## _wp0575

## Source details

**Canonical URL:** [_wp0575](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0575.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0575.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0575.pdf.json)

---

### Introduction and research scope
- Objective: assess links between legal central bank independence (CBI) and inflation for 24 Latin American and Caribbean countries during the 1990s and early 2000s (1990–2002).
- Theoretical motivation: time-inconsistency models (Kydland and Prescott, 1977; Barro and Gordon, 1983) and Rogoff (1985) suggest delegating monetary policy to an independent and conservative central bank can reduce inflationary bias.
- Empirical strategy: construct three de jure CBI measures (GMT 1991 index; Cukierman 1992 index; Modified Cukierman Index (MCI)), use panel regressions with country fixed effects and FGLS, control for Banking Crisis Dummy, Fixed Exchange Regime Dummy, International Inflation (IFS series 110), and Structural Reform Index; test for endogeneity using instrumental-variable methods.

### Measures of legal central bank independence
- GMT index (Grilli, Masciandaro, Tabellini, 1991): 15 binary criteria, summed.
- Cukierman index (1992): 16 criteria, continuous 0–1 scale, weighted; 50 percent weight to central bank lending to government, 20 percent to appointment/term/dismissal, 30 percent to policy formulation and objective.
- Modified Cukierman Index (MCI): retains Cukierman structure and adds accountability; evaluates appointment/dismissal rules for entire board, exchange-rate policy role, lender-of-last-resort (LOLR) rules, safeguards for central bank capital, and reporting/financial-statement transparency.
- Limitation emphasized: indexes are de jure measures based on constitutions and central bank laws and may not reflect de facto independence.

### Map of central bank reform in LAC during the 1990s (cluster analysis)
- Five legislative feature categories: (i) primary policy objective, (ii) political independence for design/execution of monetary policy, (iii) economic independence, (iv) financial autonomy, (v) accountability and transparency.
- Main patterns:
  - Reforms widespread in Latin America (LAC); almost absent in Caribbean (CAR).
  - Prominent reform features: mandate focus on price stability, formal restrictions on central bank lending to government, increased autonomy in interest rate policy.
  - Areas needing improvement: limits on unilateral appointment/removal by government, clearer restrictions/conditions for LOLR, improved financial independence and transparency.
- Cluster analysis results — group averages (exact figures preserved):
  - Modified Cukierman Index (average by group): Group 1 = 0.84 ; Group 2 = 0.70 ; Group 3 = 0.44
  - Political Independence: Group 1 = 0.91 ; Group 2 = 0.55 ; Group 3 = 0.27
  - Central Bank primary objective: Group 1 = 0.96 ; Group 2 = 0.69 ; Group 3 = 0.23
  - Central Bank lending: Group 1 = 0.78 ; Group 2 = 0.77 ; Group 3 = 0.45
  - LOLR: Group 1 = 0.50 ; Group 2 = 0.50 ; Group 3 = 0.50
  - Accountability: Group 1 = 0.93 ; Group 2 = 0.73 ; Group 3 = 0.54
  - Fiscal Deficit (percent of GDP): Group 1 = 2.60 ; Group 2 = 2.31 ; Group 3 = 1.08
  - Inflation (in percent): Group 1 = 9.83 ; Group 2 = 15.8 ; Group 3 = 35.65
  - Structural Reform Index: Group 1 = 0.60 ; Group 2 = 0.53 ; Group 3 = 0.49
  - Number of countries: Group 1 = 9/6 ; Group 2 = 9 ; Group 3 = 23
- Interpretation: Groups 1 and 2 show stronger legal independence and lower average inflation with higher fiscal deficits and more advanced structural reforms; Group 3 comprises pre- and non-reform countries (many CAR) with weaker legal CBI.

### Legal CBI and inflation: descriptive statistics and correlations
- Regional summary statistics (selected exact figures):
  - Caribbean countries, Inflation overall mean = 18.84 ; Std. Dev. = 50.17 ; Min = -1.30 ; Max = 368.50 ; N = 88.00
  - Latin American countries, Inflation overall mean = 19.43 ; Std. Dev. = 33.36 ; Min = -1.20 ; Max = 413.30 ; N = 198.00
  - GMT total overall: Caribbean = 7.23 ; Latin America = 10.20
  - Cukierman overall: Caribbean = 0.50 ; Latin America = 0.66
  - Modified Cukierman overall: Caribbean = 0.50 ; Latin America = 0.64
- Pair-wise correlations:
  - Inflation is negatively correlated with CBI indexes in both sub-samples.
  - High positive correlation between Structural Reform Index and CBI indexes, especially in LAC.
  - GMT, Cukierman, and Modified Cukierman indexes are highly correlated.
  - Selected index-to-index correlations:
    - Latin American countries: Cukierman, Modified Cukierman total index correlation = 0.896
    - Caribbean countries: Cukierman, Modified Cukierman total index correlation = 0.845

### Regression evidence and robustness checks
- Dependent variable: inflation measured as yearly change in CPI rescaled π/(1+π); alternative log(1+π) used with similar results.
- Estimation methods: OLS with country fixed effects and robust standard errors; FGLS allowing heteroscedasticity and common AR(1) error; IV methods (G2SLS, EC2SLS).
- Baseline FGLS results (whole sample excluding hyperinflation cases; exact coefficients reproduced):
  - Dummy Banking Crises coefficients: 0.023 , 0.023 , 0.026* , 0.036** , 0.037** , 0.038**
  - Dummy Exch. Regime coefficients: -0.082*** , -0.080*** , -0.081*** , -0.079*** , -0.079*** , -0.076***
  - World Inflation (in percent) coefficients: 0.035*** , 0.035*** , 0.033***
  - GMT Total coefficient where included: -0.009*** (column [4]) and -0.005** (column [5])
  - Cukierman Weighted coefficient where included: -0.129*** and -0.066**
  - Total Modified Cukierman coefficient where included: -0.137*** and -0.079***
  - Observations = 297 ; Number of countries = 24 ; reported Rho examples: 0.74, 0.72, 0.67
- Interpretation of baseline: higher legal CBI (any index) is associated with lower inflation after controlling for banking crises and exchange rate rigidity; fixed exchange regimes associated with lower inflation; results robust to adding world inflation.
- Role of Structural Reforms (contemporaneous and lagged specifications; exact coefficients preserved):
  - Index Structural Reform coefficient (contemporaneous): -0.421*** , -0.428*** , -0.412***
  - Index Structural Reform coefficient (lagged-CBI specifications): -0.399*** , -0.416*** , -0.408***
  - Including Structural Reform renders contemporaneous CBI coefficients statistically non-significant in columns [1]–[3]; lagged CBI specifications (three- and four-lag) restore CBI significance in columns [4]–[6].
  - Observations fall to 198 when Structural Reform Index is included (data coverage limits: missing years 2001–2002 and some CAR countries).
- Interpretation: structural reforms and increased legal CBI appear complementary; CBI effects may materialize with lags (reputation, regulatory implementation).

### Accounting for endogeneity: instrumental-variable results
- IV methods: G2SLS (Balestra and Varadharajan-Krishnakumar, 1987) and EC2SLS (Balgati and Chang, 2000); instruments include lagged values of MCI and, in some specs, lagged Structural Reform.
- Selected IV regression coefficients (Table 8; exact values preserved):
  - Dummy Banking Crises: 0.117*** , 0.102*** , 0.112*** , 0.097*** , 0.124*** , 0.108***
  - Dummy Exch. Regime: -0.056** , -0.073*** , -0.054** , -0.070*** , -0.060*** , -0.074***
  - World Inflation (in percent): 0.030** , 0.037*** , 0.028** , 0.032*** , 0.029** , 0.033***
  - Index Structural Reform: -0.444*** , -0.547*** , -0.483*** , -0.643*** , -0.524*** , -0.643***
  - Total Modified Cukierman: -0.100 , 0.070 , -0.109 , 0.068 , -0.073 , 0.072 (none statistically significant)
  - Observations = 149 ; Number of countries = 19 ; R-Sq. within/between/overall examples: 0.25/0.49/0.36
  - Hausman tests: p-values reported (examples: 0.433, 0.509, 0.286) — failed to reject null of coefficient equality between random- and fixed-effects models.
- Interpretation:
  - IV evidence does not support a statistically significant causal effect of legal CBI (MCI) on inflation.
  - Structural Reform Index remains robustly and negatively associated with inflation.
  - Negative contemporaneous CBI–inflation correlations appear biased downward when endogeneity is not addressed.
  - Overall conclusion: the dramatic disinflation in the region during the 1990s likely reflects a broad set of policies and reforms rather than legal central bank reform alone.
- Remaining caveat: legal CBI indexes may be poor proxies for effective (de facto) central bank independence.

### Concluding summary (three main conclusions)
- Regional differences: CBI reform widely implemented in LAC but largely absent in CAR; LAC central banks generally enjoy higher formal operational and political independence while CAR central banks remain politically dependent on executive branches.
- Observed negative correlation: legal CBI negatively correlated with inflation after controlling for international inflation, exchange regime, and banking crises; Structural Reform Index demonstrates complementarity of reforms.
- Endogeneity-adjusted inference: after accounting for endogeneity, evidence does not support a causal relationship from legal CBI to inflation; disinflation in the 1990s owed to a combination of policies, cautioning against attributing the outcome solely to legal CBI reform.
- Persistent limitation: indexes of legal CBI are imperfect proxies for effective central bank independence.

### Appendix excerpt: GMT index scoring components (selected items and weights)
- CEO term of office weight: (0.25) — scoring row: 1 0.75 0.50 0.25 0
- Who appoints the CEO weight: (0.25) — scoring row: 1 0.75 0.50 0.25 0
- Provisions for dismissal of CEO weight: (0.25) — scoring row: 1 0.83 0.67 0.50 0.33 0.17 0
- CEO allowed to hold another office weight: (0.25) — scoring row: 1 0.5 0
- Central Bank Board weights and scoring examples:
  - Term of office of Governor (0.20): score row 1 0.67 0.33 0
  - Who appoints the Governor (0.20): score row 1 0
  - Appointment and term of office rest of the Board (0.20): score row 1 0.75 0.50 0.25 0
  - Dismissal of Board members (0.30): score row 1 0.75 0.50 0.25 0
  - CEO allowed to hold another office (0.10): score row 1 0.5 0
- Central Bank objectives (0.15) and Fundamental objective (1.00) scoring rows preserved as presented.
- Policy formulation components (examples of weights and score rows preserved).
- Central Bank lending section (various breakdowns preserving weights and score rows; examples include Limits on advances score row 1 0.67 0.33 0; LOLR (0.15) score row 1 0.75 0.50).
- Financial autonomy (0.10) score row examples: 1 0.5
- Accountability (0.10) and transparency (0.25)—disclosure criteria listed verbatim; associated numeric entries as presented include: 0.25, 0, 1, 0.67, 0.33, 0, 1, 0.75, 0.50, 0.25, 0, 1, 0.75, 0.50, 0.25, 0.

*Source: _wp0575 - “2. International Inflation and Inflation in Sampled Countries, 1990–2002” and appended GMT index material from the supplied IMF working paper content.*

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

### _wp0575 - References

### Tables
- 1. List of Country Groups Based on Cluster Analysis.............................................................12
- 2. Summary Statistics Across Country Groups .......................................................................13
- 3. Summary Statistics by Regions, 1990–2002 .......................................................................16
- 4. Pair-Wise Correlations Between Selected Variables, by Regions, 1990–2002...................17
- 5. Correlations Between Indexes of Legal Central Bank Independence, by Regions .............17
- 6. Panel Regressions of Inflation on Contemporaneous Central Bank Independence.............20
- 7. Panel Regressions of Inflation on Contemporaneous Central Bank Independence 
  (Including Structural Reforms) ........................................................................................21
- 8. Panel Regressions of Inflation on MCI Using Instrumental Variables................................23

### Figures
- 1. Cluster Analysis Based on the Components of the Modified Cukierman ...........................11

*Source: _wp0575 - References (PDF).*

### 2. International Inflation and Inflation in Sampled Countries, 1990–2002 .............................15

### 2. International Inflation and Inflation in Sampled Countries, 1990–2002

### Introduction
- The 1990s saw deep changes to central bank legislation in many countries, increasing autonomy and accountability with the objective of attaining price stability.
- Reforms were widespread in Latin American countries (LAC) during the 1990s; changes were largely absent among Caribbean countries (CAR).
- Theoretical motivation:
  - Time-inconsistency models (Kydland and Prescott, 1977; Barro and Gordon, 1983) and Rogoff (1985) imply that delegating monetary policy to an independent and conservative central bank can reduce inflationary bias.
- Empirical background:
  - Industrial-country evidence often shows a negative association between central bank independence (CBI) and inflation (Berger et al., 2000), though causality is contested and measurement/error issues exist.
  - Evidence for developing countries is scarcer and less conclusive; prior studies (Cukierman et al., 1992; De Haan and Kooi, 2000) did not find robust CBI–inflation links for developing-country samples without addressing omitted variables, time dimension, and endogeneity.
- This paper:
  - Extends Cukierman-style indexes to a regional sample of developing countries in the 1990s, tracking CBI pre- and post-reform.
  - Builds three alternative de jure CBI measures: GMT (1991) index, the Cukierman (1992) index, and a Modified Cukierman Index (MCI) adding appointment rules for boards, exchange-rate policy CBI, lender-of-last-resort (LOLR) rules, and accountability/transparency requirements.
  - Uses panel regressions, controls for international inflation, banking crises, exchange regimes, and structural reforms, and tests for endogeneity of central bank reform.
- Key high-level result summary:
  - A strong negative correlation is found between increased legal CBI and inflation after controlling for international inflation, banking crises, and exchange regimes; robust across three CBI measures.
  - Structural reforms also show beneficial effects on inflation, indicating complementarity among reform dimensions.
  - After accounting for endogeneity, a causal relationship from legal CBI to inflation is not established; implying the 1990s disinflation in LAC and CAR cannot be straightforwardly attributed to increased legal CBI.
  - Caveat: legal CBI measures may be poor indicators of effective CBI.

### Measures of Legal Central Bank Independence
- Sample and approach:
  - Quantification applies to 24 Latin American and Caribbean countries—before and after reform where relevant.
  - Assessment based exclusively on legal provisions (constitutions and central bank laws); de jure may differ from de facto independence.
- Three indexes described:
  1. GMT index (Grilli, Masciandaro, and Tabellini, 1991):
     - 15 criteria, binary scoring (0 or 1), overall index by simple addition; measures political and economic independence.
  2. Cukierman index (1992):
     - 16 criteria, continuous scale 0–1, weighted average; highest weight (50 percent) given to legal provisions on central bank lending to the government; 20 percent to appointment/term/dismissal; remaining 30 percent split between policy formulation and central bank fundamental objective.
  3. Modified Cukierman Index (MCI):
     - Builds on Cukierman structure, keeps four general classification criteria (appointment/dismissal, independence for policy formulation, objectives, central bank lending) and adds a new category: accountability.
     - Innovations and additions:
       - Political autonomy: evaluates appointment/dismissal rules for entire central bank board (not only governor); considers overlapping of board/term with presidential term; extra weight on dismissal protections for board members.
       - Policy formulation: includes assessment of CBI in exchange rate policy formulation; shifts assessment of role in government budget to approval of public sector debt.
       - Economic autonomy: simplifies evaluation of central bank lending provisions; adds criteria for LOLR faculties and safeguards for central bank financial autonomy (government required to preserve integrity of central bank capital); rewards limits on central bank involvement in banking crises.
       - Accountability: rewards legal requirements for regular reporting on policy targets/achievements and mandatory publication of financial statements, especially when following international accounting standards and certified by independent auditor.
- Limitations noted:
  - Legal provisions may not reflect de facto independence; discrepancies often arise in political independence, and also in operational independence and transparency.

### A Map of Central Bank Reform in LAC during the 1990s
- Five categories used to organize central bank legislation features:
  - (i) definition of primary policy objective, (ii) political independence in design/execution of monetary policy, (iii) economic independence, (iv) financial autonomy, (v) accountability and transparency.
- Broad findings:
  - Central bank reform was widespread in LAC but almost absent in CAR; post-reform LAC central banks generally enjoy higher formal operational and political independence.
  - Prominent reform features: more focused mandate on price stability and formalized restrictions on central bank lending to the government; increased autonomy in interest rate policy.
  - Areas needing improvement: reduce government capacity to unilaterally appoint/remove governors/directors; limit excessive discretion in LOLR function (size, conditions, triggers); improve financial independence and transparency (e.g., disclosure of financial statements).
- Cluster analysis (based on MCI, Euclidean distance, hierarchical single linkage) results:
  - Three groups emerged (see Table 1 listing country-periods by group; reform countries treated as pre-reform “-0” and post-reform “-1” cases).
  - Summary averages by group (Table 2) — select exact figures preserved:
    - Modified Cukierman Index (average by group): Group 1 = 0.84 ; Group 2 = 0.70 ; Group 3 = 0.44
    - Political Independence: Group 1 = 0.91 ; Group 2 = 0.55 ; Group 3 = 0.27
    - Central Bank primary objective: Group 1 = 0.96 ; Group 2 = 0.69 ; Group 3 = 0.23
    - Central Bank lending: Group 1 = 0.78 ; Group 2 = 0.77 ; Group 3 = 0.45
    - LOLR: Group 1 = 0.50 ; Group 2 = 0.50 ; Group 3 = 0.50
    - Accountability: Group 1 = 0.93 ; Group 2 = 0.73 ; Group 3 = 0.54
    - Fiscal Deficit (percent of GDP): Group 1 = 2.60 ; Group 2 = 2.31 ; Group 3 = 1.08
    - Inflation (in percent): Group 1 = 9.83 ; Group 2 = 15.8 ; Group 3 = 35.65
    - Structural Reform Index: Group 1 = 0.60 ; Group 2 = 0.53 ; Group 3 = 0.49
    - Number of countries: Group 1 = 9/6 ; Group 2 = 9 ; Group 3 = 23
  - Interpretation:
    - Groups 1 and 2 display stronger legal independence along most dimensions and lower average inflation combined with higher fiscal deficits and more advanced structural reforms.
    - Group 3 includes pre- and non-reform countries, many from CAR; displays weaker legal CBI.

### Legal Central Bank Independence and Inflation in LAC and CAR: Is There a Link?
- Contextual point:
  - Regional disinflation: decline from an average of about 500 percent at the beginning of the decade to 7 percent in 2002 (excluding hyperinflation episodes as filtered in the paper).
  - Disinflation coincided with global trend, but magnitude suggests country-specific policies/institutional reforms were main drivers.
- Data overview (Tables 3–5 highlights; exact values preserved where cited):
  - Summary statistics by region (Table 3): examples of exact figures:
    - Caribbean countries, Inflation overall mean = 18.84 ; Std. Dev. = 50.17 ; Min = -1.30 ; Max = 368.50 ; N = 88.00
    - Latin American countries, Inflation overall mean = 19.43 ; Std. Dev. = 33.36 ; Min = -1.20 ; Max = 413.30 ; N = 198.00
    - GMT total overall for Caribbean: 7.23 ; for Latin America: 10.20
    - Cukierman overall for Caribbean: 0.50 ; for Latin America: 0.66
    - Modified Cukierman overall for Caribbean: 0.50 ; for Latin America: 0.64
  - Pair-wise correlations (Table 4):
    - Inflation is negatively correlated with CBI indexes in both sub-samples (specific correlation coefficients and significance levels presented in the paper).
    - High positive correlation between Structural Reform Index and CBI indexes, especially in LAC.
    - GMT, Cukierman, and Modified Cukierman indexes are highly correlated.
  - Correlations between indexes and subcomponents (Table 5) — selected exact coefficients:
    - Latin American countries: Cukierman, Modified Cukierman total index correlation = 0.896
    - Caribbean countries: Cukierman, Modified Cukierman total index correlation = 0.845
- Estimation strategy and dependent variable:
  - Inflation measured as yearly change in average consumer price index (IFS series 64), rescaled using π/(1+π); alternative log(1+π) used with similar results.
  - Regressions use OLS with country fixed effects and robust standard errors; later Feasible Generalized Least Squares (FGLS) allowing heteroscedasticity across countries and a common AR(1) error process.
  - Controls include Banking Crisis Dummy, Fixed Exchange Regime Dummy (constructed from IMF de facto exchange regimes), and International Inflation (average inflation of industrial countries, IFS series 110). Fiscal Deficit (IFS series 80 scaled by GDP series 99b) also tested.
  - Filter excludes eight cases with yearly inflation above 1,000 percent per year (exactly listed in the text).
- A. Baseline Results (Table 6; coefficients reproduced exactly where shown)
  - FGLS results for whole sample (excluding hyperinflation cases), dependent variable π/(1+π):
    - Dummy Banking Crises coefficients across models: 0.023 , 0.023 , 0.026* , 0.036** , 0.037** , 0.038** (standard errors reported in table).
    - Dummy Exch. Regime coefficients: -0.082*** , -0.080*** , -0.081*** , -0.079*** , -0.079*** , -0.076***.
    - World Inflation (in percent) coefficients (in columns including it): 0.035*** , 0.035*** , 0.033***.
    - GMT Total coefficient where included: -0.009*** (column [4]) and -0.005** (column [5]).
    - Cukierman Weighted coefficient where included: -0.129*** and -0.066** (respective columns).
    - Total Modified Cukierman coefficient where included: -0.137*** and -0.079***.
    - Observations = 297 ; Number of countries = 24 ; Rho reported per column (e.g., 0.74, 0.72, 0.67).
  - Interpretation:
    - After controlling for banking crises and exchange rate rigidity, higher legal CBI (any index) is related to lower inflation.
    - Fixed exchange regimes associated with lower inflation.
    - Results robust to adding world inflation.
    - Fiscal Deficit added as control entered not significantly and specification was rejected in favor of alternative controls.
- B. Role of Structural Reforms (Table 7; coefficients preserved)
  - Adding Index Structural Reform (Inter-American Development Bank index, ranges in [0, 1]):
    - Index Structural Reform coefficient: -0.421*** , -0.428*** , -0.412*** in contemporaneous specifications; -0.399*** , -0.416*** , -0.408*** in lagged-CBI specifications.
    - Inclusion of structural reforms renders contemporaneous CBI coefficients statistically non-significant (columns [1]–[3]).
    - Exploring lagged effects: using three- and four-lag specifications for CBI restores statistical significance of CBI coefficients (columns [4]–[6]); suggests CBI effects may take time to materialize (reputation, regulation, implementation).
    - Observations fall after including Structural Reform Index (198 observations vs. 297 previously) due to data coverage limits (missing years 2001–2002 and some CAR countries).
  - Interpretation:
    - Structural reforms and increased CBI appear complementary; both played roles in disinflation.
    - Contemporaneous estimation with structural reforms suggests omitted variable bias could explain earlier negative contemporaneous CBI–inflation relationship.
- C. Accounting for Endogeneity (Table 8; instrumental-variable results preserved)
  - Two IV methods used:
    - Generalized Two-Stage Least Squares (G2SLS) per Balestra and Varadharajan-Krishnakumar (1987).
    - Error Correction Two-Stage Least Squares (EC2SLS) per Balgati and Chang (2000).
  - Instruments:
    - Lagged values of the MCI used as instruments for contemporaneous MCI; in some specifications, Structural Reform also instrumented by its lagged values.
    - First-stage results (not reported) indicated instruments appropriate (coefficients significant at one-percent level).
  - IV regression results (selected coefficients from Table 8):
    - Dummy Banking Crises: 0.117*** , 0.102*** , 0.112*** , 0.097*** , 0.124*** , 0.108***.
    - Dummy Exch. Regime: -0.056** , -0.073*** , -0.054** , -0.070*** , -0.060*** , -0.074***.
    - World Inflation (in percent): 0.030** , 0.037*** , 0.028** , 0.032*** , 0.029** , 0.033***.
    - Index Structural Reform: -0.444*** , -0.547*** , -0.483*** , -0.643*** , -0.524*** , -0.643***.
    - Total Modified Cukierman: -0.100 , 0.070 , -0.109 , 0.068 , -0.073 , 0.072 (none statistically significant in these IV specifications; standard errors reported in table).
    - Observations = 149 ; Number of countries = 19 ; R-Sq. within/between/overall reported per column (e.g., 0.25/0.49/0.36).
    - Hausman specification tests (p-values reported) failed to reject null of coefficient equality between random- and fixed-effects models in all cases (p-values: 0.433, 0.509, 0.286 reported).
  - Interpretation:
    - IV results fail to show a statistically significant causal relation from legal CBI to inflation.
    - Structural reforms remain robustly associated with lower inflation.
    - The negative contemporaneous correlations of CBI with inflation appear biased downward due to potential endogeneity of central bank reform.
    - At face value, the extraordinary reduction in inflation during the 1990s in the region is likely attributable to a broad range of economic policies rather than legal central bank reform alone.
  - Remaining caveat: validity of legal CBI indexes as measures of effective CBI remains a crucial limitation.

### Concluding Remarks
- Three main conclusions based on legal assessments of central bank charters and constitutions for 24 countries in the 1990s:
  1. Significant regional differences: CBI reform widely implemented in LAC but largely absent in CAR; LAC central banks generally enjoy higher formal operational and political independence while CAR central banks remain politically dependent on executive branches.
  2. Negative correlation between legal CBI and inflation is observed after controlling for international inflation, exchange regime, and banking crises; the inclusion of Structural Reform Index highlights complementarity of reforms.
  3. After accounting for endogeneity, evidence does not support a causal relationship from legal CBI to inflation; suggests disinflation in the 1990s owed to a combination of policies, warning against excessive optimism about legal CBI’s direct inflation effects.
- Persistent caveat:
  - Indexes of legal CBI are imperfect proxies for effective central bank independence; results should be interpreted with this limitation in mind.

*Source: “2. International Inflation and Inflation in Sampled Countries, 1990–2002” (chapter/section) from the supplied IMF working paper content.*

### 1. Term of office of CEO (0.25)

### _wp0575 - 1. Term of office of CEO (0.25)

### 1. Term of office of CEO (0.25)
- Equal or more than 8 years
- 6  years or more but less than 8 years
- Equal to 5 years
- Equal to 4 years
- Less than 4 years
- Score row: 1 0.75 0.50 0.25 0

### 2. Who appoints the CEO (0.25)
- The Central Bank Board
- Council composed by executive and legislative branch and Central Bank Board
- By legislative branch
- By executive branch
- By one or two members of executive branch
- Score row: 1 0.75 0.50 0.25 0

### 3. Provisions for dismissal of CEO (0.25)
- No provision
- Only for non-policy reasons (e.g., incapability, or violation of law)
- At a discretion of Central Bank Board
- For policy reasons at legislative branch’s discretion
- At legislative branch’s discretion
- For policy reasons at executive branch’s discretion
- At executive branch’s discretion
- Score row: 1 0.83 0.67 0.50 0.33 0.17 0

### 4. CEO allowed to hold another office in government (0.25)
- Prohibited by law
- Not allowed unless authorized by executive branch
- No prohibition for holding another office
- Score row: 1 0.5 0

---

### Central Bank Board (0.20)

### 1. Term of office of Governor(0.20)
- More than presidential period
- The period does not coincide
- Same period as the executive branch
- Less than executive branch or not specified in the law
- Score row: 1 0.67 0.33 0

### 2. Who appoints the Governor (0.20)
- Double process (Executive/Legislative), or through the Central Bank Board if also appointed in a double process, or for longer or overlapped periods with respect to the executive branch
- The executive branch directly or through the Central Bank Board, when this is directly appointed by the executive branch
- Score row: 1 0

### 3. Appointment and term of office rest of the Board (0.20)
- More than presidential period or for a non-defined period
- For the same period as the President of the Republic with overlap
- Double process for the same period
- Executive and private sector appoint the majority of directors for same period or less
- Executive branch appoints the majority for the same period or less
- Score row: 1 0.75 0.50 0.25 0

### 4. Dismissal of Board members (0.30)
- Double process approved by the Senate or by a qualified majority and for violations codified in legislation
- By an independent Central Bank Board
- Double process with simple majority, based on policy decisions or due to subjective reasons
- By executive branch or subordinated Central Bank Board due to legal reasons
- By executive branch or subordinated Central Bank Board due to policy or subjective reasons, or no legal provision
- Score row: 1 0.75 0.50 0.25 0

### 5. CEO allowed to hold another office in government (0.10)
- Prohibited by law
- Not allowed unless authorized by executive branch
- No prohibition for holding another office
- Score row: 1 0.5 0

---

### Central Bank objectives (0.15)

### 5. Central Bank objectives
- Price stability is the only or major goal, and in case of conflict with government, the Central Bank has final authority
- Price stability is the only goal
- Price stability along with other objectives that do not seem to conflict with the former
- Price stability along with other objectives of potentially conflicting goals (e.g., full employment)
- Central Bank charter does not contain any objective
- Some goals appear in the charter but price stability is not one of them
- Score row: 1 0.8 0.6 0.4 0.2 0

### 6. Fundamental objective (1.00)
- Price stability is the single or primary objective
- Price stability together with non-conflicting objectives but without priority
- Price stability plus others goals including stability of financial system that may conflict with the former, without priority
- Price stability together with objective of economic growth / economic development with no priority
- Objectives do not include price stability
- Score row: 1 0.75 0.50 0.25 0

---

### Policy formulation (0.15)

### 6. Who formulates monetary policy (0.25)
- Central Bank has the legal authority
- Central Bank participates together with government
- Central Bank in an advisory capacity
- Government alone formulates monetary policy
- Score row: 1 0.67 0.33 0

### 7. Government directives and resolution of conflicts (0.50)
- Central Bank given final authority over issues defined in the law as objectives
- Government has final authority over issues not clearly defined as Central Bank goals
- Final decision up to a council whose members are from the Central Bank, executive branch, and legislative branch
- Legislative branch has final authority
- Executive branch has final authority, but subject to due process and possible protest by Central Bank
- Executive branch has unconditional authority over policy
- Score row: 1 0.8 0.6 0.4 0.2 0

### 8. Central Bank given active role in formulation of government’s budget (0.25)
- Yes
- No
- Score row: 1 0

---

### Policy formulation (0.15) — alternate breakdown

### 7. Who formulates monetary policy (0.50)
- Central Bank has the legal authority
- Executive branch holds the final decision on exchange rate policy
- Central Bank participates on monetary policy formulation in an advisory capacity or faces legal limitations on monetary instruments or interest rates
- Government formulates monetary policy alone
- Score row: 1 0.67 0.33 0

### 8. Government directives and resolution of conflicts (0.30)
- Central Bank given final authority over issues defined in the law as objectives
- Government has final authority over issues not clearly defined as Central Bank goals
- Final decision up to a council whose members are from the Central Bank, executive branch, and legislative branch
- Legislative branch has final authority
- Executive branch has final authority, but subject to due process and possible protest by Central Bank
- Executive branch has unconditional authority over policy
- Score row: 1 0.8 0.6 0.4 0.2 0

### 9. Central Bank involvement in debt approval (0.20)
- Approves government debt
- Legally required to provide opinion on technical aspects
- No involvement at all
- Score row: 1 0.67 0.33

---

### Central Bank lending (0.50)

### 9. Limitations on advances (0.30)
- Advances to government prohibited
- Permitted but subject to limits in terms of absolute cash amounts or relative limits (government revenues)
- Permitted subject to relatively accommodative limits (more than 15 percent of government revenues)
- No legal limitations on advances. Subject to negotiations with government
- Score row: 1 0.67 0.33 0

### 10. Limitations on securitized lending (0.20)
- The same as in 9
- Score row: 1 0.67 0.33 0

### 11. Who decides control of terms of lending to government (0.20)
- Central bank controls terms and conditions
- Terms of lending specified in law, or Central Bank given legal authority to set conditions
- Law leaves decision to negotiations between the Central Bank and government
- Executive branch alone decides and imposes to the Central Bank
- Score row: 1 0.67 0.33 0

### 12. Beneficiaries of Central Bank lending (0.10)
- Only central government
- Central and state governments, as well as further political subdivisions
- Also public enterprises can borrow
- Central Bank can lend to all of the above and to the private sector
- Score row: 1 0.67 0.33 0

### 13. Type of limits when they exist (0.05)
- As an absolute cash amount
- As a percentage of Central Bank capital or other liabilities
- As a percentage of government revenues
- As a percentage of government expenditure
- Score row: 1 0.67 0.33 0

### 14. Maturity of loans (0.05)
- Limited to a maximum of 6 months
- Limited to a maximum of 1 year
- Limited to a maximum of more than one year
- No legal upper bounds
- Score row: 1 0.67 0.33 0

---

### Central Bank lending (0.40) — alternate breakdown

### 10. Limitations on advances (0.15)
- Advances to government prohibited
- Limited by small percentage of government revenues or by monetary program
- Allowed under lax limits (more than 15 percent of government revenues)
- Allowed without limits
- Score row: 1 0.67 0.33 0

### 11. Lending to Government (0.30)
- Not allowed
- In the secondary market with restricted limits
- In the secondary market with lax or without limits
- In the primary market with limits or approved by Central Bank Board with a qualified majority
- In the primary market without limits
- Score row: 1 0.75 0.50 0.25 0

### 12. Who decides financing conditions to government (0.10)
- Central Bank defines terms and conditions
- Defined by law
- The law allows negotiations between government and Central Bank
- Executive decides independently
- Score row: 1 0.67 0.33 0

### 13. Beneficiaries of Central Bank financing (0.10)
- Only the government
- Government plus local governments
- All of the above plus public enterprises
- All of the above and to the private sector
- Score row: 1 0.67 0.33 0

### 14. Interest rates in advances or lending (0.10)
- At market rates
- Interest rates not specified in law
- At below market rates
- Score row: 1 0.5 0

### 15. LOLR (0.15)
- For liquidity purposes with limitations (up to 180 days or up to banks’ equity), or no legal provision for emergency lending.
- For liquidity at conditions defined by the central bank
- Provisions for constructive ambiguity or rediscount of commercial bank loans
- Open assistance to cope with solvency problems
- To finance bank restructuring and/or paying deposit insurance
- Score row: 1 0.75 0.50

### 15. Restrictions on interest rates (0.05)
- Must be at market rate
- On loans to government can not be lower than a certain floor
- Interest rate on Central Bank loans can not exceed a certain ceiling
- No explicit legal provisions regarding interest rate in Central Bank loans
- No interest rate charge on government’s borrowing from Central Bank
- Score row: 1 0.75 0.50 0.25 0

### 16. Prohibition on Central Bank lending in primary market to Government (0.05)
- Prohibition from buying government securities in primary market
- No prohibition
- Score row: 1 0

---

### Financial autonomy (0.10)
- Government should maintain central capital integrity
- Government is legally allowed to capitalize the central bank
- The law does not allow the government to capitalize the central bank
- The Central Bank conducts quasi-fiscal operations.
- Score row: 1 0.5

---

### Accountability (0.10)

### 17. Accountability of Central Banks (0.75)
- Reports to executive branch and informs at least annually to Congress
- Reports to the executive once a year and submits an annual report to Congress
- Annual report to the executive. Informs to the executive branch whenever fundamental disequilibria emerge, or reports through the media without specific periodicity
- Issues annual report at specific time
- Distributes an annual report without establishing particular period of time for it
- Score row: 1 0.75 0.50 0.25 0

### 18. Central Bank transparency (0.25)

### 18. Central Bank transparency (0.25)

### Disclosure criteria (verbatim)
- Discloses detailed financial statements at least once a year with a certification of an independent auditor
- Discloses consolidated financial statements at least once a year with seal of the Banking Superintendent or other public sector authority
- Discloses financial statements at least once a year, certified by an internal auditor
- Publishes partial financial statements
- Does not publish financial statements or the law authorizes the central bank to deviate from international accounting standards

### Associated numeric entries (as presented)
- 0.25
- 0
- 1
- 0.67
- 0.33
- 0
- 1
- 0.75
- 0.50
- 0.25
- 0
- 1
- 0.75
- 0.50
- 0.25
- 0

### Contextual placement in the appendix
- Appears in APPENDIX II under "GMT Index and the Independence of Central Banks (Latin America and the Caribbean)" alongside country-level indices and related tables (pages -29 to -37).
- The transparency item is one component of the assessment framework used in the appendix (heading shows weight "(0.25)").

*Source: _wp0575 - 18. Central Bank transparency (0.25).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0575.pdf_