## _wp11112 - 4. Dynamic Effect of De Jure and De Facto Pegs on Inflation

## Source details

**Canonical URL:** [_wp11112 - 4. Dynamic Effect of De Jure and De Facto Pegs on Inflation](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11112.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11112.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11112.pdf.json)

---

### Introduction and research question
- Context: rising inflationary pressures in emerging market and developing countries (EMDCs) after the global financial crisis, driven in part by capital inflows and higher food and fuel prices.
- Core hypothesis: a formal (de jure) commitment to a peg matters for inflation performance because it provides credibility that complements any de facto behavior; many de facto pegs lack a corresponding de jure commitment (closet pegs), while some de jure pegs are not de facto maintained (soft pegs).
- Empirical strategy advantage: use a unique IMF dataset that contains both de jure and de facto exchange rate regime classifications from a common source to enable directly comparable tests.

### Theoretical framework — setup and intuition
- Model foundation: builds on policy credibility models of Barro and Gordon (1983a, b) applied to exchange rate regimes.
- Key equations and relations (as presented):
  - Output: y = θ(πd − πed) + η, with shock η mean zero and variance 2
η
σ
.
  - CPI inflation: π = βπd + (1−β)πi, with β ∈ [0,1].
  - Money demand / domestic inflation link: πd = Δm + vπed, with 0 ≤ v < 1.
  - Imported inflation via PPP: πi = Δe (foreign inflation normalized to zero).
  - Monetary supply: Δm = ΔDC + ΔR; under a peg Δe = 0 ⇒ Δm = 0 when reserves sterilize flows.
  - Central bank loss: L = E{(π)^2 + A(y − y*)^2} with A relative weight on output.
- Model regime comparisons and ranking (equation references preserved):
  - Pure peg (de jure commit and de facto maintain): π = 0 under full credibility; welfare loss LPeg given by expressions in equations (9) and (10).
  - Free float: discretionary ΔDC yields positive expected inflation (equations (11)–(18)).
  - Hybrid regimes:
    - Closet peg (de facto peg, no formal commitment): private sector assigns probability ρ that peg survives; expected inflation given by equation (22); positive CPI inflation can result even if Δe = 0 (equation (23)).
    - Soft peg (de jure peg, de facto non-peg): low credibility (ρ < 1) and positive inflation (equations (24)–(26)); soft pegs unlikely to last.
  - Ranking of average inflation rates (model result, equation (29)):
    - πFlt > πSoftPeg > πClosetPeg > πPurePeg

### Empirical specification and identification
- Benchmark regression (equation (30)):
  - πit = β0 + βPeg Pegit + βInt Intit + βMon Δm it + X it βX + v t + ε it
  - Dependent variable: inflation transformed to π/(1+π) to reduce hyper-inflation effects.
  - Controls X: real GDP growth, trade openness, central bank independence (proxied by central bank governor turnover rate), fiscal balance, terms of trade shocks.
  - Year effects vt included; country fixed effects omitted in main specification but used in robustness checks.
- Money growth endogeneity addressed:
  - Money-growth equation (31): Δm it = α0 + αPeg Pegit + αInt Intit + αX X it + υ it
  - Total effect of peg computed as γPeg = βPeg + βMon · αPeg.
- Hybrid-regime distinction (equation (32)):
  - πit = β0 + βPP PurePegit + βCP ClosetPegit + βSP SoftPegit + βMon Δm it + X it βX + vt + ε it
- Estimation methods:
  - Instrumental variable estimation to address endogeneity (instruments include lagged real GDP growth, fiscal balance, money growth).
  - Time effects included; t-statistics based on clustered standard errors.

### Data and sample
- Dataset: unique IMF dataset with de jure and de facto classifications (AREAR for de jure; de facto compiled by Anderson (2008)).
- Sample period: 1980–2010.
- Sample composition: 145 EMDCs emphasized for main empirical results.
- Key variable descriptive statistics (selected, as reported in Appendix A):
  - Broad money growth: Obs. 1,937; Mean 0.186; Std. Dev 0.224.
  - Inflation (transformed inflation/(1+inflation)): Obs. 1,937; Mean 0.085; Std. Dev 0.095.
  - Central bank turnover rate: Obs. 1,937; Mean 0.195; Std. Dev 0.200.
  - Fiscal balance to GDP: Obs. 1,932; Mean -0.029; Std. Dev 0.051.
  - Population size (millions): Obs. 1,937; Mean 48.423; Std. Dev 170.150.
  - Trade openness: Obs. 1,937; Mean 0.768; Std. Dev 0.407.

### Main empirical findings
- Aggregate relationships and stylized facts:
  - Inflation is systematically lower by about 5 percentage points per year in countries with pegs compared to countries with intermediate or floating regimes.
  - Broad money growth is, on average, 9 percentage points per year lower under pegs than under other regimes.
  - For a given rate of money growth, inflation is lower under pegged regimes (a “confidence effect”).
- Benchmark estimation (selected coefficients and interpretation):
  - Table 4 (All countries, De Jure classification):
    - Pegged regimes coefficient: -0.029 (t-stat -4.83 ***) implying lower inflation under pegs relative to floats.
    - Money growth coefficient: 0.391 (t-stat 6.41 ***).
    - GDP growth coefficient: -0.459 (t-stat -3.86 ***).
    - Openness coefficient: -0.026 (t-stat -6.50 ***).
    - Central bank turnover rate: 0.040 (t-stat 4.44 ***).
    - Fiscal balance (pct. of GDP): -0.108 (t-stat -2.20 **).
    - Number of observations, R2: 2,024; 0.432.
  - Table 4 (All countries, De Facto classification):
    - Pegged regimes coefficient: -0.063 (t-stat -11.76 ***) — example interpretation: "coefficient of -0.063 for pegged regimes implies 6.3 percent per year lower inflation under pegged exchange rate regimes compared to floats."
    - Money growth coefficient: 0.404 (t-stat 5.69 ***).
    - Number of observations, R2: 1,881; 0.419.
- De jure versus de facto pegs:
  - De jure pegs: about 6 percentage points lower inflation vs floats (full sample).
    - Decomposition: about 3 percentage points from lower money growth (discipline effect), about 3 percentage points from confidence effect.
    - Subsamples: EMEs ~9 percentage points lower; developing countries ~5 percentage points lower.
  - De facto pegs: about 3 percentage points lower inflation vs floats (full sample).
    - Of this, about 2 percentage points due to the confidence effect.
    - Effect smaller in EMEs (~2 percentage points) and larger in developing countries (~6 percentage points).
  - Consensus sample (de facto pegs that are also de jure pegs): de facto peg differential rises to 4 percentage points (full sample); 4 (EME); 8 (developing).
  - Direct comparison (equation (32), Table 6): when de facto peg is supported by de jure peg, inflation relative to non-pegs is lower by about 6 and 3 percentage points in the EMEs and developing country samples, respectively, than when a de facto peg is not supported by a de jure peg.
  - Soft pegs (de jure but not de facto): rare; imply somewhat lower inflation than pure floats but do not deliver full benefits.
- Duration effects (Table 7):
  - Peg duration phases in: about a 10 percentage points reduction in inflation during the first ten years of pegging; only a 0.1 percentage point reduction in the following ten years.
  - Five-year effects:
    - Five years under a de jure peg associated with 5 percentage points lower inflation relative to a float.
    - Five years under a de facto peg associated with 3 percentage points lower inflation relative to a float.
  - Dropping de facto pegs not classified as de jure pegs strengthens de facto peg effects; consensus sample raises the five-year differential by about 1 percentage point for both EMEs and developing countries.
- Capital inflows and current account surpluses (Table 8):
  - Large capital inflows threshold: inflows above 2.5 percent of GDP (the full sample’s 30th percentile for positive net flows).
  - Large current account surplus threshold: 2 percent of GDP (about the full sample’s 30th percentile for positive balances).
  - Under large capital inflows: money growth higher for pegged regimes; inflation differential between floating and pegged regimes remains positive.
  - Under large current account surpluses: money growth sufficiently higher under pegged regimes that the differential moves in favor of floating exchange rates.
    - For the full sample, de jure and de facto pegs associated with 4-6 percentage points higher inflation than floating regimes (effect stems entirely from the money growth channel).
    - For developing countries, pegs do not have significantly higher inflation than floating regimes even under large current account surpluses.
  - Anchor-country inflation: enters strongly positive; higher anchor inflation is imported by pegged countries. For the EMDCs over the sample period, anchor inflation was not sufficiently high to make pegging yield higher inflation than intermediate regimes or floats.
- Endogeneity and robustness (Table 9 and other tests):
  - Controlling for regime endogeneity via simultaneous equations (probit for regime choice and second-stage inflation regression with exclusion variables population size and geographic concentration ratio of exports) yields little change: de jure and de facto pegs associated with about 7 and 4 percentage points lower inflation, respectively, than intermediate and floating regimes (full sample).
  - Excluding freely falling regimes (floaters with depreciation >25% per year or inflation >40% per year) does not change main finding—pegged regimes retain about a 1-3 percentage point lower inflation differential vis-à-vis floaters (full sample).
  - Excluding hard pegs (currency boards and monetary unions) from pegged category still yields statistically significant inflation differential favoring pegs vis-à-vis floats.
  - Results robust to country-specific effects, different time periods, and alternative sample splits (coefficients modestly smaller in some specifications).

### Key quantitative findings (selected)
- Inflation lower by about 5 percentage points per year under pegs (aggregate statistic).
- Broad money growth lower by about 9 percentage points per year under pegs.
- De jure pegs: ~6 percentage points lower inflation vs floats (full sample); ~9 (EMEs); ~5 (developing countries).
  - Of the 6 percentage points: ~3 from lower money growth, ~3 from credibility/confidence.
- De facto pegs: ~3 percentage points lower inflation vs floats (full sample); ~2 (EMEs); ~6 (developing countries).
- Consensus (de facto pegs that are also de jure pegs): de facto peg differential rises to 4 percentage points (full sample); 4 (EME); 8 (developing).
- Matched de facto + de jure pegs vs unmatched: inflation lower by ~6 percentage points (EME) and ~3 percentage points (developing) when matched.
- Duration: ~10 percentage points reduction in first ten years of a peg; only 0.1 percentage point reduction in next ten years.
- Large capital inflows threshold: 2.5 percent of GDP; large current account surplus threshold: 2 percent of GDP.
- Under large current account surpluses: de jure and de facto pegs associated with 4-6 percentage points higher inflation (full sample).

### Policy-relevant implications and conclusions
- Two channels through which pegs lower inflation:
  - Monetary discipline: lower money growth under pegs contributes materially to lower inflation.
  - Credibility/confidence: formal commitment (de jure peg) lowers inflation for a given money growth rate.
- Words and deeds both matter:
  - De jure commitment strengthens the inflation-reducing benefits of a peg; de facto pegs not backed by de jure commitment (closet pegs) deliver weaker inflation benefits.
  - Soft pegs (de jure peg not supported by de facto behavior) are rare and do not deliver full benefits.
- Duration matters: credibility builds over time, with largest gains in the first decade of pegging.
- Policy takeaway:
  - Promising to maintain the parity and delivering on that promise yields the greatest benefits for price stability.
  - A fuller policy assessment should weigh inflation benefits against other outcomes (for example, susceptibility to currency crises), which is left for future research.

*Source: IMF working paper chapter “4. Dynamic Effect of De Jure and De Facto Pegs on Inflation” (excerpt) from _wp11112.*

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

### _wp11112 - References .............................................................................................................

### Appendix and Tables (Major Listings)
- Appendix A. Data Description .......................................................................................................33

- Tables
  - 1. Classification of Exchange Rate Regimes, 1980–2010 .............................................................27
  - 2. Distribution of De Jure and De Facto Classifications, 1980–2010 ............................................27
  - 3. Inflation and Money Growth under Alternative Regimes, 1980–2010 .....................................27
  - 4. Inflation Performance: Benchmark Specification ......................................................................28
  - 5. Inflation Performance: Consensus Between De Jure and De Facto Pegs  ................................29
  - 6. Inflation Performance: Consensus Between De Jure and De Facto Classifications ..................29
  - 7. Inflation Performance: Regime Duration ...................................................................................30
  - 8. Inflation Performance under Balance of Payments Surpluses. ..................................................31
  - 9. Inflation Performance: Regime Endogeneity, Freely Falling, and Hard Pegs ...........................32
  - A1. Agreement of Different Exchange Rate Regime Classifications, 1980–2004 ........................34
  - A2. List of Countries in the Sample ...............................................................................................35
  - A3. Data Description and Sources .................................................................................................35

### Figures
- 1. Frequency Distribution of Exchange Rate Regimes ..................................................................24
- 2. Money Growth and Inflation under Alternative De Jure and De Facto Regimes ......................25
- 3. Inflation and Money Growth under Pegs and Non-Pegs ...........................................................25

*Source: _wp11112 - References (PDF chapter/section).*

### 4. Dynamic Effect of De Jure and De Facto Pegs on Inflation ......................................................26

### _wp11112 - 4. Dynamic Effect of De Jure and De Facto Pegs on Inflation

### Introduction and research question
- Context: rising inflationary pressures in emerging market and developing countries (EMDCs) after the global financial crisis, driven in part by capital inflows and higher food and fuel prices.
- Core hypothesis: a formal (de jure) commitment to a peg matters for inflation performance because it provides credibility that complements any de facto behavior; many de facto pegs lack a corresponding de jure commitment (closet pegs), while some de jure pegs are not de facto maintained (soft pegs).
- Empirical strategy advantage: use a unique IMF dataset that contains both de jure and de facto exchange rate regime classifications from a common source to enable directly comparable tests.

### Theoretical framework — setup and intuition
- Model foundation: builds on policy credibility models of Barro and Gordon (1983a, b) applied to exchange rate regimes.
- Key ingredients:
  - Output: y = θ(πd − πed) + η, with shock η mean zero and variance 2
η
σ
.
  - CPI inflation: π = βπd + (1−β)πi, with β ∈ [0,1].
  - Money demand / domestic inflation link: πd = Δm + vπed, with 0 ≤ v < 1.
  - Imported inflation via PPP: πi = Δe (foreign inflation normalized to zero).
  - Monetary supply: Δm = ΔDC + ΔR; under a peg Δe = 0 ⇒ Δm = 0 when reserves sterilize flows.
  - Central bank loss: L = E{(π)^2 + A(y − y*)^2} with A relative weight on output.

### Model comparisons of regimes and formal results
- Pure peg (de jure commit and de facto maintain):
  - Δe = 0, Δm = 0, expectations πed = 0 ⇒ π = 0 (central bank’s “bliss” inflation).
  - Welfare loss for a realization: LPeg = 1/2 Ayη^2 − Ayη + ... (equations (9) and (10) provide exact forms).
- Free float:
  - Central bank optimally chooses ΔDC; discretionary policy yields positive expected inflation (equations (11)–(18)).
- Hybrid regimes:
  - Closet peg (de facto peg, no formal commitment): private sector assigns probability ρ that peg survives; expected inflation given by equation (22); even if Δe = 0, positive CPI inflation can result (equation (23)).
  - Soft peg (de jure peg, de facto non-peg): credibility low (ρ < 1), discretionary monetary policy inconsistent with parity, positive inflation (equations (24)–(26)); soft pegs unlikely to last.
- Ranking of average inflation rates (model result, equation (29)):
  - πFlt > πSoftPeg > πClosetPeg > πPurePeg
  - Interpretation: both monetary discipline (de facto maintenance ⇒ Δm = 0) and confidence (credible de jure commitment ⇒ low expected inflation) matter.

### Empirical specification and identification
- Benchmark regression (equation (30)):
  - πit = β0 + βPeg Pegit + βInt Intit + βMon Δm it + X it βX + v t + ε it
  - Dependent variable: inflation transformed to π/(1+π) to reduce hyper-inflation effects.
  - Controls X: real GDP growth, trade openness, central bank independence (proxied by central bank governor turnover rate), fiscal balance, terms of trade shocks.
  - Year effects vt included; country fixed effects omitted in main specification but used in robustness checks.
- Money growth potential endogeneity addressed by estimating:
  - Money-growth equation (31): Δm it = α0 + αPeg Pegit + αInt Intit + αX X it + υ it
  - Total effect of peg: γPeg = βPeg + βMon · αPeg; total effect of intermediate regime analogous.
- Additional specification to distinguish hybrids (equation (32)):
  - πit = β0 + βPP PurePegit + βCP ClosetPegit + βSP SoftPegit + βMon Δm it + X it βX + vt + ε it

### Data and sample
- Dataset: unique IMF dataset containing both de jure and de facto classifications from a common source (AREAR for de jure; de facto classification compiled by Anderson (2008) using IMF country teams, staff reports, and consultations).
- Sample period: 1980–2010.
- Sample composition emphasized: 145 EMDCs (explicit sample size used for main empirical results).

### Main empirical findings (summary)
- Pegged exchange rates are associated with significantly lower inflation than intermediate or floating regimes after controlling for other determinants.
- The effect is much stronger for de jure pegs than for de facto pegs.
- The relatively smaller impact of de facto pegs is concentrated in cases where the de facto peg is not supported by a de jure commitment (i.e., closet pegs): when restricting the sample to cases where de facto peg is also de jure peg, the estimated association between de facto pegs and low inflation strengthens significantly.
- Duration effect: longer duration of the regime is associated with lower inflation, though at a diminishing rate.
- Robustness: results are robust to:
  - Accounting for potential endogeneity of regime choice (instrumental variables using lagged real GDP growth, fiscal balance, and money growth).
  - Exclusion of extreme cases (freely falling regimes and hard pegs).
- Empirical transformations and notes:
  - Inflation transformed to π/(1+π) to reduce impact of hyper-inflation observations.
  - Fewer than 2 percent of de jure pegs are de facto floats (noted in model discussion).

### Policy-relevant implications and interpretation
- Formal commitment (de jure peg) matters for anchoring inflation expectations in EMDCs, beyond mere de facto exchange rate stability.
- De facto interventions without a supportive formal commitment (closet pegs) yield weaker inflation-anchoring benefits than when deeds are explicitly backed by words.
- Where central banks cannot credibly commit, pegging may be less effective; conversely, where commitment is credible, pegging can solve the time-consistency problem and produce substantially lower inflation.
- Duration of credible pegging regimes reinforces credibility and lowers inflation over time, but with diminishing returns.

*Source: IMF working paper chapter “4. Dynamic Effect of De Jure and De Facto Pegs on Inflation” (excerpt), using IMF de jure and de facto classifications and a sample of 145 EMDCs over 1980–2010.*

### Appendix I establishes that our de facto classification is

### _wp11112 - Appendix I establishes that our de facto classification is

### Exchange rate regime classifications and sample
- The IMF’s de jure and de facto classifications group exchange rate regimes into eight categories: (i) exchange arrangement with no separate legal tender; (ii) currency board arrangement; (iii) conventional pegged arrangement (further divided into single and basket pegs); (iv) pegged exchange rates within horizontal bands; (v) crawling peg; (vi) crawling band; (vii) managed float with no predetermined path for the exchange rate; and (viii) independently floating arrangement.
- For the empirical analysis:
  - The first three arrangements (excluding basket pegs) are grouped into the pegged exchange rate regime category.
  - The next four (including basket pegs) are grouped as intermediate regimes.
  - The last one is classified as the floating regime.
- Sample: 145 EMDCs over the period 1980–2010 (see Figure 1 and Appendix A for country selection and data descriptions).

### Observed patterns in regime evolution
- Three discernible patterns across 1980–2010:
  - Floating regimes became more popular among EMEs since the 1997/98 Asian crises, without a corresponding decrease in the proportion of pegged regimes.
  - A marked “hollowing out of the middle” (bipolarity): proportion of intermediate regimes in EMEs shrank from around 64 percent in the 1980s and 1990s to about 34 percent in the last decade.
  - Significant divergence between de jure and de facto classifications:
    - Almost 95 percent of de jure pegs are also de facto pegs.
    - Of de facto pegs, only about 70 percent are also de jure pegs (closet pegs more common than soft pegs).
    - Only in 55 percent of cases where the currency de jure floats does it also de facto float.

### Stylized facts on inflation, money growth, and regimes
- Average differences (Table 3, Figure 2):
  - Inflation is systematically lower by about 5 percentage points per year in countries with pegs compared to countries with intermediate or floating regimes.
  - Broad money growth is, on average, 9 percentage points per year lower under pegs than under other regimes.
  - For a given rate of money growth, inflation is lower under pegged regimes (a “confidence effect”), with the difference initially increasing in money growth.
- De jure vs de facto:
  - Inflation and money growth are slightly lower under de jure pegs than de facto pegs; the confidence effect is greater for de jure pegs.

### Baseline estimation results (Table 4 and related)
- Determinants of inflation enter with expected signs (terms of trade not significant): higher real GDP per capita growth, stronger fiscal position, and greater trade openness → lower inflation; higher money growth and higher central bank turnover → higher inflation.
- Estimated impacts (benchmarks):
  - Inflation under (de jure) pegged regimes is about 6 percentage points per year lower than under floating exchange rates.
    - Of this 6 percentage points: about 3 percentage points are associated with lower money growth (discipline effect) and about 3 percentage points represent the confidence effect (lower inflation for a given rate of money growth).
  - Pegs vs intermediate regimes: pegged regimes show about a 5.5 percentage point lower inflation relative to intermediate regimes.
  - Sub-samples:
    - Emerging markets: de jure pegs associated with about 9 percentage points lower inflation than floating regimes.
    - Developing countries: de jure pegs associated with about 5 percentage points lower inflation than floating regimes.
  - De facto pegs:
    - Associated with about 3 percentage points lower inflation than floats for the full sample.
    - Of this, about 2 percentage points are due to the confidence effect.
    - The effect is smaller in EMEs (~2 percentage points) and larger in developing countries (~6 percentage points).

### Role of formal commitment (deeds vs words)
- Excluding de facto pegs that are not also de jure pegs (consensus sample) increases the estimated inflation differential of de facto pegs relative to floats to 4 percentage points for the full consensus sample (4 and 8 percentage points for the EME and developing country samples, respectively).
- Direct comparison (equation (32), Table 6):
  - When a de facto peg is supported by a de jure peg, inflation relative to non-pegs is lower by about 6 and 3 percentage points in the EMEs and developing country samples, respectively, than when a de facto peg is not supported by a de jure peg.
  - De jure pegs not matched by de facto pegs (soft pegs) imply somewhat lower inflation than pure floats (but these cases are rare).

### Dynamic effects and duration of pegs (Table 7)
- Peg duration matters: the effect of pegs phases in over time using an (inverse of) years-in-peg variable.
- Average effects:
  - About a 10 percentage points reduction in inflation during the first ten years of pegging.
  - Only a 0.1 percentage point reduction in the following ten years.
- Five-year effects:
  - Five years under a de jure peg is associated with 5 percentage points lower inflation relative to a float.
  - Five years under a de facto peg is associated with 3 percentage points lower inflation relative to a float.
- Dropping de facto pegs not classified as de jure pegs strengthens de facto peg effects; consensus sample raises the five-year differential by about 1 percentage point for both EMEs and developing countries.

### Capital inflows, current account surpluses, and anchor-country inflation
- Definitions used for conditional samples:
  - Large capital inflows: inflows above 2.5 percent of GDP (the full sample’s 30th percentile for positive net flows).
  - Large current account surplus: 2 percent of GDP (about the full sample’s 30th percentile for positive balances).
- Findings (Table 8):
  - Under large capital inflows, money growth is higher for pegged regimes; nonetheless, the inflation differential between floating and pegged regimes remains positive.
  - Under large current account surpluses, money growth is sufficiently higher under pegged regimes that the differential moves in favor of floating exchange rates.
    - For the full sample, de jure and de facto pegs are associated with 4-6 percentage points higher inflation than floating regimes (effect stems entirely from the money growth channel).
    - For developing countries, pegs do not have significantly higher inflation than floating regimes even under large current account surpluses.
- Including anchor-country inflation:
  - Anchor-country inflation enters strongly positive; higher anchor inflation is imported by pegged countries.
  - For the EMDCs over the sample period, anchor inflation was not sufficiently high to make pegging yield higher inflation than intermediate regimes or floats.

### Regime endogeneity and robustness (Table 9 and other tests)
- Addressing possible endogeneity via a simultaneous equation system (probit for regime choice and second-stage inflation regression) using exclusion variables (population size and geographic concentration ratio of exports) yields:
  - Little change in estimated effects: for the full sample, de jure and de facto pegs are associated with about 7 and 4 percentage points lower inflation, respectively, than intermediate and floating regimes.
- Freely falling regimes:
  - Excluding floaters where nominal depreciation > 25 percent per year or inflation > 40 percent per year does not change the main finding—pegged regimes retain about a 1-3 percentage point lower inflation differential vis-à-vis floaters for the full sample.
- Hard pegs and other robustness checks:
  - Excluding hard pegs (currency boards and monetary unions) from the pegged category still yields a statistically significant inflation differential in favor of pegs vis-à-vis floats.
  - Results robust to country-specific effects, different time periods, and alternative sample splits (though coefficients modestly smaller in some specifications).

### Key quantitative findings (selected)
- Inflation lower by about 5 percentage points per year under pegs (aggregate statistic).
- Broad money growth lower by about 9 percentage points per year under pegs.
- De jure pegs: ~6 percentage points lower inflation vs floats (full sample); ~9 (EMEs); ~5 (developing countries).
  - Of the 6 percentage points: ~3 from lower money growth, ~3 from credibility/confidence.
- De facto pegs: ~3 percentage points lower inflation vs floats (full sample); effect smaller in EMEs (~2) and larger in developing countries (~6).
- Consensus (de facto pegs that are also de jure pegs): de facto peg differential rises to 4 percentage points (full); 4 (EME); 8 (developing).
- Matched de facto + de jure pegs vs unmatched: inflation lower by ~6 percentage points (EME) and ~3 percentage points (developing) when matched.
- Duration: ~10 percentage points reduction in first ten years of a peg; only 0.1 percentage point reduction in next ten years.
- Large capital inflows threshold: 2.5 percent of GDP; large current account surplus threshold: 2 percent of GDP.
- Under large current account surpluses: de jure and de facto pegs associated with 4-6 percentage points higher inflation (full sample).

### Policy-relevant implications and conclusions
- Pegs operate through two channels:
  - Monetary discipline: lower money growth under pegs contributes materially to lower inflation.
  - Credibility/confidence: formal commitment (de jure peg) lowers inflation for a given money growth rate.
- Words and deeds both matter:
  - De jure commitment strengthens the inflation-reducing benefits of a peg; de facto pegs not backed by de jure commitment (closet pegs) deliver weaker inflation benefits.
  - Soft pegs (de jure peg not supported by de facto behavior) are rare but, where present, do not deliver full benefits.
- Policy takeaway:
  - Promising to maintain the parity and delivering on that promise yields the greatest benefits for price stability.
  - A fuller policy assessment would weigh inflation benefits against other outcomes (for example, susceptibility to currency crises), which is left for future research.

*Source: _wp11112 - Appendix I establishes that our de facto classification is*

### REFERENCES

### _wp11112 - REFERENCES

### References (selected citations as listed)
- Alesina, A., and Wagner, A. F., 2006, “Choosing (and Reneging on) Exchange Rate Regimes,” Journal of the European Economic Association, Vol. 4, pp. 770–799.
- Anderson, H., 2008, “Exchange Policies before Widespread Floating (1945–89),” mimeo (Washington DC: International Monetary Fund).
- Barro, R., and D. Gordon, 1983a, “A Positive Theory of Monetary Policy in a Natural-Rate Model,” Journal of Political Economy, Vol. 91, pp. 589-610.
- Barro, R. and D. Gordon, 1983b, “Rules, Discretion, and Reputation in a Model of Monetary Policy,” Journal of Monetary Policy, Vol. 12, pp. 101–121.
- Baxter, M. and A. Stockman, 1989, “Business Cycles and the Exchange Rate Regime: Some International Evidence,” Journal of Monetary Economics, Vol. 23, pp 377–400.
- Bleaney, M., and M. Francisco, 2005, “Exchange Rate Regimes and Inflation: Only Hard Pegs Make a Difference,” Canadian Journal of Economics, Vol. 38, pp 1453–1471.
- Calvo, Guillermo and Carmen Reinhart, 2002, “Fear of Floating,” Quarterly Journal of Economics, Vol. 117, pp. 379–408.
- Cooper. R, 1971, “Currency Devaluation in Developing Countries,” Essays in International Finance No. 86 (Princeton, NJ: Princeton University).
- Fischer, Stanley, 2001, “Exchange Rate Regimes: Is the Bipolar View Correct?” Journal of Economic Perspectives, Vol. 15, pp. 3-24.
- Frankel, J., 2005, “Contractionary Currency Crashes in Developing Countries,” IMF Staff Papers, Vol. 52, pp. 149-192.
- Genberg, H., and A. K. Swoboda, 2005, “Exchange Rate Regimes: Does What Countries Say Matter?” IMF Staff Papers, Vol. 52, pp. 129-141.
- Ghosh, A., A. Gulde, J. Ostry, and H. Wolf, 1997, “Does the Nominal Exchange Rate Regime Matter?” NBER Working Paper No. 5874 (Cambridge, MA: National Bureau of Economic Research).
- Ghosh, A., A. Gulde, and H. Wolf, 2003, Exchange Rate Regime: Choices and Consequences, (Cambridge and London : MIT Press).
- Ghosh, A., J. Ostry, and C. Tsangarides, 2010, Exchange Rate Regimes and the Stability of the International Monetary System, IMF Occasional Paper 270 (Washington: IMF).
- Guisinger, A. G., and D.A. Singer, 2010, “Exchange Rate Proclamations and Inflation-Fighting Credibility,” International Organization, Vol. 64, pp. 313-337.
- Levy-Yeyati, E., and F. Sturzenegger, 2001, “Exchange Rate Regimes and Economic Performance,” IMF Staff Papers, Vol. 47, pp. 62-98.
- Levy-Yeyati, E., and F. Sturzenegger, 2003, “To Float or to Trail: Evidence on the Impact of Exchange Rate Regimes,” American Economic Review, Vol. 93, pp. 1173–93.
- Levy-Yeyati, E., and F. Sturzenegger, 2005, “Classifying Exchange Rate Regimes: Deeds versus Words,” European Economic Review, Vol. 49, pp. 1603–35.
- Reinhart, C. and K. Rogoff, 2004, “The Modern History of Exchange Rate Arrangements: A Reinterpretation,” Quarterly Journal of Economics, Vol. 119, pp. 1-48.
- Rogoff, K., A. Husain, A. Mody, R. Brooks, and N. Oomes, 2003, Evolution and Performance of Exchange Rate Regimes, IMF Working Paper No. 03/243 (Washington D.C.: IMF).
- Romer, D., 1993, “Openness and Inflation: Theory and Evidence,” Quarterly Journal of Economics, Vol. 108, pp. 869-903.
- Shambaugh, J., 2004, “The Effects of Fixed Exchange Rates on Monetary Policy,” Quarterly Journal of Economics, Vol. 119, No.1, pp. 301-352.
- Tornell, A., and A. Velasco, 2000, “Fixed versus Flexible Exchange Rates: Which Provides More Fiscal Discipline?” Journal of Monetary Economics, Vol. 45, pp. 399-436.
- Wolf, H., A. R. Ghosh, H. Berger, and A. Gulde, 2008, Currency Boards in Retrospect and Prospect (Cambridge, MA: MIT Press).

### Figures: captions and methodological notes
- Figure 1. Frequency Distribution of Exchange Rate Regimes, 1980–2010 (in percent)
  - Panels: (i) Emerging Market Economies; (ii) Developing Economies.
  - Source: Authors' estimates based on IMF's AREAR, and Anderson (2008).
  - Time points shown on axes: 1980, 1983, 1986, 1989, 1992, 1995, 1998, 2001, 2004, 2007, 2010.
  - Both De Jure Classification and De Facto Classification series are plotted.

- Figure 2. Money Growth and Inflation under Alternative De Jure and De Facto Regimes, 1980–2010* (in percent)
  - Source: Authors’ estimates.
  - Note: Figure shows the relationship between broad money growth and inflation (in percent per year) under alternative de jure and de facto regime classifications. The inflation rate has been transformed by: inflation/(1+inflation).

- Figure 3. Inflation and Money Growth under Pegs and Non-Pegs, 1980–2010*
  - Source: Authors’ estimates.
  - Notes: Based on the estimated sample where inflation is transformed as follows: inflation/(1+inflation).
  - Plots include De jure peg, De jure float, De facto peg, De facto float; and categories Pure peg, De facto peg, Soft peg, Free float with axes for Inflation (in percent) and Broad money growth (in percent).

- Figure 4. Dynamic Effect of De Jure and De Facto Pegs on Inflation*
  - Source: Authors’ estimates.
  - Note: *Figure shows estimated inflation dynamics following adoption of a de jure or de facto peg. Years into peg plotted: 1,3,5,7,9,11,13,15,17,19. Panels for All countries, EMEs, Developing countries. Legend entries include De jure, De facto, De jure consensus sample, De facto consensus sample.

### Tables: key numeric findings and reported coefficients (selected)
- Table 1. Classification of Exchange Rate Regimes, 1980–2010 (in percent of observations)
  - Source: Authors’ estimates based on IMF’s AREAER and Anderson (2008).

- Table 2. Distribution of De Jure and De Facto Classifications, 1980–2010 (in percent of observations)
  - De Jure Classification / De Facto classification — counts and percentages reported:
    - Pegged: 1,430 (De Jure) / 1,999 (De Facto) — Percentage consensus 71.5
    - Intermediate: 761 / 1,699 — Percentage consensus 74.2
    - Floating: 201 / 346 — Percentage consensus 73.3
    - Total observations: 1,526 (De Jure), 1,847 (De Facto), 786, 1,386 (other columns shown).

- Table 3. Inflation and Money Growth under Alternative Regimes, 1980–2010* (in percent per year)
  - De Jure classification (Full Sample 1980-2010):
    - Pegged regimes: 36.84
      - Hard pegs: 17.31
      - Conventional pegs: 19.52
    - Intermediate regimes: 44.44
      - Basket pegs: 11.92
      - Pegged within bands: 6.5
      - Crawiling pegs/band: 2.5
      - Managed floats: 23.5
    - Floating regimes: 18.96
      - Independent floats: 18.96
  - De Facto classification (Full Sample 1980-2010):
    - Pegged regimes: 44.43
      - Hard pegs: 18.5
      - Conventional pegs: 26.0
    - Intermediate regimes: 49.24
      - Basket pegs: 22.3
      - Pegged within bands: 10.0
      - Crawiling pegs/band: 2.7
      - Managed floats: 14.2
    - Floating regimes: 6.43
      - Independent floats: 6.43
  - Inflation and Broad money growth by group (selected entries):
    - De Jure Inflation (All countries, Peg): 5.4
    - De Jure Inflation (All countries, Intermediate): 10.3
    - De Jure Inflation (All countries, Float): 10.5
    - De Facto Inflation (All countries, Peg): 5.8
    - De Facto Broad money growth (All countries, Peg): 12.7
    - Emerging markets De Jure Inflation (Peg): 4.7; Broad money growth (Peg): 13.0
    - Developing countries De Jure Inflation (Peg): 5.5; Broad money growth (Peg): 12.6

- Table 4. Inflation Performance: Benchmark Specification (instrumental variable estimates)
  - All countries (De Jure classification): reported coefficients (coef. and t-stat):
    - Constant 0.041 t-stat 2.93 ***
    - Pegged regimes -0.029 t-stat -4.83 ***
    - Intermediate regimes -0.002 t-stat -0.50
    - Money growth 0.391 t-stat 6.41 ***
    - GDP growth -0.459 t-stat -3.86 ***
    - Openness -0.026 t-stat -6.50 ***
    - Central bank turnover rate 0.040 t-stat 4.44 ***
    - Fiscal balance (in pct. of GDP) -0.108 t-stat -2.20 **
    - Number of observations, R2: 2,024   0.432
  - All countries (De Facto classification): selected entries
    - Constant 0.025 t-stat 1.79 *
    - Pegged regimes -0.063 t-stat -11.76 ***
    - Intermediate regimes -0.008 t-stat -1.68 *
    - Money growth 0.404 t-stat 5.69 ***
    - GDP growth -0.440 t-stat -3.28 ***
    - Number of observations, R2: 1,881    0.419
  - Emerging market countries (De Jure):
    - Constant 0.037 t-stat 1.54
    - Pegged regimes -0.042 t-stat -3.23 ***
    - Number of observations, R2: 790   0.524
  - Developing countries (De Jure):
    - Constant 0.041 t-stat 2.93 ***
    - Pegged regimes -0.026 t-stat -3.71 ***
    - Number of observations, R2: 1,234   0.373
  - Note: “For example, the coefficient of -0.063 for pegged regimes implies 6.3 percent per year lower inflation under pegged exchange rate regimes compared to floats.” *, **, and *** indicate significance at the 10, 5 and 1 percent levels, respectively.

- Table 6. Inflation Performance: Consensus between De Jure and De Facto Classifications (selected)
  - All countries (De Jure):
    - Constant 0.033 t-stat 2.20 **
    - Pegged regimes -0.028 t-stat -4.67 ***
    - Intermediate regimes 0.004 t-stat 0.80
    - Number of observations, R2: 1,682   0.441
  - All countries (De Facto):
    - Constant 0.012 t-stat 0.71
    - Pegged regimes -0.064 t-stat -11.28 ***
    - Intermediate regimes 0.010 t-stat 1.20
    - Number of observations, R2: 1,629   0.429
  - Pure peg / De facto peg / Soft peg (combined specifications):
    - Constant 0.041 t-stat 3.15 ***
    - Pure peg -0.033 t-stat -6.60 ***
    - De facto peg -0.012 t-stat -2.40 **
    - Soft peg 0.011 t-stat 1.22
    - Money growth 0.388 t-stat 6.36 ***
    - Fiscal balance (in pct. of GDP) -0.139 t-stat -2.84 ***
    - Number of observations, R2: 2,024 0.438; subgroup R2 values reported for EMEs (0.527) and Developing countries (0.383).

- Table 7. Inflation Performance: Regime Duration (selected)
  - All countries (De Jure):
    - Constant 0.043 t-stat 3.07 ***
    - Pegged regimes -0.035 t-stat -5.83 ***
    - Intermediate regimes -0.008 t-stat -1.60
    - Pegged regimes duration 0.056 t-stat 2.67 ***
    - Intermediate regimes duration 0.044 t-stat 2.00 **
    - Money growth 0.387 t-stat 6.34 ***
    - Number of observations, R2: 2,024 0.435
  - Interpretative note shown in table: “The overall effect of pegs on inflation is the sum of the main coefficient for pegs and the duration of pegs. For example, the coefficient of -0.074 for pegged regimes and 0.107 for the duration of pegged regimes implies that for peg duration equal to 5 years, inflation is 5.2 percent lower under pegged exchange rate regimes compared to floats.” *, **, and *** indicate significance at the 10, 5 and 1 percent levels.

- Table 8. Inflation Performance under Balance of Payments Surpluses (selected)
  - All countries [1] Capital inflows exceeding 2.5 percent of GDP:
    - Constant -0.003 t-stat -0.33
    - Pegged regimes -0.018 t-stat -3.60 ***
    - Intermediate regimes 0.009 t-stat 1.80 *
    - Number of observations, R2: 1,117 0.432
  - All countries [2] Current account balance exceeding 2 percent of GDP:
    - Constant 0.059 t-stat 3.47 ***
    - Pegged regimes -0.012 t-stat -1.33
    - Intermediate regimes -0.013 t-stat -1.63
    - Number of observations, R2: 374 0.543
  - Note: “3/ Corresponds to the 30th percentile of positive net capital flows and positive current account balances, respectively.”

- Table 9. Inflation Performance: Regime Endogeneity, Freely Falling, and Hard Pegs (selected)
  - All countries (controlling for regime endogeneity):
    - Constant 0.056 t-stat -4.51 ***
    - Pegged regimes -0.040 t-stat 3.77 ***
    - Money growth 0.389 t-stat -27.59 ***
    - GDP growth -0.517 t-stat 5.12 ***
    - Fiscal balance (in pct. of GDP) -0.138 t-stat 2.86 ***
    - Number of observations: 1,908
  - Freely Falling: Collapsing Exchange Rates (excludes floating regimes with depreciation >25% per year) and Freely Falling: High Inflation (excludes floating regimes with inflation over 40% per year) specifications are reported with corresponding coefficients and sample sizes in the table.
  - Hard Pegs: specifications exclude hard pegs (currency boards and monetary unions) from the pegs category in alternative regressions; coefficients and sample sizes are reported in table.

### Methodological and notation highlights (as presented)
- Regression dependent variable: inflation (decimal fraction, per year).
- Instrumental variable estimation used to control for endogeneity of real GDP growth, fiscal balance, and money growth where lagged values are used as instruments.
- All specifications include time effects. t-statistics are based on clustered standard errors.
- Negative coefficient on pegged or intermediate exchange rate regime dummies indicates lower inflation under that regime relative to inflation under floating exchange rate regimes (the omitted category).
- Significance notation: *, **, and *** indicate significance at the 10, 5 and 1 percent levels, respectively.
- Inflation transformation used in several figures/tables: inflation/(1+inflation).
- Direct Effect vs. Total Effect 2/: tables report both direct effect of exchange rate regime on inflation and total effect including indirect effect through money growth (labeled “Total Effect 2/” or similar).

*Source: _wp11112 - REFERENCES*

### APPENDIX A. DATA DESCRIPTION

### APPENDIX A. DATA DESCRIPTION

### Comparison of exchange rate regime classifications
- The IMF’s de facto classification is described as "less idiosyncratic" than other commonly used de facto classifications.
- A composite similarity index (value between 0 and 1, where 0 = perfect dissimilarity and 1 = perfect similarity) is constructed by:
  - Assigning each classification a value of 1 if it agrees with any other classification.
  - For every classification, a country-year observation receives a score of 1/3 for each other classification that agrees with it.
  - The overall index is the weighted sum of the scores for the three regimes, with weights equal to the proportion of pegs, intermediate, and floats in the particular classification.
- Reported average similarity scores:
  - IMF classification: about 0.75 if the de jure classification is included in the comparator category.
  - IMF classification: about 0.74 if the de jure classification is not included.
  - LYS: 0.67.
  - RR: 0.58.
- Interpretation: A low score indicates greater disagreement with other classification methods; disagreement does not necessarily imply incorrectness but suggests empirical results may be less robust across classifications.
- Sample period and coverage for comparisons:
  - Sample period: 1980-2004.
  - Country coverage: expanded to include all countries for which data is available (advanced, EMEs, and developing countries).

### Agreement with Shambaugh (JS) and pairwise comparison metrics
- JS classification is binary (pegs versus nonpegs) and therefore compared using binary grouping of other classifications and correlation matrices.
- Key summary statements:
  - About 87 percent of the observations in the IMF de facto classification are coded (as pegs or nonpegs) in the same way as in the JS classification.
  - The overall correlation between the IMF de facto series and the JS series is 0.76.
  - The IMF’s de facto classification is found to be the closest to the JS classification and the least similar to LYS.
- Percentage matching matrix (as presented):
  - IMF's De jure — 100.0
  - IMF's De facto — 84.5 100.0
  - JS — 80.0 87.8 100.0
  - LYS — 69.8 59.6 62.9 100.0
  - RR — 77.6 77.8 72.0 54.3 100.0
- Correlation matrix (as presented):
  - IMF's De jure — 1.0
  - IMF's De facto — 0.7 1.0
  - JS — 0.6 0.8 1.0
  - LYS — 0.5 0.6 0.6 1.0
  - RR — 0.5 0.7 0.6 0.5 1.0

### Figures and tables referenced
- Figure A1: Similarity Index across different exchange rate regime classifications (sample period 1980-2004).
- Table A1: Agreement of Different Exchange Rate Regime Classifications, 1980–2004 (percentage matching and correlation matrices shown above).
- Table A2: List of Countries in the Sample (country list provided in the source).
- Table A3: Data Description and Sources (variables, observations, descriptive statistics, and sources summarized below).

### Sample and key variables (selected descriptive statistics and sources)
- Broad money growth
  - Obs.: 1,937
  - Mean: 0.186
  - Std. Dev: 0.224
  - Source: IMF's WEO database
- Central bank turnover rate
  - Description: Central Bank governor turnover rate per 5 years
  - Obs.: 1,937
  - Mean: 0.195
  - Std. Dev: 0.200
  - Source: Ghosh et al. (2007)
- Fiscal balance to GDP
  - Description: Overall government balance to GDP
  - Obs.: 1,932
  - Mean: -0.029
  - Std. Dev: 0.051
  - Source: IMF's WEO database
- Geographic export concentration
  - Description: Share of exports to top three destinations in total exports
  - Obs.: 1,824
  - Mean: 0.574
  - Std. Dev: 0.169
  - Source: IMF's DOTS
- Inflation
  - Description: Growth rate of Consumer Price Index. Transformed to inflation/(1+inflation)
  - Obs.: 1,937
  - Mean: 0.085
  - Std. Dev: 0.095
  - Source: IMF's WEO database
- Population size
  - Description: In millions
  - Obs.: 1,937
  - Mean: 48.423
  - Std. Dev: 170.150
  - Source: IMF's WEO database
- Real GDP per capita growth
  - Description: Growth rate of GDP (in constant prices) per capita
  - Obs.: 1,937
  - Mean: 0.020
  - Std. Dev: 0.046
  - Source: IMF's WEO database
- Terms of trade change
  - Description: Growth rate of terms of trade index
  - Source: IMF's WEO database
  - (Obs., mean, and std. dev. not shown in the excerpt)
- Trade openness
  - Description: Sum of exports and imports to GDP
  - Obs.: 1,937
  - Mean: 0.768
  - Std. Dev: 0.407
  - Source: IMF's WEO database

### IMF exchange rate regime classification variables (descriptive stats)
- Coarse classification (1=Peg; 2=Intermediate; 3=Float) — Source: Anderson (2008)
  - Obs.: 1,937 (for coarse classification)
- De jure peg
  - Description: Includes currency unions, currency boards, conventional pegs (excl. basket pegs)
  - Obs.: 1,937
  - Mean: 0.366
  - Std. Dev: 0.481
  - Source: Anderson (2008)
- De jure intermediate
  - Description: Includes pegs within bands, crawling pegs, basket pegs, crawling band, and managed floats
  - Obs.: 1,937
  - Mean: 0.460
  - Std. Dev: 0.499
  - Source: Anderson (2008)
- De jure float
  - Description: Includes independently floating regime
  - Obs.: 1,937
  - Mean: 0.174
  - Std. Dev: 0.379
  - Source: Anderson (2008)
- De facto peg
  - Description: Includes currency unions, currency boards, conventional pegs (excl. basket pegs)
  - Obs.: 1,937
  - Mean: 0.460
  - Std. Dev: 0.499
  - Source: Anderson (2008)
- De facto intermediate
  - Description: Includes pegs within bands, crawling pegs, basket pegs, crawling band, and managed floats
  - Obs.: 1,937
  - Mean: 0.443
  - Std. Dev: 0.497
  - Source: Anderson (2008)
- De facto float
  - Description: Includes independently floating regime
  - Obs.: 1,937
  - Mean: 0.010
  - Std. Dev: 0.296
  - Source: Anderson (2008)

*Source: APPENDIX A. DATA DESCRIPTION (from the provided IMF content).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11112.pdf_