## _wp04223 - 1.   Consumer Price Index Market Basket

## Source details

**Canonical URL:** [_wp04223 - 1.   Consumer Price Index Market Basket](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04223.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04223.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04223.pdf.json)

---

### Introduction and scope
- Since October 2002, consumer price inflation in Guinea has been on a progressively increasing trend, reaching a peak of 16 percent year-on-year in May 2003.
- The paper develops stylized facts about the inflationary process in Guinea, focusing particularly on the relationship between money growth and inflation.
- Quarterly data are used for the period September 1991–March 2003.
- Main methodological approach: a bivariate inflation model containing monetary growth and CPI inflation, applying cointegration techniques and an error-correction framework.
- Prior related applications cited: Sacerdoti and Xiao (2001) for Madagascar; Kuijs (1998) for Nigeria; Nachega (2001a and b) for Cameroon and Uganda.
- Data limitations noted: statistical weaknesses in the computation of the CPI and the large weight of food in the index.

### Key findings on money and inflation
- There is a significant long-run relationship between money growth and consumer price inflation during the past ten years (relative to the data period).
- Evidence suggests the long-run relationship between money growth and inflation has been reinforced in recent years.
- Using an error-correction model and impulse response analysis, a monetary shock has an immediate impact on inflation.
- Monetary policy may operate through two related channels:
  - (i) direct liquidity management, to contain reserve money growth; and
  - (ii) a policy mix that does not lead to excessive broad money expansion in the economy.
- Despite exogenous factors (such as rainfall and food supply), episodes of high money supply growth have a significant impact on CPI inflation, even if limited by those exogenous factors.

### Guinea’s inflation experience: periods and stylized facts
- Post-independence inflation divided into three periods:
  - Administrative control of prices (first two decades after independence):
    - Inflation remained between 5 to 10 percent; it averaged 6.7 percent from 1964 to 1979.
    - Administrative controls distorted markets and price formation.
  - High inflation (roughly 1979–1989):
    - Surge in 1979 and early 1980s driven by partial economic liberalization, fiscal imbalances, loose monetary policy, and weak supply responses.
    - Reforms following the 1984 coup culminated in an annual average rate of 65 percent in 1986 when:
      - the national currency was devalued by 92 percent,
      - trade was liberalized, and
      - price controls (except those on fuel and rice) were removed.
  - Disinflation and relative stability (post-1986 until end-1999):
    - Annual average rate of inflation dropped sharply and remained low until the end of 1999.
    - Disinflation supported by favorable climatic conditions increasing agricultural output and food production.
    - Other contributing factors: low price increases in food products and basic commodities (housing and transport), exchange rate stability, and a decline in imported prices (especially imported rice).
    - Except in 1995, the declining trend continued; annual average CPI inflation was under 2 percent in 1997.
    - Broad money growth was limited, bank credit to the government was contained, and net foreign assets remained adequate.

### Recent dynamics and risks
- Since 2000, the 12-month rate of change in the CPI has been volatile.
- A period of high inflation in 2000 was driven mostly by a rise in nonfood prices.
- Resurgence of inflation in late 2002 and 2003:
  - inflation rose to 6.1 percent in December 2002 and 10.4 percent in March 2003, with simultaneous increases in food and nonfood price indices.
- Twelve-month CPI inflation returned to double digits in the second half of 2000: it reached 10.4 percent, and remained around or above 7 percent until the last quarter of 2001.
- Monetary volatility mirrored inflation: substantial increases in base and broad money in late 2002–2003 correlated with rising credit to the government and falling net foreign assets of the central bank.
- Correlation between money growth and CPI inflation has grown over time, consistent with monetization of large fiscal deficits and large increases in public consumption.

### Composition and structure of the Guinean CPI
- CPI coverage: 165 items since 1987; food items account for a large proportion of the basket.
- CPI weights (In percent):
  - Food 41.9
  - Nonfood 58.0
  - Clothing 7.0
  - Health 9.3
  - Housing 15.1
  - Electricity and water 7.4
  - Transport 12.0
  - Leisure and culture 7.2
  - All items 100.0
  - Domestic products 58.9
  - Imported products 41.2
- Key influences on CPI:
  - Domestic inputs, transport costs to retail markets, and climatic conditions (rainfall) drive food price movements and seasonal price jumps between harvests.
  - Domestic product prices are more volatile than imported product prices and weigh more heavily on CPI movements.
- Data limitations:
  - Weights based on a 1991 survey; some items obsolete; index frozen for items not consumed in Guinea (these account for more than 3 percent of the CPI).
  - Sample restricted to five markets in Conakry; supermarkets and small retail shops excluded.
  - No rural coverage or urban/rural breakdown.
  - Planned revision: expand items to 317 and new weights from a 4,416-household survey (vs. 300 households previously); extend coverage to all major urban centers.

### Theoretical framework and empirical specification
- Framework: small open economy with flexible exchange rate; CPI is weighted average of domestic and imported goods; focus on domestic determinants due to limited data.
- Money market equilibrium and money demand depend on real income y, inflationary expectations ep, domestic dr and foreign fr interest rates.
- Two bivariate models considered: Δz_t = resp_t and Δz_t = Mp_t (monetary growth and inflation).
- Monetary variables: reserve money (RES) and broad money (M2).
- Sample restricted to September 1991–March 2003 due to quarterly monetary data availability from September 1991.
- Estimation: unrestricted VAR, cointegration tests using Johansen (1988); variables in logarithms, not seasonally adjusted.

### Unit root and cointegration findings
- ADF unit root tests suggest variables are I(1): nonstationary in levels, stationary in first differences.
  - Examples from Table 2:
    - LCPI levels: ADF t-value -3.070, Probability 0.1261; first differences ADF t-value -6.127, Probability 0.000
    - LRES levels: ADF t-value -1.579, Probability 0.786; first differences ADF t-value -9.649, Probability 0.000
- Johansen tests: hypothesis of no cointegration between money and inflation is rejected in all four models at the 5 percent level; one long-run relationship identified between price level and money.

### Long-run estimated relationships (normalized cointegrating coefficients)
- Four long-run relationships reported (coefficients preserved exactly as in source):
  - Trend LRES LNONF: 007.0160.0 +=
  - Trend LM LNONF: 004.02257.0 +=
  - Trend LRES LCPI: 004.0477.0 +=
  - Trend LM LCPI: 004.02317.0 +=
- Interpretation and implications:
  - Increases in the money stock have a substantial long-term effect on inflation.
  - Impact of reserve money on composite CPI is larger than impact of broad money (currency in circulation ~80 percent of reserve money and ~55 percent of broad money).
  - For the nonfood index, broad money has a larger impact than reserve money, and impact of monetary variables is more subdued.
  - Coefficients imply a doubling in the money supply results in a price increase of only about 16 to 48 percent; possible reasons include price-level response only to large money changes, positive trend term indicating persistent inflation, data limitations, and potential omitted variable bias.

### Structural break analysis and changes over time
- Potential breakpoints identified from inflation trends: December 1992–December 1997 (disinflationary), March 1998–March 2003 (controlled around lower rates); December 1999 (greater volatility); December 2002 (accelerating inflation and dramatic money growth).
- Chow breakpoint tests reject the null of no structural change for three of the four long-run relationships across tested breaks (1997:4 onward and 1999:4 onward).
  - Examples of test statistics reported:
    - Structural break from 1997:4 onward — F-statistic 6.11 (0.0046) for LCPI-LM2; Log likelihood ratio 11.76 (0.0028).
    - Structural break from 1999:4 onward — F-statistic 8.85 (0.0006) for LCPI-LM2; Log likelihood ratio 16.21 (0.0003).
- Reestimated long-run relationships for subperiods show larger money coefficients:
  - 1997:4 – 2003:1:
    - LCPI = 0.445 LM2 + 0.005 Trend
    - LCPI = 0.544 LRes + 0.006 Trend
    - LNONF = 0.375 LM2 + 0.004 Trend
    - LNONF = 0.520 LRes + 0.030 Trend
  - 1999:4 – 2003:1:
    - LCPI = 0.357 LM2 + 0.004 Trend
    - LCPI = 0.564 Lres + 0.006 Trend
    - LNONF = 0.837 LM2 + 0.007 Trend
    - LNONF = 1.450 Lres + 0.043 Trend
- Conclusion: long-run pass-through between money and prices has become more significant in recent years.

### Short-run inflation dynamics (VECM and impulse responses)
- Error-correction model specification: Δx_t = αβ′ x_{t-1} + Σ Γ Δx_{t-i} + ε_t.
- Error-correction term coefficients in inflation equations significant at 5 percent, ranging between -0.604 and -0.733, implying about two-thirds of excess money supply is reabsorbed through inflation each quarter.
- Lagged money growth effect in the short run appears counterintuitive in estimated model (acceleration of money growth associated with reductions in inflation).
- Lagged inflation coefficients significant and positive, indicating inflation persistence.
- Impulse-response analysis (Cholesky decomposition):
  - A shock in money growth raises inflation significantly over about two years, then stabilizes at a higher level — consistent with permanent increase in price levels following money shocks.
  - A contemporaneous shock in inflation causes an immediate jump in price levels that decreases over time; after between five and six periods the impact turns negative, suggesting a rapid policy response to high inflation that reduces inflation subsequently.

### Policy implications and recommendations
- Evidence supports both short-term and long-term significant links between money and inflation.
- Monetary policy should focus on reserve money because of its closer link to inflation.
- Active liquidity management (presently issuance of sterilization bills) is needed to meet reserve money targets.
- To prevent excessive broad money expansion, an adequate macroeconomic policy mix is required, notably sustainable fiscal policy.
- Financing the government deficit through treasury bills, while preferable to monetary financing, would expand net domestic assets and broad money, with substantial impact on price levels.
- Given the role of exogenous supply-side factors (notably food supply and rainfall), policy should account for limits on the potency of monetary measures while preventing monetary conditions that would exacerbate inflationary pressures.

### Appendix — Measures of inflation and early warning indicators
- One-year (12-month) growth rates can be misleading and slow to reflect emerging inflation; one-month and three-month annualized rates provide earlier signals.
- Example: one-month and three-month rates indicated inflationary pressures as early as July 2002 while 12-month rates remained low until October 2002.
- Outliers and base effects significantly distort 12-month rates. Illustration:
  - September 2000: a 5 percent monthly increase (largest monthly increase) caused the 12-month rate to be 4.8 percentage points higher for the next 11 months than it would have been without the outlier.
- Proposed mitigation: use three-month growth rates to better capture current inflation developments while reducing noise from one-month series.
- Selected data points from Jan-2000 to Mar-2003 (as reported):
  - Sep-00 Index 166.6; 12-Month Growth Rate 10.4; 1-Month Growth rate 10.1; Corrected 12-Month Growth Rate 5.4
  - Dec-02 Index 180.4; 12-Month Growth Rate 6.1; 1-Month Growth rate 6.0; Corrected 12-Month Growth Rate 5.4
  - Mar-03 Index 188.7; 12-Month Growth Rate 10.4; 1-Month Growth rate 9.9; Corrected 12-Month Growth Rate 9.6

*Source: _wp04223 - 1. Consumer Price Index Market Basket (PDF chapter/section).*

### 1.   Consumer Price Index Market Basket.................................................................................

### _wp04223 - 1.   Consumer Price Index Market Basket

### Introduction and scope
- Since October 2002, consumer price inflation in Guinea has been on a progressively increasing trend, reaching a peak of 16 percent year-on-year in May 2003.
- The paper develops stylized facts about the inflationary process in Guinea, focusing particularly on the relationship between money growth and inflation.
- Quarterly data are used for the period September 1991–March 2003.
- The main methodological approach is a bivariate inflation model containing monetary growth and CPI inflation, applying cointegration techniques and an error-correction framework.
- Prior related applications cited in the text include Sacerdoti and Xiao (2001) for Madagascar, Kuijs (1998) for Nigeria, and Nachega (2001a and b) for Cameroon and Uganda.
- Data limitations are noted, and the paper cautions about statistical weaknesses in the computation of the CPI and the large weight of food in the index.

### Key findings on money and inflation
- There is a significant long-run relationship between money growth and consumer price inflation during the past ten years (relative to the data period).
- Evidence suggests the long-run relationship between money growth and inflation has been reinforced in recent years.
- Using an error-correction model and impulse response analysis, a monetary shock has an immediate impact on inflation.
- The analysis includes, alternatively, broad money and reserve money variables, indicating monetary policy may operate through two related channels:
  - (i) direct liquidity management, to contain reserve money growth; and
  - (ii) a policy mix that does not lead to excessive broad money expansion in the economy.
- Despite the importance of exogenous factors (such as rainfall and food supply), episodes of high money supply growth have a significant impact on CPI inflation, even if this impact is limited by those exogenous factors.

### Guinea’s inflation experience: periods and stylized facts
- Post-independence inflation can be divided into three periods (Figure 1):
  - Administrative control of prices (first two decades after independence):
    - Inflation remained between 5 to 10 percent; it averaged 6.7 percent from 1964 to 1979.
    - Administrative controls distorted markets and price formation.
  - High inflation (roughly 1979–1989):
    - Surge in 1979 and the early 1980s prompted by partial economic liberalization, fiscal imbalances, loose monetary policy, and weak supply responses.
    - With reforms following the 1984 coup, inflation increased dramatically, culminating in an annual average rate of 65 percent in 1986 when:
      - the national currency was devalued by 92 percent,
      - trade was liberalized, and
      - price controls (except those on fuel and rice) were removed.
  - Disinflation and relative stability (post-1986 until end-1999):
    - Annual average rate of inflation dropped sharply and remained low until the end of 1999.
    - Disinflation supported by favorable climatic conditions that significantly increased agricultural output and food production.
    - Other contributing factors: low price increases in food products and basic commodities (housing and transport), exchange rate stability, and a decline in imported prices (especially imported rice).
    - Except in 1995, when elections and heavy rains pushed prices upward, the declining trend continued.
    - Annual average CPI inflation was under 2 percent in 1997.
    - Broad money growth was limited, bank credit to the government was contained, and the level of net foreign assets remained adequate.

### Recent dynamics and risks
- Figure 2 (described) depicts prices and broad money year-on-year growth over the past 10 years and highlights stylized facts:
  - Since 2000, the 12-month rate of change in the CPI has been volatile.
  - A period of high inflation in 2000 was driven mostly by a rise in nonfood prices (see Figure 3), followed by a drastic reduction in the 12-month rate of CPI inflation (series continues beyond provided excerpt).

### Policy implications and recommendations
- The empirical results support the argument in favor of an active monetary policy to maintain inflation at low levels.
- Monetary policy should emphasize:
  - Containing reserve money growth through direct liquidity management.
  - Avoiding policy mixes that permit excessive broad money expansion.
- Given the role of exogenous supply-side factors (notably food supply and rainfall), policy should account for these limits on the potency of monetary measures while preventing monetary conditions that would exacerbate inflationary pressures.

*Source: _wp04223 - 1. Consumer Price Index Market Basket (PDF chapter/section).*

### 1.1 percent in December 2001. The rate of inflation remained below 3 percent until

### _wp04223 - 1.1 percent in December 2001. The rate of inflation remained below 3 percent until

### Inflation developments (Dec 1992–Mar 2003)
- Resurgence of inflation in late 2002 and 2003: inflation rose to 6.1 percent in December 2002 and 10.4 percent in March 2003, with simultaneous increases in food and nonfood price indices.
- Twelve-month CPI inflation returned to double digits in the second half of 2000: it reached 10.4 percent, and remained around or above 7 percent until the last quarter of 2001.
- Monetary volatility mirrored inflation: substantial increases in base and broad money in late 2002–2003 correlated with rising credit to the government and falling net foreign assets of the central bank.
- Correlation between money growth and CPI inflation has grown over time, consistent with monetization of large fiscal deficits and large increases in public consumption.

### Composition and structure of the Guinean CPI
- CPI coverage: 165 items since 1987; food items account for a large proportion of the basket.
- Tabled CPI weights (In percent):
  - Food 41.9
  - Nonfood 58.0
  - Clothing 7.0
  - Health 9.3
  - Housing 15.1
  - Electricity and water 7.4
  - Transport 12.0
  - Leisure and culture 7.2
  - All items 100.0
  - Domestic products 58.9
  - Imported products 41.2
- Key influences on CPI:
  - Domestic inputs, transport costs to retail markets, and climatic conditions (rainfall) drive food price movements and seasonal price jumps between harvests.
  - Domestic product prices are more volatile than imported product prices and weigh more heavily on CPI movements.
- Noted data limitations:
  - Weights based on a 1991 survey; some items obsolete; index frozen for items not consumed in Guinea (these account for more than 3 percent of the CPI).
  - Sample restricted to five markets in Conakry; supermarkets and small retail shops excluded.
  - No rural coverage or urban/rural breakdown.
  - Planned revision: expand items to 317 and new weights from a 4,416-household survey (vs. 300 households previously); extend coverage to all major urban centers.

### Theoretical framework and empirical specification
- Framework: small open economy with flexible exchange rate; CPI is weighted average of domestic and imported goods; focus on domestic determinants due to limited data.
- Money market equilibrium: sd mm = . (money demand and supply relationship specified; money demand depends on real income y, inflationary expectations ep, domestic dr and foreign fr interest rates).
- Two bivariate models considered: Δz_t = resp_t and Δz_t = Mp_t (monetary growth and inflation).
- Monetary variables: reserve money (RES) and broad money (M2).
- Sample restricted to September 1991–March 2003 due to quarterly monetary data availability from September 1991.
- Estimation: unrestricted VAR, cointegration tests using Johansen (1988); variables in logarithms, not seasonally adjusted.

### Unit root and cointegration findings
- ADF unit root tests suggest variables are I(1): nonstationary in levels, stationary in first differences (see reported ADF t-values and probabilities).
  - Examples from Table 2:
    - LCPI levels: ADF t-value -3.070, Probability 0.1261; first differences ADF t-value -6.127, Probability 0.000
    - LRES levels: ADF t-value -1.579, Probability 0.786; first differences ADF t-value -9.649, Probability 0.000
- Johansen tests: hypothesis of no cointegration between money and inflation is rejected in all four models at the 5 percent level; one long-run relationship identified between price level and money.

### Long-run estimated relationships (normalized cointegrating coefficients)
- Four long-run relationships reported (coefficients shown exactly as in source):
  - Trend LRES LNONF: 007.0160.0 +=
  - Trend LM LNONF: 004.02257.0 +=
  - Trend LRES LCPI: 004.0477.0 +=
  - Trend LM LCPI: 004.02317.0 +=
- Interpretation:
  - Increases in the money stock have a substantial long-term effect on inflation.
  - Impact of reserve money on composite CPI is larger than impact of broad money (currency in circulation ~80 percent of reserve money and ~55 percent of broad money).
  - For the nonfood index, broad money has a larger impact than reserve money, and impact of monetary variables is more subdued.
  - Coefficients imply a doubling in the money supply results in a price increase of only about 16 to 48 percent; possible reasons include price-level response only to large money changes, positive trend term indicating persistent inflation, data limitations, and potential omitted variable bias.

### Structural break analysis and changes over time
- Potential breakpoints identified from inflation trends: December 1992–December 1997 (disinflationary), March 1998–March 2003 (controlled around lower rates); December 1999 (greater volatility); December 2002 (accelerating inflation and dramatic money growth).
- Chow breakpoint tests reject null of no structural change for three of the four long-run relationships across tested breaks (1997:4 onward and 1999:4 onward); F-statistics and log-likelihood ratios reported (examples):
  - Structural break from 1997:4 onward — F-statistic 6.11 (0.0046) for LCPI-LM2; Log likelihood ratio 11.76 (0.0028).
  - Structural break from 1999:4 onward — F-statistic 8.85 (0.0006) for LCPI-LM2; Log likelihood ratio 16.21 (0.0003).
- Reestimated long-run relationships for subperiods show larger money coefficients:
  - 1997:4 – 2003:1:
    - LCPI = 0.445 LM2 + 0.005 Trend
    - LCPI = 0.544 LRes + 0.006 Trend
    - LNONF = 0.375 LM2 + 0.004 Trend
    - LNONF = 0.520 LRes + 0.030 Trend
  - 1999:4 – 2003:1:
    - LCPI = 0.357 LM2 + 0.004 Trend
    - LCPI = 0.564 Lres + 0.006 Trend
    - LNONF = 0.837 LM2 + 0.007 Trend
    - LNONF = 1.450 Lres + 0.043 Trend
- Conclusion: long-run pass-through between money and prices has become more significant in recent years.

### Short-run inflation dynamics (VECM and impulse responses)
- Error-correction model specification: Δx_t = αβ′ x_{t-1} + Σ Γ Δx_{t-i} + ε_t.
- Error-correction term coefficients in inflation equations significant at 5 percent, ranging between -0.604 and -0.733, implying about two-thirds of excess money supply is reabsorbed through inflation each quarter.
- Lagged money growth effect in the short run appears counterintuitive in estimated model (acceleration of money growth associated with reductions in inflation).
- Lagged inflation coefficients significant and positive, indicating inflation persistence.
- Impulse-response analysis (Cholesky decomposition):
  - A shock in money growth raises inflation significantly over about two years, then stabilizes at a higher level — consistent with permanent increase in price levels following money shocks.
  - A contemporaneous shock in inflation causes an immediate jump in price levels that decreases over time; after between five and six periods the impact turns negative, suggesting a rapid policy response to high inflation that reduces inflation subsequently.

### Policy implications and recommendations
- Evidence supports both short-term and long-term significant links between money and inflation.
- Monetary policy should focus on reserve money because of its closer link to inflation.
- Active liquidity management (presently issuance of sterilization bills) is needed to meet reserve money targets.
- To prevent excessive broad money expansion, an adequate macroeconomic policy mix is required, notably sustainable fiscal policy.
- Financing the government deficit through treasury bills, while preferable to monetary financing, would expand net domestic assets and broad money, with substantial impact on price levels.

### Appendix — Measures of inflation and early warning indicators
- One-year (12-month) growth rates can be misleading and slow to reflect emerging inflation; one-month and three-month annualized rates provide earlier signals.
- Example: one-month and three-month rates indicated inflationary pressures as early as July 2002 while 12-month rates remained low until October 2002.
- Outliers and base effects significantly distort 12-month rates. Illustration:
  - September 2000: a 5 percent monthly increase (largest monthly increase) caused the 12-month rate to be 4.8 percentage points higher for the next 11 months than it would have been without the outlier.
- Proposed mitigation: use three-month growth rates to better capture current inflation developments while reducing noise from one-month series.
- Selected data from Table 7 (Index (1991=100), 12-Month Growth Rate, and related series) preserved in source for Jan-2000 to Mar-2003; sample observations include:
  - Sep-00 Index 166.6; 12-Month Growth Rate 10.4; 1-Month Growth rate 10.1; Corrected 12-Month Growth Rate 5.4
  - Dec-02 Index 180.4; 12-Month Growth Rate 6.1; 1-Month Growth rate 6.0; Corrected 12-Month Growth Rate 5.4
  - Mar-03 Index 188.7; 12-Month Growth Rate 10.4; 1-Month Growth rate 9.9; Corrected 12-Month Growth Rate 9.6

*Source: Guinean authorities; DNS; Banque Centrale de la Republique de Guinee (BCRG); staff calculations (from the supplied IMF working paper text).*

### References

### _wp04223 - References

### References
- Blinder, A., 1997, “Commentary,” Federal Reserve Bank of St. Louis Review, (May/June), pp. 157-60.            
- Cecchetti, S., 1997, “Measuring Short-Run Inflation for Central Bankers,” Federal Reserve Bank of St. Louis Review, (May/June), pp. 143-55. 
- Dickey, D., and W.A. Fuller, 1981, “Likelihood Ratio Statistics for Autoregressive Time Series with a Unit Root,” Econometrica, Vol. 49 (July), pp.1057–72. 
- Durevall, D., and N.S. Ndung’u, 1999, “A Dynamic Model of Inflation for Kenya, 1974-1996,” IMF Working Paper 99/97 (Washington: International Monetary Fund). 
- Johansen, S., 1988, “Statistical Analysis of Cointegration Vectors,” Journal of Economic Dynamics and Control, Vol. 12 (June–September), pp. 231-54. 
- Kuijs, L., 1998, “Determinants of Inflation, Exchange Rate, and Output in Nigeria,” IMF Working Paper 98/160 (Washington: International Monetary Fund). 
- Maehle, N., 2002, “Inflation in Kenya—Signal Extraction for Policy Purposes,” Kenya: Selected Issues and Statistical Appendix, IMF Country Report No. 02/84 (Washington: International Monetary Fund). 
- Mankiw, N.G., 2003, Macroeconomics, New York: Worth Publishers. 
- Nachega, J-C., 2001a, “A Cointegration Analysis of Broad Money Demand in Cameroon,” IMF Working Paper 01/26 (Washington: International Monetary Fund). 
- ______, 2001b, “Financial Liberalization, Money Demand and Inflation in Uganda,” IMF Working Paper 01/118 (Washington: International Monetary Fund). 
- Sacerdoti, E., and Y. Xiao, 2001, “Inflation Dynamics in Madagascar, 1971-2000,” IMF Working Paper 01/168 (Washington: International Monetary Fund). 
- Sriram, S., 2000, “A Survey of Recent Empirical Money Demand Studies,” IMF Staff Papers, Vol. 47, no. 3, pp. 334-65.

*Source: _wp04223 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04223.pdf_