).
  - Hyperinflation treatment: exclude countries with any period of inflation rates higher than the 50 percent threshold for baseline diff-in-diff (entire country series excluded if any year > 50 percent).
  - Countries excluded for >50 percent inflation episodes include: Brazil, Indonesia, Peru, Albania, Armenia, Azerbaijan, Georgia, Kyrgyz Republic, Moldova, Mongolia, Mozambique, Nigeria, Vietnam, Argentina, Dominican Republic, Jamaica, Kazakhstan, Suriname, Turkmenistan, Uruguay, Venezuela.
  - Baseline diff-in-diff sample sizes after exclusions: 26 countries when using LIC control group; 36 countries when using control group combining LICs and non-IT emerging markets.
- Timing:
  - For control (non-IT) countries the pre/post partition uses the average IT adoption date, the year 2002.
  - IT adoption dates taken from Roger (2010) for consistency.

### Methodology (difference-in-difference specification and robustness)
- Baseline diff-in-diff follows Ball and Sheridan (2005), controlling for reversion to the mean by including the pre-period level of the economic indicator.
- Core idea: regress the change of an economic indicator (post minus pre) on an IT policy dummy and the pre-period value of the indicator; coefficient on the IT dummy captures the IT effect, while pre-period value controls for regression-to-the-mean.
- Robustness checks:
  - Alternative control group composition: include non-IT emerging market economies to examine sensitivity of IT effects to control group choice.
  - Exclusion of high-inflation country series (>50 percent) for baseline; additional robustness checks with different country samples are conducted.
- Propensity score matching not used in this analysis because reviewed literature suggests similar results to diff-in-diff and the majority of studies use diff-in-diff or panel estimation methods.

### Key empirical observations and comparative averages (descriptive)
- ITers and non-ITers (LIC and emerging controls) share common macroeconomic characteristics in the pre-period: average inflation, growth, and growth volatility are broadly similar; non-ITers exhibit much higher inflation volatility.
- Average inflation fell in both IT and non-IT groups from pre- to post-period; decline was stronger in the IT group.
- Inflation and growth volatility fell from pre- to post-period in both groups.
- Average growth increased in both IT and control groups; post-period average growth in the LIC control group exceeded that in IT emerging market and non-IT emerging market economies, suggesting group-specific factors.

*Italic source: _wp11276 - 2004. Implementation was effective in 2009. The inflation target for the 2009-2011 period is*

### Appendix III provides more details on the methodology.

### 1.4 percentage points in the baseline case. The same results hold in all our robustness checks.

### _wp11276 - 1.4 percentage points in the baseline case. The same results hold in all our robustness checks.

### Major findings
- Adopting inflation targeting (IT) is associated with lower average inflation and lower inflation volatility.
- Baseline result: "1.4 percentage points in the baseline case. The same results hold in all our robustness checks."
- Cross-section (difference-in-difference) estimates find inflation reductions on the order of around 2 percentage points in IT countries relative to low-income control groups; panel estimates corroborate this magnitude.
- There is limited evidence that IT may be associated with smaller increases in real GDP growth in some specifications.
- Growth volatility is generally not affected in the cross-section (difference-in-difference) analysis but panel analysis finds significant reductions in growth volatility under some specifications.
- Estimated effects are generally robust to alternative control groups (LICs vs. non-IT emerging markets) except for the effects on growth, which do differ by control-group choice.

### Difference-in-difference (cross-section) results — key coefficients and patterns
- Inflation (Low-Income Country control group, Table 2):
  - IT Dummy: -1.96*, -2.06*, -2.21**, -2.33**, -1.62* (models 1–5 as reported).
  - Initial Conditions: -0.73***, -0.76***, -0.79***, -0.86***, -0.78***.
  - Constant: 3.90***, 5.06***, 5.45***, 5.91***, 4.64***.
  - R-squared: 0.76, 0.71, 0.75, 0.79, 0.78 (models 1–5).
- Inflation Volatility (Low-Income Country control group, Table 2):
  - IT Dummy: -1.41**, -1.54**, -1.64***, -1.49**, -1.47*** (models 1–5).
  - Initial Conditions: -0.89***, -0.90***, -0.93***, -0.89***, -0.89***.
  - Constant: 2.81***, 3.27***, 3.51***, 2.91***, 2.96***.
  - R-squared: 0.83, 0.52, 0.81, 0.82, 0.83 (models 1–5).
- Inflation (Combined Country control group, Table 3):
  - IT Dummy: -2.56**, -2.49*, -3.07***, -2.61**, -3.24*** (models 1–5).
  - Initial Conditions: -0.66***, -0.66***, -0.76***, -0.76***, -0.66***.
  - LIC Dummy (indicating differences between LICs and emerging markets): -0.38, -0.02, -0.83, -0.13, -1.57* (models 1–5).
  - R-squared: 0.74, 0.65, 0.72, 0.78, 0.72 (models 1–5).
- Inflation Volatility (Combined control group, Table 3):
  - IT Dummy: -0.57, -0.54, -1.14**, -0.73, -1.48*** (models 1–5).
  - Initial Conditions: -0.91***, -0.97***, -0.93***, -0.89***, -0.85***.
  - LIC Dummy: 0.88*, 1.05*, 0.50, 0.75, -0.13 (models 1–5).
  - R-squared: 0.87, 0.72, 0.83, 0.84, 0.83 (models 1–5).
- Growth (Low-Income Country control group, Table 4):
  - IT Dummy: -1.35**, -0.81, -1.55*, -1.66***, -2.10*** (models 1–5).
  - Initial Conditions: -0.71***, -0.64***, -0.62***, -0.80***, -1.19***.
  - Constant: 4.30***, 3.74***, 4.76***, 4.95***, 7.25***.
  - R-squared: 0.38, 0.46, 0.22, 0.45, 0.56 (models 1–5).
- Growth Volatility (Low-Income Country control group, Table 4):
  - IT Dummy: -0.45, -0.61, -0.80, -0.55, -0.65 (models 1–5; none significant).
  - R-squared: 0.76, 0.43, 0.38, 0.76, 0.72 (models 1–5).
- Growth (Combined Country control group, Table 5):
  - IT Dummy: -0.19, -0.13, -0.16, -0.44, -0.67 (models 1–5; generally not significant).
  - LIC Dummy: 1.16**, 0.70, 1.37**, 1.21**, 1.43* (models 1–5) — indicates growth differential is significantly different when using an emerging economy control group.
  - R-squared: 0.44, 0.52, 0.25, 0.53, 0.55 (models 1–5).
- Growth Volatility (Combined control group, Table 5):
  - IT Dummy: -0.09, -0.10, -0.57, -0.12, -1.12* (models 1–5).
  - R-squared: 0.75, 0.57, 0.40, 0.75, 0.67 (models 1–5).

### Panel analysis — data, methods, and key quantitative outcomes
- Data:
  - Country sample: 10 emerging market IT countries and 29 LICs (total N=39).
  - Sample period: 1990 to 2008 (T=27); yearly data summarized into three-year intervals to reduce instrument proliferation.
  - Inflation transformation: y_{i,t} = 100·log(1+Y_{i,t}/100) as in Brito and Bystedt (2010).
  - High inflation dummy = 1 if transformed inflation > 40 percent, 0 otherwise.
- Methodology:
  - Model estimated: dynamic panel with lagged dependent variable, IT dummy, high inflation dummy, time fixed effects, country fixed effects.
  - Estimators: pooled OLS, time-effects OLS (TE-OLS), country-and-time-effects OLS (CTE-OLS), difference GMM (D-GMM), two-step system GMM treating some variables as predetermined (S-GMM P) or endogenous (S-GMM E).
- Panel regression key outcomes (Tables 6 and 7):
  - Log Inflation (Table 6):
    - Inflation Targeting Dummy (columns 1–6 representative): -2.11***, -1.70***, -1.98, -1.91, -1.75**, -3.13*** (various estimators).
    - Lagged inflation coefficient: 0.10***, 0.09***, 0.04*, 0.03, 0.06, 0.07 (various estimators).
    - High Inflation Dummy: 33.06***, 29.98***, 30.08***, 34.08***, 33.94***, 34.20***.
    - Observations: 190 (most specifications); Instrument columns reported as 24, 34, 31 depending on estimator.
    - R-squared examples: 0.62, 0.68, 0.71 (OLS variants).
  - Log Inflation Volatility (Table 6, columns 7–12 representative):
    - Inflation Targeting Dummy: -2.20***, -1.02***, 3.28*, 7.61, -1.02***, -1.25*** (note some specifications produce positive coefficients; see text discussion).
    - Lagged volatility coefficient: 0.09***, 0.09***, 0.08***, 0.07***, 0.08***, 0.08***.
    - High Inflation Dummy: 19.06*, 16.66, 12.99***, 11.25, 74.57, 71.84 (estimates vary by estimator).
  - Log Growth (Table 7):
    - Inflation Targeting Dummy (columns 1–6 representative): -0.39, -1.22***, -0.33, -3.80, -1.35***, -2.03**.
    - Lagged growth coefficient: 0.29***, 0.22**, 0.01, 0.09, 0.15**, 0.12*.
    - High Inflation Dummy: -4.76, -4.87, -7.14***, -6.15, -1.70, 0.08.
    - Observations: 190 (most specifications).
  - Log Growth Volatility (Table 7, columns 7–12 representative):
    - Inflation Targeting Dummy: -0.73***, -0.47**, -0.03, 1.73, -0.56**, -0.59* (estimates vary by estimator).
    - Lagged volatility coefficient: 0.24***, 0.23***, -0.10, 0.09***, 0.15, 0.14***.
- Panel-summary interpretation from the authors:
  - Panel estimates support the cross-section finding that IT reduces inflation and inflation volatility.
  - Panel results show IT is associated with reductions in growth volatility (contrasting some previous literature).
  - The negative effect of IT on real GDP growth appears in some specifications but is not uniformly robust.
  - The high inflation dummy has large, significant coefficients for inflation in most specifications.

### Interpretation, caveats, and robustness
- The reduction in inflation and inflation volatility associated with IT is robust across methods (difference-in-difference and panel GMM), though magnitudes and significance vary by specification.
- The potential negative effect of IT on output growth:
  - Appears in some specifications, especially when using low-income control groups in cross-section analyses.
  - May reflect characteristics of LIC control groups rather than a causal effect of IT; in combined-control regressions the LIC dummy is significant and positive for growth (Table 5), suggesting the growth differential may be due to LIC characteristics.
- Caveats highlighted by the authors:
  - Comparisons of advanced LICs with emerging markets that adopted IT may be imperfect; sample selection attempts to match macroeconomic performance but limitations remain.
  - Short post-adoption time spans: many countries have only short-to-medium run experience with IT; observed effects may not reflect steady-state outcomes.
  - IT adoption often coincides with broader structural and policy reforms, complicating causal attribution.
  - Endogeneity concerns addressed via panel GMM, but specification sensitivity remains (different estimators yield different significance for some coefficients).

### Conditions and contextual factors for adopting IT
- Standard preconditions often cited (and noted as generally unmet in practice) include:
  - Full central bank independence.
  - Well-developed technical infrastructure to forecast/model inflation.
  - Economic structures with limited sensitivity of domestic prices to commodity prices and exchange rates.
  - A healthy financial system.
- Empirical observation: "no inflation targeter had these preconditions fully in place before adopting inflation targeting," implying failure to meet them is not by itself an impediment to adopting IT.
- Factors that may weaken the efficacy of IT include:
  - Managed exchange rate regimes associated with a fear of floating.
  - A narrow base of domestic nominal financial assets.
  - Lack of market instruments to hedge exchange rate risk.
  - Dollarization.
- Evidence exists that IT can be adapted successfully in highly dollarized economies (references noted by the authors).

*Italicized source attribution provided from the content unit.*

### Box 3. PreConditions for the Adoption of Inflation Targeting

### Box 3. PreConditions for the Adoption of Inflation Targeting

### Preconditions (four broad categories)
- Institutional independence: central bank must have full legal autonomy and be free from fiscal and/or political pressure that would create conflicts with the inflation objective.
  - (i) absence of fiscal obligation
  - (ii) operational independence
  - (iii) inflation-focused mandate
  - (iv) favorable fiscal balance
  - (v) low public debt
  - (vi) central bank independence
- A well-developed technical infrastructure: inflation forecasting and modeling capabilities, and the data needed to implement them, must be available at the central bank.
- Economic structure: prices must be fully deregulated, the economy should not be overly sensitive to commodity prices and exchange rates, and dollarization should be minimal.
  - Indicators capturing relevant economic conditions:
    - (i) low exchange rate pass-through
    - (ii) low sensitivity to commodity prices
    - (iii) low dollarization
    - (iv) extent of trade openness
- A healthy financial system: banking system should be sound, and capital markets well developed to minimize conflicts with financial stabilization objectives and guarantee effective monetary policy transmission.

### Factors that can enhance adoption (and caveats)
- Enhancing factors:
  - (a) a well functioning monetary transmission mechanism
  - (b) a degree of independence of monetary policy
  - (c) absence of commitment to a particular level for the exchange rate
- Caveat: A country satisfying these requirements could choose to conduct its monetary policy under an IT framework despite not having all the preconditions. Nonetheless, several features can impact the effectiveness of an IT framework.

### A. Weak Monetary Transmission Mechanism (implications)
- Implementing an inflation targeting regime requires a clear understanding of the monetary policy transmission mechanism and methods of projecting inflation consistent with that understanding.
- Adopting IT in LICs is often constrained by:
  - weak monetary policy transmission mechanism
  - rudimentary understanding of transmission
- Additional impediments in LICs (Mishra et al., 2010) that impair traditional channels (interest rate, bank lending, and asset price):
  - weak institutional frameworks
  - reduced role of securities markets
  - imperfect competition in the banking sector
  - resulting high cost of bank lending to private firms

### B. Independent Monetary Policy (constraints and manifestations)
- Scope for independent monetary policy tends to be hampered by:
  - heavy reliance on seigniorage
  - shallow financial markets
  - fragile banking systems (Masson et al., 2007)

- Reliance on Seigniorage (fiscal dominance indicators):
  - link between government's inability to raise conventional revenue and recourse to seigniorage
  - in developing countries, link is often strong due to structural features (concentrated and unstable sources of tax revenue, poor tax collection procedures, skewed income distributions)
  - government monetizing debt or pressuring central bank to maintain expansionary monetary policy are manifestations

- Financial Markets:
  - shallow capital markets can indicate fiscal dominance
  - common causes include government extraction of revenue via financial repression: interest rate ceilings, high reserve requirements, sectoral credit policies, compulsory placements of public debt
  - underdeveloped foreign exchange markets have held back LICs from adopting IT

- Banking Systems:
  - fragile banking systems often result from prolonged financial repression
  - post-reform banking systems can influence monetary policy independently, but conflicts may arise between price stability and restoring banking sector profitability

### C. Conflicts with other Objectives
- Necessary IT features (primacy of inflation target, forward-looking procedure using inflation forecasts) are difficult to satisfy when exchange rate stability is a stated or implicit objective.
- If an inflation target co-exists with other monetary policy objectives, tension between the inflation target and other objectives is unavoidable, diminishing the benefits of IT.

### Conclusion (main findings)
- Inflation targeting has been gaining popularity over the last two decades but is fairly new in LICs.
- To date only three LICs have officially declared themselves inflation targeters: Ghana and Armenia, and Albania.
- Several countries (such as Moldova, Georgia, etc.) could be described as informal inflation targeters.
- Two different methods used in the paper indicate that shifting to IT will lower inflation and its volatility in LICs.
- There is limited evidence of a trade-off between inflation and output in the data.
  - "The negative impact on growth might be due to the LIC characteristics and not the adoption of inflation targeting."
- Benefits of IT can be limited when basic conditions are not in place; economic structure of LICs often makes IT adoption more challenging.
- Adoption of IT may still be appropriate for LICs because it allows central bankers to pay considerable attention to monetary aggregates or the exchange rate when these variables contain valuable information or play a role in the monetary transmission mechanism.

*Box 3. PreConditions for the Adoption of Inflation Targeting — source document excerpt*

### Appendix III. Difference–in–Difference Method

### Appendix III. Difference–in–Difference Method

### Model specification
- Core dynamic process (equation (1)):
  - X_i,t = φ[α_T d_i,t + α_N (1 – d_i,t)] + (1 – φ)X_i,t–1    (1)
- Definitions:
  - X_i,t is the value of a macroeconomic performance indicator X for country i at time t.
  - α_T is the mean to which X reverts for inflation targeters.
  - α_N is the mean to which X reverts for non inflation-targeters.
  - d_i,t is a variable equal to 1 for inflation targeters and 0 for non-inflation targeters.
  - φ represents the speed with which X reverts to its group-specific α:
    - φ = 1 means X reverts completely after one period.
    - φ = 0 means X depends only on its past history, with no tendency to revert to any particular value.

### Regression form used (two-period difference–in–difference)
- Change form with an error term e and two periods (“pre” and “post”) (equation (2)):
  - X_i,post – X_i,pre = φα_T d_i + φα_N (1 – d_i) – φX_i,pre + e_i    (2)
- Reparameterization letting a_0 = φα_N, a_1 = φ(α_T – α_N), and b = –φ yields (equation (3)):
  - X_i,post – X_i,pre = a_0 + a_1 d_i + b X_i,pre + e_i    (3)

### Interpretation of parameters and estimand
- Relevant treatment effect:
  - a_1 (the coefficient on the inflation targeting dummy d_i) is the parameter for gauging inflation targeting’s economic impact.
- Role of a_0:
  - a_0 captures whether there has been a generalized improvement in macroeconomic performance across countries independently of differences in monetary regimes.
- Special case when φ = 0:
  - If φ were known to be zero (i.e., complete mean reversion), the estimated a_1 would be the difference in average X_post – X_pre for inflation targeters versus non-inflation-targeters.
  - The regression’s advantage in that case is primarily controlling for the initial level X_pre.

### Scope and limitations
- Focus on relatively long periods:
  - The analysis largely compares steady states.
  - It “says nothing about what happens during the transition to inflation targeting (or any other) policy framework.”
- Transition effects:
  - Identifying transition dynamics would “require a very careful control of cyclical conditions to distinguish transition effects from the normal trajectory of the business cycle.”

*Appendix III. Difference–in–Difference Method — Source: _wp11276 - Appendix III. Difference–in–Difference Method*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11276.pdf_