Building Integrated Economies in West Africa
Chapter

Chapter 18. Monetary Policy and Inflation

Author(s):
Alexei Kireyev
Published Date:
April 2016
Share
  • ShareShare
Show Summary Details
Author(s)

The objective of this chapter is to determine the impact of a change in the key interest rates of the Central Bank of West African States (BCEAO), as well as other economic, monetary, and financial aggregates, on inflation in the West African Economic and Monetary Union (WAEMU). Specifically, the effects of interest rates, the money supply, imported inflation from the euro area, domestic credit, the fiscal deficit, and the output gap on inflation are evaluated. To this end, error correction models (ECMs) are used to calculate short- and long-term elasticities, while vector autoregressive (VAR) models are used to evaluate the generalized impulse response functions. The results show that most monetary and financial variables have an impact on the inflation rate. The impact of domestic credit on inflation seems to be greater than that of interest rates. Furthermore, the impact of the fiscal deficit on price trends is generally felt only over the long term. Therefore, efforts to control inflationary pressures should focus on the use of monetary policy instruments as well as on structural reforms to increase supply, which partially explains the evolution of the overall price level in the short term.

Targeting price stability

The institutional reform of the West African Economic and Monetary Union (WAEMU), which took effect on April 1, 2010, established price stability as an explicit objective for the central bank. The framework implemented by the bank with the aim of achieving this objective comprises three components: a definition of price stability as an annual variation in the Harmonized Index of Consumer Prices (IHPC) of between 1 and 3 percent; the use of the inflation forecast as the principal indicator for the implementation of monetary policy; and the adoption by the Monetary Policy Committee, on a quarterly basis, of the measures necessary to achieve this objective.

The overall aim of this paper is to determine the impact of different economic, monetary, and financial variables on inflation in economies of WAEMU. Specifically, the intent is to measure the impact of a change in the key BCEAO rates, as well as other economic and financial aggregates, on inflation. In this process, owing to the diverse factors that could affect the inflation forecasts, it would be important to look at all of the variables that could influence changes in the overall price level. Monetary and credit aggregates, the economic outlook, and measurements of the output gaps are of interest, in particular for inflation rates over the medium term. Over the short term, factors such as the exchange rate, prices for major commodities such as petroleum products, administered prices, and profit margins would seem to be more significant a priori (Assenmacher-Wesche and Gerlach 2007).

In empirical terms, a number of studies have been performed within the BCEAO since 1996, with the aim of achieving a better understanding of the implementation of monetary policy. These empirical studies have made it possible to identify the determining factors in price dynamics within the Union. They have addressed issues such as imported inflation (Dembo Toé 2010;Doe and Diallo 1997), interest rates (Nubukpo 2003; Diop and Adoby 1997), the money supply (Dembo Toé and Hounkpatin 2007), domestic output (Nubukpo 2003), food production (Diallo 2003), public spending (Doe and Diallo 1997), and the nominal effective exchange rate (Dembo Toé 2010).

Most of these studies use relatively similar methodologies. They usually construct and estimate econometric models of the VAR or ECM type. Specifically, Diop and Adoby (1997) deal with the determination of “simple ratios for measuring the impact of monetary policy on prices” and it contains calculations of elasticity coefficients that allow measuring the impact of monetary policy on prices. Based on monthly data, they estimate a model derived from the balance in the money market. This model has as explanatory variables the money supply (M2) and the interest rate from money market auctions. The results obtained revealed differences in the effects of variations in the money supply and the interest rate on prices within the Union’s member countries. An increase of 1 percentage point in the money supply over the long term resulted in a rise in the consumer price index of 0.43 percentage point in Benin, 0.28 percentage point in Burkina Faso, 0.45 percentage point in Côte d’Ivoire, and 0.41 percentage point in Mali. The influence of the money supply on prices appeared to be minor in Niger, Senegal, and Togo. The authors explain the absence of a relationship in certain countries, Senegal in particular, by the fact that a strictly administered price regime was in place during the period of the study. An increase of 1 percentage point in the interest rate over the long term caused a drop in prices by 0.30 percentage point in Benin, 0.18 percentage point in Burkina Faso, 0.41 percentage point in Niger, and 0.36 percentage point in Senegal. In Côte d’Ivoire, Mali, and Togo, the interest rate did not have a significant influence on inflation.

Empirical studies performed in other African countries reveal the impact of money supply and the output gap on inflation. Barnichon and Peiris (2008) use the following as explanatory variables for inflation in countries in sub-Saharan Africa: the output gap, money (the real money gap, that is, the difference between the money supply and demand), and rainfall. The results of their study produce elasticities of 0.28 for the output gap, 0.34 for money, and −0.13 for rainfall. The elasticity of the output gap is 0.42 for the countries outside the CFA zone and 0.32 for countries in the sample within the CFA zone (Cameroon, Côte d’Ivoire, Mali, Niger, Senegal). The elasticity of money is also higher in the countries outside the CFA zone (0.37) than in countries within the CFA zone (0.15). Ocran (2007), in an inflation model for Ghana, obtains a money supply elasticity of 0.42. Kovanen (2011), however, shows that money explains only a small part of the evolution of prices in Ghana and obtained an elasticity of the output gap with respect to inflation of 0.91.

A number of studies also point to the role of inflation inertia in the majority of countries in sub-Saharan Africa and the limited ability to explain the evolution of inflation by changes in the money supply. For the WAEMU, Dembo Toé and Hounkpatin (2007) show that the current level of inflation depends heavily on lagged price changes. Thus, 82.6 percent of the forecast error of the IHPC in the WAEMU is due to its own innovations (residuals), 3.8 percent is due to those in the nominal effective exchange rate, 8.8 percent is due to the evolution of imported inflation, and 4.8 percent to variation in the money supply.

Adebiyi (2007) obtains similar results for Ghana and Nigeria. Over a four-year horizon, 72 percent of the forecast error for inflation in Nigeria and 93 percent of the forecast error in Ghana is due to its own innovations (residuals), while 27 percent in Nigeria and 6 percent in Ghana is due to fluctuations in the money supply. Furthermore, the elasticity of the inertial component of inflation is estimated to be 0.54 for countries in sub-Saharan Africa (Barnichon and Peiris 2008). The present study expands upon the approach taken by Diop and Adoby (1997) by taking other economic and financial variables into account in the modeling, on the one hand, and by considering more recent statistical data, on the other hand.

A number of studies conducted for WAEMU member countries have looked at the influence of different economic and financial variables, in addition to the money supply and the interest rate, on inflation. These variables include, in particular, lagged inflation, imported inflation, oil prices, the evolution of liquidity, and variables related to pressures on the commodity markets and food production. The study by Cecchiti, Chu, and Steindel (2000) defines three main categories of variables used to predict inflation. These are, first, the prices for raw materials (oil prices, gold prices, price indices for a set of commodities, and so on). A steady rise in prices for these products would result in an increase in inflation. Second, financial indicators (the exchange rate, monetary aggregates, the difference between long-term and short-term interest rates) would affect the inflation rate. A decline in the exchange rate or a rapid increase in monetary aggregates could signal a rise in inflation. Finally, indicators of the state of the real economy (capacity utilization rate, unemployment rate, and the like) could serve as variables for predicting inflation. A steady rise in the capacity utilization rate or a drop in the unemployment rate beyond a certain threshold would result in inflationary pressures.

Inflation dynamics

During the period January 2001 to December 2014, the year-over-year average inflation rate was 3.6 percent. If 2008 is excluded—a year marked by a sharp rise in prices for foodstuffs—the average inflation rate was 2.8 percent (Figure 18.1).

Figure 18.1.Evolution of Year-Over-Year Inflation in WAEMU

(Percent)

Source: BCEAO.

Figure 18.2 shows a similar profile for the evolution of inflation in the WAEMU and trends in overall liquidity. A lag between the fluctuations in the money supply and the changes in prices is visible.

Figure 18.2.Evolution of Inflation and Growth of Money Supply in WAEMU

Source: BCEAO.

The BCEAO’s marginal lending rate was modified a number of times during the period January 1997 to December 2014 (Figure 18.3). These changes are generally consistent with the evolution of inflation. Indeed, most of the reductions in the marginal lending rate occurred within the context of a slowdown in inflation, while increases in the rate took place during periods of an acceleration of inflation. Nevertheless, it should be noted that these rates remained fixed during certain episodes of inflationary pressures in the Union, in particular in 2001, 2005, and 2011.

Figure 18.3.Comparative Trends in the Inflation Rate in the WAEMU and the BCEAO Key Rate

(Year-over-Year)

Source: BCEAO.

Methodology

The specification of the relationship between inflation and the different variables is based on the theoretical and empirical foundations of the determinants of inflation. In theoretical terms, the changes related to monetary policy instruments, primarily the key lending rates and control over the money supply, can affect the overall price level. According to Mishkin and others (2010), monetary policy influences the level of economic activity through four channels—interest rates, bank lending, asset prices, and exchange rates—as well as expectations. In their study of the monetary policy being pursued by the BCEAO, Ary Tanimoune and Tenou (2010) found that the two main channels used in the WAEMU economies appear to be interest rates and lending. In this connection, the present study includes variables related to these channels, including the maximum marginal lending rate, the money market rate, the bank borrowing rate, and lending to the economy by banks.

Furthermore, with the aim of identifying other variables that could interact with the overall price level, two main theoretical models were considered—the traditional monetarist model and a model derived from the new Phillips curve. According to the monetarist approach, growth in the money supply, its lagged and current values, are the key variables affecting inflation. In addition to these variables, changes in food and energy prices (supply shocks), as well as an increase in budget spending, could affect overall price levels. The new formulation of the Phillips curve is based on the Keynesian school of thought. This theoretical approach identifies three main determinants of inflation, namely, the output gap, which represents the difference between actual output and potential output; expected inflation, lagged inflation, or both; and supply shocks. It should be mentioned that other models could be used to identify explanatory factors related to the overall price level (Mehra 1988). The choice of models takes into account empirical studies performed for WAEMU countries, as well as economic shocks, such as the currency devaluation of 1994 and the sharp rise in food prices in 2008.

To measure the link between inflation and the different variables, error correction and VAR models were used. The ECM models allowed for a determination of the short- and long-term elasticities among several variables. The VAR models were used to analyze the impact of each of the variables on the others. Finally, the impact of a shock on the variables and the reaction times of each variable following the shock on other variables were evaluated based on the impulse response functions of the VAR models.

In practical terms, an analysis of the stationarity properties of the variables showed that the variables are not stationary. They do have a first order of integration and there is a cointegrating relationship among them. Therefore, several versions of the ECM (equation 18.1) following the specification of Davidson and Hendry (1978) were estimated:

in which IHPCt represents the inflation rates for year t, M2t is the money supply at time t, and Xt is a set of explanatory variable or variables of the selected models; t is the index for the year, α is the constant, εt are the error terms,

The β1 and β2 parameters are coefficients that describe the short-term dynamics, while β4 and β5 are the long-term coefficients. The β3 parameter is the error correction coefficient. For the ECM to be valid, this coefficient needs to be negative, statistically significant, and have an absolute value of less than 1. All data series are derived from the BCEAO database. The study was performed on the basis of monthly and annual data, depending on the models used. The monthly data cover the period from January 2001 to December 2014. Given the absence of intra-annual data on GDP, annual data from 2001 to 2014 were used to estimate the elasticities of the output gap.

IHPC refers to the Harmonized Index of Consumer Prices in WAEMU countries, which has been calculated by the national statistics offices according to a harmonized methodology since 1997. The selection of the gross IHPC, rather than the underlying inflation indicator is due both to a desire not to deviate from the measurement used in the definition of target inflation, and to a desire to have benchmarks similar to those of other relevant studies that deal primarily with the gross IHPC.

The following dependent variables are subsumed for the estimated equations.

CRED refers to domestic credit

IHPC_ZEURO is the Harmonized Index of Consumer Prices in the euro area. It is a proxy for imported inflation in the equations based on monthly data.

The output gap (GAPUEMOA) represents the gap between the real level of GDP and its potential level. Two quite different paths can be taken to define potential GDP: the statistical approach, which consists of extracting a posteriori the trend, deterministic or stochastic, of a GDP series; and the economic approach, which attempts to determine the maximum level of activity compatible with price stability. In this study, the output gap is measured by the difference between the logarithm of actual output (real GDP) and the logarithm of potential output (Y*). Potential output is determined by the Hodrick-Prescott (HP) filter H-P filter (see Diop 2000 for details).

TGPM is the maximum marginal lending rate of the BCEAO. The central bank’s key lending rate is the minimum auction rate for one week. This rate was established in August 2008. The marginal lending rate (repo rate) is also a key rate of the BCEAO. It is equal to the minimum auction rate plus 100 basis points. Thus, the variations in these two key rates are technically synchronous. The marginal lending rate is used here primarily because of its availability throughout the entire period covered by the study.

TMM is the money market rate, while TDEBITEUR represents the bank borrowing rate.

DEP is the ratio of the government spending to the real GDP of the Union.

DEFICIT refers to the ratio of deficits to the real GDP of the Union.

The dummy variables DUM94 and DUM08 capture the respective effects of the shocks of 1994 and 2008.

The expected signs of variables are the following:

A decline in the key rates (−) allows banks to obtain refinancing at a lower cost. Banks can thus pass on the reduction in refinancing costs to the borrowing rates that they offer to households and businesses, which encourages activity, but this can also place upward pressure on demand and on prices. Conversely, a rise in key rates tends to boost borrowing rates, which could curb activity and could lead to a drop in prices.

The money supply should have a positive impact on inflation over the short and long terms. In fact, an increase in overall liquidity, in particular through domestic credit, leads to an increase in total demand, which results in additional inflation, all else equal. The same is true for the other monetary variables (currency in circulation, narrow money, base money).

The sign of the output gap (GAPUEMOA) should be positive. Theoretically, during peaks in the economic cycle, the output gap is positive: output is temporarily higher than its equilibrium level and enterprises can achieve this level of output only by incurring higher costs, in particular salary costs, which are reflected in prices. In this case, there is an increase in inflation. Conversely, during economic downturns, the gap is negative and inflation drops.

Imported inflation (IHPC_ZEURO) can also be responsible for cost inflation. Owing to globalization and the fact that a number of firms import a significant proportion of their raw materials or semifinished products, companies may be forced to raise the prices of their products to deal with a decline in the exchange rate, a rise in raw materials prices in the international market, or an external shock.

The fiscal deficit ratio (DEFICIT) is expected to have a positive effect on the overall price level. The effect of fiscal policy on prices in countries in the Union can be explained by the nature of public spending and the method used for financing fiscal deficits.

For credit (CRED), an increase in this lending could cause a rise in the overall price level in connection with the pressure that an easing of credit terms could put on the economy as a whole.

An increase in government spending (DEP) could induce an upward pressure on the overall price level. Thus, a positive sign would be expected.

Finally, economic shocks related to devaluation and to a steep rise in food prices could have a positive impact on inflation in the countries in the Union. Thus, the dummy variables (DUM94 and DUM08) should have positive signs.

Policies and inflation

The results of the estimates presented in Table 18.1 show the influence of monetary, financial, and economic variables on the dynamics of the IHPC. In an effort to avoid the risks of multicolinearity among the explanatory variables and to show the impact of each of them on the IHPC, they have been included gradually in the different regressions (see models 1 through 7 in Table 18.1). The econometric tests showed that the variables with a monthly frequency all have a first order of integration (I(1)), with the exception of the money market rate (TMM) and the fiscal deficit ratio (DEFICIT) for the annual data (see Annex 18.1, Annex Table 18.1.1). The variables in models 1 through 6 are all cointegrated, however, with the exception of those in model 7.

Table 18.1.Error Correction Models of the Link between Inflation and Different Variables in the WAEMU*
Explanatory VariablesModel 1Model 2Model 3Model 4Model 5Model 6Model 7
DLOG(TGPM)−0.0104***
(−8.9)
DLOG(TMM)−0.4016***
(−4.4)
DLOG(TDEBITEUR)−0.152**
(−2.64)
DLOG(M2)0.0199**
(2.5)
DLOG(IHPC_ZEURO)0.1086***
−2.2
DLOG(CRED)0.5432***
−13.89
D(DEFICIT)0.0087*0.0066***
−1.94(5.1)
LOG(IHPC(−1)−0.0167***−0.4560***−0.0503***−0.1359***−0.0976***−0.1469**
(−7.1)(−5.28)(−4.78)(−9.1)(−9.6)(−6.8)
LOG(TGPM(−1))−0.0752***
(−4.8)
LOG(TMM(−1))−0.2860***
(−4.92)
LOG(TDEBITEUR(−1))−0.1121***
(−5.40)
LOG(M2(−1))0.0229***
(8.3)
LOG(IHPC_ZEURO(−1))0.1815***
−6.3
LOG(CRED(−1))0.3278***
(9.01)
DEFICIT(–1)0.0105***0.0067***
(2.34)(4.23)
LOG(GAPUEMOA)0.1376***
−8.06
LOG(DEP)0.5718***
(7.15)
DUM940.2018***0.1095***
(32.5)(4.62)
DUM080.0030**0.0092***0.0043***
(2.5)(8.8)(10.2)
Constant0.1265***2.5130***−0.01670.21870.0803***−0.1099***−4.5898***
(7.1)(5.23)(−0.29)−1.1−5.8(−4.0)(−3.85)
R20.7180.8380.870.8180.6780.8370.907
Prob(F-Stat)00.0110.0010000
S.E.0.0010.0190.0060.0040.0020.0010.074
DW2.5712.8492.6862.5012.0982.3741.81
* Note: The values between (…) represent the Student’s t values. Models 1, 4, 5, and 6 are estimated on the basis of monthly data and models 2, 3, and 7 are based on annual data. (***) significant at the 1% level; (**) significant at the 5% level; (*) significant at the 10% level.The results of the Eviews program are provided in the Annex and the results of the robustness tests are available.
* Note: The values between (…) represent the Student’s t values. Models 1, 4, 5, and 6 are estimated on the basis of monthly data and models 2, 3, and 7 are based on annual data. (***) significant at the 1% level; (**) significant at the 5% level; (*) significant at the 10% level.The results of the Eviews program are provided in the Annex and the results of the robustness tests are available.

The adjustment coefficient, which appeared to be negative, significant (in models 1 through 6), and less than 1 in absolute terms, validates the ECM. This coefficient represents the speed at which any imbalance between the desired and actual levels of the IHPC is absorbed in the month following any shock. Consequently, the influence of the maximum marginal lending rate (TGPM), the money market rate (TMM), the borrowing rate (TDEBITEUR), the money supply (M2), domestic credit (CRED), imported inflation (IHPC_ZEURO), and the fiscal deficit ratio (DEFICIT) on the Harmonized Index of Consumer Prices in the WAEMU is sufficiently captured by the short-and long-term dynamics. Furthermore, the tests indicate a lack of a cointegration relationship between the output gap (GAPUEMOA) and the Harmonized Index of Consumer Prices (IHPC). In other words, long-term dynamics should be sufficient to explain the influence of a percentage variation in the gap between actual output and potential output on inflation in the WAEMU (model 7, Table 18.1).

The estimates show that the models are generally statistically significant with coefficients that present the expected signs. In addition, 67.8 percent and 90.7 percent of the fluctuations in the IHPC in the WAEMU can be explained by the short-term and long-term variables used, respectively. Furthermore, the coefficients of all of the explanatory variables are statistically significant at the 5 percent level over the short and long terms. Following a robust estimate of heteroscedasticity, the tests performed on the residuals showed that they are normally distributed, and that they are not autocorrelated but homoscedastic. In addition, the Ramsey test (RESET) was used to confirm that the models are properly specified.

Interest rates and inflation. The estimates indicate a negative and significant relationship between changes in the marginal lending rate and changes in inflation in the WAEMU over the short and long terms (Model 1). An increase of 1 percent in the marginal lending rate would result in a drop in prices in the Union by 0.01 percent and 0.08 percent over the short and long terms, respectively. In addition, the impulse response function (Annex Figure 18.1.1) of the VAR model shows that a shock to the BCEAO key rate, specifically the marginal lending rate, would result in a drop in inflation, which would stabilize after about 14 months. The impact of the marginal lending rate and the money market rate on inflation is transmitted primarily through bank borrowing rates. In fact, it is expected that an increase in the central bank’s key rates would result in a rise in bank borrowing rates and ultimately in a decline in the demand for credit and in lower prices.

On the whole, changes in the main interest rates have an impact on inflation in the WAEMU. Thus, by modifying the key rates, the central bank can influence the future level of inflation in the WAEMU. In this connection, the impact of interest rates on price trends is clearly evident over the short and long terms. At the same time, the impact is relatively weak compared with that of domestic credit and the bank borrowing rate, and particularly compared with the level of domestic output and imported inflation. An increase of 25 basis points in the BCEAO’s marginal lending rate would lead to a reduction in inflation of just 0.26 percentage point over the short term and by 0.19 percentage point over the long term. On this basis, actions aimed at countering inflationary pressures should be aimed both at the use of monetary policy instruments and at structural reforms to increase supply.

Money supply, domestic credit, and inflation. Changes in the money supply (M2) and in domestic credit have a significant impact on inflation in the WAEMU over both the short term and the long term (models 4 and 5). An increase in the money supply of 1 percentage point would result in additional inflation of 0.02 percentage point over the short term and 0.023 percentage point over the long term. These results are in line with those obtained by Barnichon and Peiris (2008) for countries in the CFA zone. The impulse from a monetary shock would result in a positive inflationary reaction that would dissipate after 14 months. The elasticity of inflation with respect to the money supply in the Union also appears to be weaker than that estimated for countries outside the CFA zone, in particular Ghana and Nigeria. Furthermore, growth in domestic credit of 1 percent would result in an increase in inflation of 0.54 percentage point over the short term and 0.33 percentage point over the long term. This finding is consistent with the results obtained by Héricourt and Matei (2007) for central and eastern European countries. The impact of domestic credit on inflation in the WAEMU is greater than the impact of the BCEAO’s key rates.

On the whole, changes in the money supply and in domestic credit have an influence on the dynamics of inflation in the WAEMU. The impact, however, is relatively weaker than that observed in other countries in sub-Saharan Africa. The low level of elasticity of inflation in relation to the money supply is a reflection of the prudent monetary policy stance in the WAEMU, which is characterized by the monitoring of injections of liquidity and moderate changes in base money. Indeed, the monetary policy implemented by the BCEAO has ensured that changes in the money supply are in line with changes in GDP, so as not to contribute to inflation in the Union.

The output gap and inflation. The impact of the output gap on inflation is estimated on the basis of annual data owing to the absence of quarterly GDP data over an extended period. The growth rate in potential output dropped from about 4 percent in 1972 to 1.5 percent on average in the 1980s. It fell following the devaluation in 1994 and remained at about 0.3 percent between 1995 and 2014 (Figure 18.4). The output gap calculated by the percentage variation in the difference between actual output and potential output has been consistently negative throughout the past 10 years. This situation reflects actual output that falls short of the economy’s potential. Figure 18.5 also shows an essentially positive relationship between the output gap and the inflation rate. An inverse relationship between the trends in inflation and the output gap has been observed during several periods, however, including 2002, 2004, 2006–07, 2009, and 2011, most of which were marked by droughts, when the theoretical relationship between the output gap and inflation does not hold up.

Figure 18.4.Actual Output and Potential Output in the WAEMU

(as a percentage)

Figure 18.5.Inflation and the Output Gap in the WAEMU

(as a percentage)

In the WAEMU countries, changes in GDP do, in fact, depend primarily on the performance of the primary sector (agricultural production). Thus, a substantial decline in agricultural production linked to a period of drought is reflected in a drop in actual output, a contraction of the output gap, and a rise in inflation. Likewise, a positive output gap is not often the result of an overheated economy, but rather strong agricultural production. A dummy variable has been introduced to neutralize some of the phenomena described above (DUM08).

The estimation of an ECM indicates the absence of a relationship between inflation and the production gap over the short term. Over the long term, an increase in the output gap by 1 percentage point would result in a rise in the inflation rate by 0.57 percentage point (model 7, Table 18.1). This elasticity of 0.14 is lower than the figure of 0.3 obtained by Barnichon and Peiris (2008) for countries in sub-Saharan Africa, the figure of 0.4 estimated by Diop (2000) for WAEMU member countries, and the indicator of 0.91 for Ghana (Kovanen 2011). This result is related to the stability of the inflation rate observed in the WAEMU zone.

The fiscal deficit and inflation. The estimations indicate that the fiscal deficit (and public spending, accordingly) in relation to GDP has a positive and significant impact on inflation in the WAEMU over both the short term and the long term. In fact, an increase of 1 percent in the fiscal deficit would result in a rise in inflation of between 0.67 percentage point and 0.87 percentage point over the short term and between 0.66 percentage point and 1.05 percentage point over the long term (models 2 and 3). In addition, an increase in the ratio of public spending to GDP would mean an increase in the inflation rate of 0.14 percentage point.

Domestic inflation and inflation imported from the euro area. The results show a positive link between inflation in the WAEMU and inflation imported from the euro area. Thus, a 1 percent decline in inflation in the euro area would result in a decline in inflation in the WAEMU by 0.11 percentage point over the short term and by 0.18 percentage point over the long term. Indeed, a rise in imported inflation causes an upturn in domestic inflation through imports of goods. These imported products represented on average about 32 percent in 2005–13) of goods and services in the consumer basket in the WAEMU. Furthermore, a significant number of local goods consumed by households have considerable import content. In view of this, the evolution of prices in the Union’s main trading partners in the euro area has an impact on domestic inflation. This finding is in line with the results obtained by Doe and Diallo (1997), who find that imported inflation, from France in particular, is the principal factor affecting the evolution of prices in the WAEMU over both the short term and the long term. The impulse response function indicates that domestic inflation reacts positively to a positive imported inflation shock, with a return to its long-term equilibrium after 15 months.

Certain transitory shocks, in particular the currency devaluation of 1994 (DUM94) and the shock to food prices (DUM08), significantly affected the dynamics of the inflation rate in the WAEMU.

Table 18.2 summarizes the elasticities determined using various equations based on the different models.

Table 18.2.Elasticities of Inflation with Respect to Different Variables
VariableDirection and Amount of Variation in the VariableElasticities (as %)Maximum Time Period
Short TermLong Term
Marginal lending rate1% increase−0.01−0.0515 months
Money market rate1% increase−0.4−0.2915 months
Borrowing rate1% increase−0.15−0.11
Money supply (M2)1% increase0.120.2114 months
Domestic credit1% increase0.540.3314 months
Imported inflation (IHPC_ZEuro)1% increase0.110.1815 months
Output gap1% increase0.14
Fiscal deficit/GDP1% increase0.67 to 0.870.66 to 1.05
Annex 18.1
Annex Table 18.1.Unit Root Test
LevelFirst Difference
ADF caADF thTRENDConstantADF caADF thTRENDConstantCONCLUSION
LOG (IHPC)−2.1−2.95yesyes−3.51−2.95noyesI(1)
LOG (IPC_ZEURO)−3.43−3.48yesyes−5.58−2.91noyesI(1)
LOG (M2)−3.29−3.55yesyes−3.62−2.95noyesI(1)
LOG (CRE)−2.05−2.95noyes−3.05−2.91noyesI(1)
LOG (TGPM)−1.5−2.9nono−4.9−1.9nonoI(1)
LOG(TMM)−3.84−3.55yesyesI(0)
LOG (TDEBITEUR)−3.34−3.55yesyes−4.37−3.55nonoI(1)
LOG (GAPUEMOA)−2.92−3.95nono−6.16−4.95nonoI(1)
LOG (DEFICIT)−4.95−3.55yesyesI(0)
LOG(DEP)−1.923.95yesno−10.03−3.55nonoI(1)
(*) I(0) at the 5% level; ADF ca: calculated value of the augmented Dickey-Fuller (ADF) statistic; ADF th: critical theoretical value of the ADF statistic.
(*) I(0) at the 5% level; ADF ca: calculated value of the augmented Dickey-Fuller (ADF) statistic; ADF th: critical theoretical value of the ADF statistic.

Figure 18.A1.Response of Inflation to a Shock to the Marginal Lending Rate, the Money Market Rate, the Money Supply, and Domestic Credit

Source: BCEAO estimates.

References

    AdebiyiM. A.2007. Inflation Targeting: Can We Establish a Stable and Predictable Relationship between Inflation and Monetary Policy Instruments in Nigeria and Ghana?University of LagosNigeria.

    • Search Google Scholar
    • Export Citation

    Ary TanimouneN. et TenouK.2010. « La politique monétaire dans l’UEMOA : le canal strict du crédit », présenté au cinquième Colloque BCEAOUniversités et Centres de recherche sur le thème : Quel secteur bancaire pour le financement des économies de l’UEMOA? Dakar les 25 et 26 mai 2010.

    • Search Google Scholar
    • Export Citation

    Assenmacher-WescheK. and S.Gerlach.2007. “Understanding the Link between Money Growth and Inflation in the Euro Area.” In The Travails of the Eurozone: Economic Policies Economic Development edited by DavidCobham1029. Basingstoke, United Kingdom: Palgrave Macmillan.

    • Search Google Scholar
    • Export Citation

    BarnichonR. and S. J.Peiris.2008. “Sources of Inflation in Sub-Saharan Africa.Journal of African Economies17 (5): 72946.

    CechettiG.Rita S.Chu and C.Steindel.2000. “The Unreliability of Inflation Indicators.Current Issues in Economics and Finance6 (4). Federal Reserve Bank of New York. Published in French as “Comment prévoir l’inflation.” Problèmes Economiques, No. 2683October2000.

    • Search Google Scholar
    • Export Citation

    DavidsonJ. E. H. and D. F.Hendry.1978. “Econometric Modeling of the Aggregate Time-Series Relationship between Consumers’ Expenditure and Income in the United Kingdom.Economic Journal88 (352): 66192.

    • Search Google Scholar
    • Export Citation

    Dembo ToéM.2010. Modèles de prévision de l’inflation dans les pays de l’UEMOA[Models for Inflation Forecasting in WAEMU Countries] (in French). Document d’Etude et de Recherche DER/10/03 BCEAO décembre 2010 BCEAO.

    • Search Google Scholar
    • Export Citation

    Dembo ToéM. and M.Hounkpatin.2007. Lien entre la masse monétaire et l’inflation dans les pays de l’UEMOA[Link between the money supply and inflation in WAEMU countries] (in French). Document d’Etude et de Recherche DER/07/02 BCEAO.

    • Search Google Scholar
    • Export Citation

    DoeL. et DialloM. L.1997«Déterminants empiriques de l’inflation dans les pays de l’UEMOA»BCEAO Etudes et Recherches Notes d’Information et Statistiques (476) décembre 1997.

    • Search Google Scholar
    • Export Citation

    DialloM. L. A.2003. L’impact de l’offre locale des produits vivriers sur les prix dans l’UEMOA[The impact of the local supply of foodstuffs on prices in the WAEMU] (in French). Document d’Etude et de Recherche DER/03/01 BCEAO.

    • Search Google Scholar
    • Export Citation

    DiopP. L.2000. Estimation de la production potentielle dans les pays de l’UEMOA[Estimation of Potential Output in WAEMU Countries] (in French). BCEAO.

    • Search Google Scholar
    • Export Citation

    DiopP. L. et AdobyC.1997. « Ratios simples de mesure de l’impact de la politique monétaire sur les prix »Document d’Etude et de Recherche DRS/SR/97/01 BCEAO Juin 1997.

    • Search Google Scholar
    • Export Citation

    HéricourtJ. and MateiI.2007. « Transmission de la politique monétaire dans les Pays d’Europe Centrale et Orientale: que savons-nous vraiment ? »Économie et Prévision Programme National Persée vol. 180 (4): 22138.

    • Search Google Scholar
    • Export Citation

    KovanenA.2011. “Does Money Matter for Inflation in Ghana?IMF Working Paper 11/274International Monetary FundWashington.

    MehraY. P. (1988) “The Forecast Performance of Alternative Models of Inflation.Economic ReviewSeptember/October1988.

    MishkinF.BordesC.HautcoeurP-C.Lacoue-LabartheD. et X.Ragot (2010) : « Monnaie, banque et marchés financiers »9ème édition Pearson.

    • Search Google Scholar
    • Export Citation

    NubukpoK.2001. L’impact de la variation des taux d’intérêt directeurs de la BCEAO sur l’inflation et la croissance dans l’UMOA[The impact of a variation in key interest rates of the BCEAO on inflation and growth in the WAMU] (in French). Document d’Etude et de Recherche DER/01/07 BCEAO.

    • Search Google Scholar
    • Export Citation

    NubukpoK..2003. « Dépenses publiques et croissance des économies de l’UEMOA »CIRAD Monpellier (France). pp. 128.

    OcranM. K.2007. “A Modelling of Ghana’s Inflation Experience: 1960–2003.AERC Research Paper 169African Economic Research ConsortiumNairobi, Kenya.

    • Search Google Scholar
    • Export Citation

    Other Resources Citing This Publication