Assessing the Exchange-rate pass-through to inflation in Burundi:
A RE -EXAMINATION

doi:10.21203/rs.3.rs-3777011/v1

Download PDF

Research Article

Assessing the Exchange-rate pass-through to inflation in Burundi: A RE -EXAMINATION

https://doi.org/10.21203/rs.3.rs-3777011/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study assesses the empirical determinants of inflation in Burundi over the period from January 2007 to July 2022. A single error correction model is used in which the underlying determinants of the food and non-food components of inflation as well as structural factors are incorporated into the headline inflation equation. The major contribution to the existing literature is the study of the role of the tax sector in modeling inflation, an aspect that has been neglected in existing studies on inflation in Burundi.

The empirical results reveal that the long-term sources of headline inflation are determined in the external sector market, where the exchange rate premium, world food and non-food prices determine the domestic prices as well as in the fiscal sector. In the short term, exchange rate movements, energy price adjustments, imported non-food inflation and changes in money growth as well as public sector borrowing influence headline inflation where public sector borrowing plays an important role in inflationary pressures.

Three conclusions can be drawn from these empirical results: First, the dominant influence of the exchange rate on headline inflation and its sub-components deserves special attention from the authorities, who must take consistent measures to curb the depreciation excessive BIF against the US dollar. Second, with regard to the pass-through of imported inflation, expanding and diversifying the manufacturing base to limit the current heavy reliance on imports of manufactured goods and capital goods remains a policy priority. Finally, the role of supply shocks, evident in the impact of food prices on inflation, requires immediate and significant reforms in the agricultural sector to boost productivity through the use of modern production methods.

JEL: E31, 055, C22, E62

Econometrics

Macroeconomics

Inflation

exchange rate

error correction model

fiscal sector

In recent years, the East African region has experienced an inflationary surge linked essentially to the rise in fuel and raw material prices, to the conditions adverse weather conditions, periods of drought and excessive flooding.

Indeed, inflation rose in all East African Community (EAC) countries on average, and exceeded the 8% threshold of the macroeconomic convergence criteria, except Tanzania in 2022, under the effect of the rise in food and oil prices caused by the war in Ukraine on the economy world. Inflation accelerated in the third quarter of 2022 in EAC countries. Related to previous quarter, the inflation rate stood at 19.8 against 17.4% in Burundi, at 16.4 against 12.1% in Rwanda, at 9.0 against 6.0% in Uganda, at 8.7 against 7.2% in Kenya and at 4.6 against 4.1% in Tanzania. (BRB, Monetary Policy Committee Report, 3thquarter 2022). In Burundi, episodes of high inflation were observed combined with a low GDP growth rate. Bad weather conditions; economic and financial difficulties due to the instability politics are identified as key factors in the rise in inflation. For example, imbalances macroeconomic shocks that followed the 2015 crisis, led to worsening state budget deficits and of the balance of payments, which has led to a shortage of reserve currencies and pressures inflationary.

In response, the Bank of the Republic of Burundi adopted, in 2016, a "no conventional monetary policy” aimed at supporting the economy in general and the state budget in particular. The consequence logic of such an orientation of monetary policy and external imbalance was an increase of inflation.

Indeed, in 2017; inflation reached its highest level after five years of relative price stability, supporting the arguments of the detractors of the expansionary monetary orientation of the time. However, at the same time, there was a negative supply shock, with the primary sector registering a decline of 17.3% of food production. (BRB, 2017 annual report). A year later, inflation fell to a negative level that persisted until the end of 2019 and despite the continued quantitative easing, inflation remained below the 8% target through 2020.

In an open economy, inflation is assumed to come from the monetary and external sectors acting through money demand and purchasing power parity respectively (Durevall and Ndung’u, 2001). Generally, demand and supply factors drive inflation. The factors demand-related are those that increase aggregate demand in the economy through policies expansionary monetary and fiscal policies. External factors refer to rising world food and energy prices.

Since most developing countries import a considerable amount of food, a rise in international food prices translates directly into higher food prices at the local level. Given that food products are a large part of the average household basket in these countries, an increase in domestic food prices in turn leads to a general increase in prices.

Similarly, the global energy price shock, such as the price of oil, affects domestic prices almost instantly. Internal factors generally refer to the supply-side constraints, represented by agricultural shocks. Periods of drought exert an upward pressure on food prices and therefore on the level of domestic prices. These shocks can also be captured by the seasonal pattern of agricultural production. The dry seasons are followed by a rise in inflation, while inflation is moderate during the rainy seasons.

The purpose of this paper is to empirically explore the drivers of inflation in Burundi, drawing on the existing literature (Bararuzunza, 2008; Ndikumwenayo, 2017). The inflation modeling is based on structuralist approaches, given the country's dependence on shock-prone agriculture and its vulnerability to external shocks; it also incorporates demand-side factors, namely the monetary and fiscal variables.

The empirical model includes the output gap, import prices (food, non-food, and oil), the exchange rate premium, money growth, and public sector borrowing. Overall inflation is decomposed into several components (food, non-food), as each of these can be determined by different factors. The long run and short-run effects of domestic and external factors are studied using the co-integration method.

The main contribution of this paper to the inflation literature in Burundi is the consideration of the role of the fiscal sector in modeling inflation, a factor that has been neglected in the existing literature. Fiscal theory of inflation suggests that the fiscal deficit can affect the price level through its impact on the demand for money and thus on money market equilibrium (Catão and Terrones 2005; Gordon and Leeper 2006).

The monetary expansion resulting from government borrowing to finance the budget deficit creates pressure on the price level and exacerbates the effects of other factors that drive inflation. This modeling innovation is motivated by the predominance of public sector borrowing in Burundi's financial system, and the active role of the central bank in financing government deficits through advances to the Treasury.

The results of this paper also confirm that disequilibria in the food sector have long-term effects on food and headline inflation. In addition, in the short run, inflation is driven by structural factors, including the prices of imported goods (i.e., oil and non-food items). The results also suggest that inflation shows some persistence, particularly in the case of food and non-food inflation.

The remainder of this paper is organized as follows. Section 2 provides an overview of the economic context and existing studies on inflation in Burundi. Section 3 presents a brief review of the literature on inflation. Section 4 presents stylized facts and the main characteristics and relationships between the key variables used in the regression analysis. Section 5 describes the specification of the empirical inflation model and the methodology used. The regression results are discussed in section 6, and section 7 concludes with a summary of the results and some policy implications.

This section first gives a brief description of the Burundian economy and its experience in relation to with inflation. He then outlines some of the hypotheses put forward to explain the phenomenon of inflation in Burundi.

1.1. Context and Inflation

The recurrent political crises that have been part of Burundi's history since 1993 have had serious economic consequences and have weighed on the country's economic performance. Since the 2000s, economic performance has been modest with a low average GDP growth rate and significant inflationary pressures. Inflation in Burundi has been volatile over the past two decades, linked to weather conditions on harvests, which are a key driver of price dynamics.

The first inflation spike occurred in the late 2000s, mainly due to money printing to finance the budget deficit, disruptions in agricultural production, and difficulties in supplying the market with food due to security problems in some areas. The inflation rate measured by consumer prices fell from 24.7 percent in 2000 to 9.7 percent in 2001 and then rose to 10.7 percent in 2003. (Fig. 1)The low level of inflation resulting from the restrictive measures taken by the Bank of Burundi under the IMF program in terms of credit ceilings and the increase in the central bank's key interest rate made it possible to keep inflation under control at around 8% in 2004.

The second peak occurred in 2005 with annual inflation of 13.2%, including high food inflation (17.1%). This could be due to high demand for food and election spending, in addition to food price shocks.

The third inflation spike occurred in 2008 and was mainly related to the effects of the global financial crisis and rising world food and fuel prices. The increase was in food inflation, which reached 32.7 percent (BRB, Annual Report, 2008), exacerbated by a drought that hit several regions of the country. In 2009, the inflation rate, while still high, dropped to 10.5 percent due to improved food production resulting from good rainfall and a stable exchange rate.

This was followed by a two-year period of relative stability before the fourth inflationary episode appeared in 2012. The rise in prices in 2012 involved both the food (14.1%) and non-food (23.3%) components (Fig. 2). The smaller increase in the food component compared to non-food prices is attributable to the government's decision to subside specific food products in the last two quarters of 2012, in an effort to contain the inflationary pressures of that period.

From 2013 to 2016, inflation was stable and low, possibly due to low demand resulting from political uncertainty and mass emigration to other countries. Afterwards, inflation suddenly jumped in 2017 to 16.1 percent as a result of two years of massive liquidity injections into the banking system but also an increase in food inflation of 24.1 percent induced by a 17.3 percent decline in food production in connection with the long dry season that prevailed in some parts of the country (BRB, Annual Report, 2017).

In contrast, in 2018 and 2019, the Burundian economy experienced a deflationary episode despite continued monetary quantitative easing. This decline in inflation is due to lower food inflation because of increased food production. In 2020, inflation surged again after two years of deflation, mainly due to the increase in food prices (+ 12.5 percent) and to a lesser extent non-food prices (+ 2.6 percent).The inflationary surge of 2020 due to disruptions in global supply chains continued into 2021 with an inflation rate of 8.3% following the increase in non-food inflation (9.2

In general, overall inflation tends to follow the pattern of its food component (Fig. 2). First, the good harvest of maize (the country's main cereal) and tubers, due to a favorable growing season, contributed to deflation in 2018 (BRB, annual report, 2018). Further evidence of the importance of (positive) agricultural shocks is evident in 2010 and 2014 when inflation fell largely due to good weather conditions.

Second, the droughts of 2000/2001, 2004/2005, 2008/2009, and 2016/2017 triggered higher food prices, leading to high overall inflation.

Third, since food is an important component of the consumption basket of most of the population, changes in food prices determine the overall inflation dynamics.

Finally, (negative) shocks to domestic food production lead to increased food imports. These effects can be illustrated in Fig. 3, which shows the movements of food imports and food inflation around periods of drought.

Thus, the dependence of agricultural production on rainfall and the considerable weight of food products in the CPI basket make inflation sensitive to changes in weather conditions. A large number of studies on weather dependence and its effects on inflation exist for several Sub-Saharan African countries (Diouf, 2007). This highlights the importance of supply shocks on aggregate inflation.

1.2. Inflation and exchange rate pass-through

The importance of external shocks transmitted by the exchange rate and energy (oil) prices on inflation was notable in 2008 and 2012. In 2012, for example, Burundi experienced shocks related to rising food and oil prices, which had led to inflation of 25 percent by March. Rising world oil prices, combined with the fall in world coffee prices², have exacerbated the exchange rate shocks.

The inflationary surge in 2021 (8.3 percent inflation rate), combined with declining exports (especially mining and coffee) and rising imports, widened both the trade and current account deficits. The latter has exacerbated the shortage of foreign exchange, leading to a 3% depreciation of the Burundian franc against the U.S. dollar by 2022. (African Economic Outlook (AEO), 2022).

The pass-through coefficient of the exchange rate to consumer prices in Burundi is about 0.62 (62% of domestic price changes are due to fluctuations in the official short-term exchange rate) (N. Anita et al, 2022)

From Fig. 4, it is important to note that periods of sharp depreciation of the FBU against the US dollar (2002/2003, 2007/2008, 2011/2012) are associated with higher inflation, while the decline in domestic prices, particularly between 2005 and 2006 and between 2017 and 2018, is associated with the appreciation of the FBU.

1.3. Budget Deficits, Money Supply and Inflation

In addition to fluctuations in domestic production, fiscal policy is another factor that can affect the level of domestic prices in Burundi as the government seeks to expand its tax base (increase taxes).

A budget deficit can be a source of inflation, but this depends on the duration of the deficit and how it is financed. On the one hand, a temporary budget deficit can only lead to a temporary increase in the price level, regardless of how the deficit is financed (Khieu, 2021). On the other hand, if budget deficits are permanent and financed by money creation, inflation occurs. For example, the central bank and commercial banks buy government bonds, which leads to an increase in the money supply. The main fiscal policy instruments in Burundi are public debt, tax revenues and public expenditures.

Burundi has been consistently facing budget deficits since the early 2000s. The increase in the ratio of budget deficits to GDP has gone hand in hand with rising inflation during the period 2008–2009.

Monetary financing of the budget deficit can be inflationary. In Burundi, monetary growth is driven, among other things, by the evolution of domestic credit, which in turn is influenced by the evolution of claims on the government. This leads the government to borrow from the Central Bank or to rely on tax revenues. In 2012, for example, because of the 24.1% decrease in grants, the budget deficit increased by 32.9% compared to the previous year, while the level of inflation, net claims on the state and the money supply increased by 18.2%, 27.3% and 9.6% respectively.

Although the ratio of budget deficits to gross domestic product (GDP) decreased from 6% in 2009 to 1% in 2010, it tended to increase over the period 2011–2020 (Fig. 5). This means that government spending increased during this period, which not only made the budget deficit worse, due to low tax collection capacity, but also the inflation rate as the money supply increased.

The budget deficit then narrowed to 2.9 percent of GDP in 2021 from 6 percent in 2020, as government revenues increased more than government expenditures. Foreign grants and loans, as well as domestic borrowing financed the deficit. (African Development Bank, 2021)

This brief discussion highlights the different sources of inflation in Burundi (demand, structural and supply factors) which will be formalized empirically in section 5

The dynamics of Burundi's inflation have received relatively little empirical attention. Nevertheless, a few studies have emerged in the wake of the 2008 financial crisis, relying mainly on descriptive analysis of the determinants of inflation.

This section provides a review of the empirical literature on inflation in Burundi. The review covers demand-side and supply-side factors, as well as structural factors of inflation. This discussion motivates the empirical approach developed in the next section.

a. Demand-side factors of inflation

The demand side or the supply side of the economy can drive inflation. Monetarists like Friedman (1970) state that inflation is a purely monetary phenomenon caused by an increase in the money supply that is more than proportional to the increase in the production of goods and services, which leads to an increase in the general price level.

On the demand side, the empirical literature has focused on the role of monetary factors in modeling inflation, but little attention has been devoted to the role of public credit and budget deficits on inflation in Burundi.

a.1. Monetary factors

The monetary theory of inflation stems from the quantitative theory of money of the monetarists, who consider that inflation is primarily a monetary phenomenon stimulated by an expansionary monetary policy. In the long run, inflation is driven by a surplus of money balances, which leads to an increase in the general price level.

First, the link varies depending on the empirical methodology used in the analysis. Bararuzunza (2008) assesses the real and monetary fundamentals of inflation with the ordinary least squares method and concludes that inflation in Burundi is influenced by the budget deficit and the growth rate of the money supply. N. Vénuste (2017), using an error correction model, shows that inflation is influenced by money supply growth in the short run while the relationship is not significant but positive in the long run.

a.2. Budget deficits and Inflation

The fiscal theory of inflation has been particularly important in developing countries, which have long recognized that less efficient tax revenue collection and more limited access to external borrowing tend to increase their dependence on the inflation tax (Alesina et al, 1991; Calvo et al, 1999).

Conceptually, the relationship between budget deficits and inflation stems from the government's need to finance its deficit. Fiscal-dominated governments that run persistent deficits must eventually finance these deficits through money creation (seigniorage), which produces inflation (Sargent and Wallace, 1981).

In the case of low tax collection, the government has several options for covering its budget deficit: money creation (government borrowing from the central bank) and using it to pay for the goods and services it buys (i.e., seigniorage)³ or debt raising (by issuing debt to domestic banking and non-banking institutions and/or borrowing money on international financial markets).

The existence of a unidirectional causal effect of budget deficits on inflation has been established by some SSA country-specific studies, for example studies on Nigeria (Oladipo and Akinbondola, 2011); Zimbabwe (Makochekanwa, 2008); Uganda (Kurayish et al, 2016), and in Egypt (Omneia A.Helmy, 2008).

Solomon and De Wet (2004) found a coexistence of inflation and high budget deficit for Tanzania's economy. The research established a causal link from budget deficit to inflation rate using cointegration analysis over the period 1967-2001. The study concluded, "due to the monetization of the budget deficit, significant inflationary effects are found for increases in the budget deficit."

In summary, a budget deficit can be the source of sustained inflation if it is persistent and if the government finances it by creating money (Mishkin, 2006).

b. Structural factors of inflation

The underlying dynamics of inflation are not always driven by the demand or the supply side of the market. Structural models of inflation incorporate certain features of underdeveloped economies where conventional theories of inflation do not fit. (Olivera, 1964).

The first characteristic is the dominance of the agricultural sector, which is exposed to supply shocks and to the volatility of food prices, (negative shocks, in particular droughts, are among the main causes of price increases). Second, these countries are highly dependent on export goods, which are subject to exchange rate volatility. The third characteristic is the dependence on imports of primary goods (manufactured goods and capital goods), which contributes to imported inflation.

The empirical literature shows evidence of the role of structural factors in inflation in Burundi, notably the importance of shocks to agricultural production. F. Bararuzunza (2008) emphasizes that the structural causes of inflation in Burundi are originally the result of a disequilibrium in the market for goods and services, more precisely a supply deficit resulting from insufficient national production.

c. External factors of inflation

External factors play an important role in the transmission of domestic inflation, notably through imported food and energy products and exchange rate fluctuations.

In principle, exchange rate movements (depreciation) can have direct or indirect effects on consumer prices in an open economy, depending on whether they affect prices directly or through their impact on the composition or level of aggregate demand. Changes in the exchange rate are transmitted directly to the prices of imported products, whether inputs or finished goods. By raising the cost of imported inputs, depreciation increases consumer prices because of higher production costs. Similarly, the depreciation of the domestic currency can be transmitted indirectly to domestic prices through changes in the composition of domestic demand.

Indeed, on the one hand, the rise in the price of imported products leads to an increase in the demand for substitute local products, which in turn exerts upward pressure on the price of the latter.

On the other hand, the depreciation of the local currency makes Burundian export products more competitive, and their price in foreign currency automatically decreases. As demand increases, the price of exportable local products may in turn rise, adding to the upward pressure on domestic prices already exerted by the rising price of imported goods.

As Burundi is a net importer, the increase in the price of imported goods is directly reflected in consumer prices (with the exception of so-called "strategic" products with administered prices). In addition, Burundian industries rely heavily on imported inputs in their production process, and their rising prices inflate production costs and ultimately increase domestic prices.

The article by N. BIGENDAKO et al (2017) showed that the exchange rate, as well as changes in the money supply, affected the long-run level of domestic prices in the short run. However, according to the study, the magnitude of the pass-through from the exchange rate to inflation is small.

A.NKURUNZIZA et al (2021), shows a weak and incomplete pass-through of the exchange rate to inflation in Burundi in the short run while in the long run, the effect of the exchange rate on inflation is insignificant which would suggest other determinants of inflation dynamics in Burundi.

This paper uses a monthly series covering the period from January 2007 to July 2022. The dataset comes from the BRB except for imported food prices and world crude oil prices, which come from the World Bank's primary goods price database (The Pink Sheet), and the data for the non-food goods proxy were obtained from the Asian Development Bank database.

The consumer price index, a measure of the inflation rate (pt), includes both the food (fp) and non-food (nfp) price index components presented in the Appendix. The monthly gross domestic product (GDP) series was obtained from the annual GDP data compiled by ISTEEBU. The list of variables used in the regressions and their sources are presented in the Appendix.

Before moving on to the empirical methodology, let us look at the main highlights of the variables of interest in the regression.

Money Supply Growth and Inflation

The data show that inflation is more volatile than money supply growth due to agricultural shocks. Money supply growth and non-food inflation show a downward trend from April 2013 to November 2016 that will be reversed thereafter from 2017.

Domestic prices vs. external prices

Domestic factors are not the only factors behind higher domestic prices in East Africa in general, and in Burundi in particular. Leoning (2011) and Durevall et al (2012) have shown that agricultural economies are often exposed to increases in foreign prices, which include import prices or foreign prices of tradable goods. Figure 7 depicts the dynamics of domestic food prices versus foreign food prices. It is clear that since 2007, (domestic) food prices have followed the same trend as (foreign) food prices and energy prices. All foreign prices show an increase since the end of 2007, which seems to be reflected in domestic prices one year later.

Overall (world) inflation fell significantly in 2008, the result of the fallout from the financial crisis that hit the real sector. The impact was transmitted to Burundi, followed by a slight recovery and then a decline

In inflation in the second half of 2011. It is evident from the graph that the inflation dynamics in Burundi are associated with developments in domestic food prices that respond to external shocks.

Recent studies have used a single equation error correction model (ECM) to model inflation. These include Nell (2004), Diouf (2007), Durevall et al (2013), Adam et al (2016), N. Léonce, and Janvier (2021).

The approach adopted in this study takes into account the structural specificities of the Burundian economy, notably: (1) exposure to both agricultural shocks that affect food prices and the general price level, (2) heavy dependence on imports and exposure to external shocks, particularly to food and oil prices, and finally (3) the inflationary nature of the fiscal deficit.

To model inflation, the long-run equilibrium in the monetary, fiscal and external sectors is estimated using the error correction model. First, in addition to overall inflation, this study estimates the determinants of food and non-food inflation. Second, the analysis includes dummy variables to capture the increases and decreases in food inflation that are associated with shocks to agricultural production.

The long-run equilibrium derived from each block is then incorporated into the general inflation model along with all the first-difference variables used in the long-run equations and the other variables that can explain inflation in addition to the structural variables. The general-to-specific approach is used to obtain a parsimonious model of inflation. The long-run relationships are:

$$\left({m}_{t}-{p}_{t}\right)= \alpha {y}_{t}+\beta {r}_{t}$$

${fp}_{t }=e+ {wfp}_{t}$ + c (2)

${nfp}_{t}$ = $e$ + ${p}_{t}$ + c (3)

Equation (1) represents the equilibrium in the monetary sector. The demand for money is assumed to grow through the component r that represents the interest rate on assets (foreign bonds, domestic assets in foreign currency); wfp and p represent the prices of food and non-food (imported) goods respectively and e is the exchange rate.

In an open economy, the money sector equilibrium in Eq. (1) can be extended to include the exchange rate because the depreciation of the local currency against foreign currencies can lead to a change in the holding of assets held in foreign currency. Considering this, the equation can be extended as follows:

$$\left({m}_{t}-{p}_{t}\right)= \alpha {y}_{t}+\beta {r}_{t}+\gamma e$$

The variables $\left({m}_{t}-{p}_{t}\right), {m}_{t},$ ${p}_{t}, {y}_{t}$ $,{r}_{t}$and we are real balances, money supply, and log of domestic price level, log of real income, interest rates and exchange rate

Equations (2) and (3) represent the long-run Purchasing Power Parity (PPP) equilibrium relationship for the external market for food and non-food products, respectively. According to Eq. (2), domestic prices for food goods are assumed to adjust with respect to foreign food prices and the exchange rate in the long run. Domestic non-food prices are also influenced by the exchange rate and foreign non-food prices in the long run.

For the Purchasing Power Parity (PPP) relationship to hold, the long-run relationships for equations (2) ${fp}_{\begin{array}{c}t\\ \end{array}}-e- {wfp}_{t}$ - c $=0$ and (3) ${nfp}_{\begin{array}{c}t\\ \end{array}}-e- {p}_{t}$ - c $=0$ must be stationary. This is not the case for the non-food goods sector. To do this, in order to have a stationary long-run relationship, a trend is added to Eq. (3) (Durevall and Ndung'u, 2001; Diouf, 2007; Durevall et al, 2009) and re-specified as follows:

${pnal}_{t}$ = $e$ + ${p}_{t}$ + c - τ where τ is the trend in the relative prices of non-food goods³

The fiscal theory of inflation suggests that the budget deficit can affect the price level through its impact on the demand for money and thus the equilibrium in the money market. In particular, the government's recourse to borrowing to finance its deficit can result in monetary expansion, which leads to inflation. The fiscal theory of inflation can be motivated by an intertemporal optimization model in which equilibrium inflation is related to the ratio of the budget deficit to the money stock (Catão and Terrones 2005; Gordon and Leeper 2006). In such a model, consumers maximize their utility in the presence of a set of budget constraints given by (Catão and Terrones, 2005):

$${C}_{t}+\frac{{B}_{t+1 }^{p}}{{R}_{t}^{*}}+\frac{{M}_{t+1}}{{P}_{t}}= {Y}_{t}+ {T}_{t}+{b}_{t }^{p}+\frac{{M}_{t}}{{P}_{t}}$$

Where C is consumption; B and b are the nominal and real value of risk-free bonds at a period t held by households, respectively; R* is the yield on the bonds; M is the money supply held by households; 𝑦 is income; 𝑇 is a flat tax; and P is the price level.

Households allocate their time between free time (𝑙) and purchases (s) (with 𝑙+𝑠=1), with the amount spent on purchases directly related to consumption and inversely related to the amount of real cash holdings M_(t+1)/P_t held by the household at time t and t + 1.

In each period t, the government finances its expenditures (𝐺) through taxation (T), borrowing, and seigniorage (money creation), so its budget constraint is as follows:

$$\varDelta {p}_{t}= {\alpha }_{0}+{\alpha }_{1}{ce\_m}_{t-1}+ {\alpha }_{2}{ce\_fp}_{t-1}+ {\alpha }_{3} {ce\_nfp}_{t-1}+\sum _{i=1}^{k}{\alpha }_{4i}{\varDelta p}_{t-i}+\sum _{i=1}^{k}{\alpha }_{5i}{\varDelta m}_{t-i}+ \sum _{i=1}^{k}{\alpha }_{6i}{\varDelta prem}_{t-i}+\sum _{i=1}^{k}{\alpha }_{7i}{\varDelta fpe}_{t-i}+ \sum _{i=1}^{k}{\alpha }_{8i}{D}_{t}+{\epsilon }_{t}$$

The model is estimated for three measures of inflation (Δ𝑝): food prices (fp), non-food prices (nfp), and prices in general (𝑝). The variable m1r represents real money balances; fpe represents foreign prices (fp is

imported food prices, ppi_chn is the proxy for imported non-food inflation, and pp are international oil prices), prime is the parallel market premium, and D represents the dummy variables associated with the hike (hfp) as well as the drop (dfp ) in food prices but also the main harvest season (mhse ).

ce_fp _t−1, ce_nfp_t−1 and ce_m_t−1, are the adjustment speeds (residuals of the cointegration equations) towards the long-run equilibrium for Eqs. (2), (3) and (4)

The long-run equation for the money sector is a modified version of Eq. (4) that replaces the exchange rate with the parallel market premium:

$$\left({m}_{t}-{p}_{t}\right)= \alpha {y}_{t}+\beta {r}_{t}+\gamma prem$$

The alternative model incorporating the fiscal sector is specified as follows:

$$\varDelta {p}_{t}= {\alpha }_{0}+{\alpha }_{1}{ce\_fisc}_{t-1}+ {\alpha }_{2}{ce\_fp}_{t-1}+ {\alpha }_{3} {ce\_nfp}_{t-1}+\sum _{i=1}^{k}{\alpha }_{4i}{\varDelta p}_{t-i}+\sum _{i=1}^{k}{\alpha }_{5i}{\varDelta credpub}_{t-i}+ \sum _{i=1}^{k}{\alpha }_{6i}{\varDelta prem}_{t-i}+\sum _{i=1}^{k}{\alpha }_{7i}{fpe}_{t-i}+ \sum _{i=1}^{k}{\alpha }_{8i}{D}_{t}+{\epsilon }_{t}$$

Where${ ce\_fisc}_{t-1}$ is derived from cointegration Eq. (8) for the fiscal sector; the variable credpub represents public s

a. Discussion of empirical results from the baseline model

The empirical results confirm that inflation in Burundi is determined by several long run and short-run factors (Appendix). The long-run determinants are associated with the monetary sector and the external sector (food and non-food prices), while the short-run effects are associated with the monetary sector, external and domestic non-food prices, oil prices, domestic food prices and the exchange rate premium. The results thematically reported by theme below.

Impact of monetary factors on inflation

The empirical results suggest that monetary policy and developments in the monetary sector are important determinants of inflation in Burundi in the long run. As shown by the cointegration equation (the EC term for the monetary sector), (m₁-p)-0.74*y+0.089*prem+0.46* r - 3.5, the increase in real money balances has a positive long-run impact on the price level. In the long run, a 1% increase in GDP increases the money demand by 0.74%, a 1% increase in the premium reduces the money demand by 0.089%, and an increase in interest rates (the latter being a proxy for the opportunity cost of holding cash) reduces inflation by 0.46%.

The results show that disequilibrium in the monetary sector (represented by the coefficient of the residual of the cointegration equation - CE) exert inflationary pressure on food prices and the general price level (all-items CPI), but adjustments to equilibrium appear to be relatively slow. When there is a disequilibrium in the money market, about 2 percent of it is absorbed each month by changes in domestic food prices. The speed of return to equilibrium in the non-food sector is even slower, with only 0.6 percent per month. For the overall CPI, 0.1 percent of the imbalance is corrected each month.

In the short run, money growth strongly affects prices in the food and non-food sectors, but it also has an impact on the overall price level. The effect of a shock on money growth occurs in the ninth month after the shock for food prices and in the tenth month after the shock for non-food prices. As for the magnitude of the short-run response of prices to an increase in money growth, the econometric results suggest that a 1 percent increase in money growth increases food prices by 0.078 percent and non-food prices by 0.025 percent.

This response may seem small, but given the large increases in the money supply, particularly in recent years, the impact on prices is indeed non-negligible. These results are consistent with those found by Bararuzunza (2008) and N. Venuste (2017) who showed that inflation is determined by the growth of the money supply in the short run.

• Impact of foreign prices (food and non-food sector) on inflation

In the long run, international food prices have a significant effect on domestic food prices as indicated by the estimated cointegration equation. This highlights Burundi's integration into international food markets and, in particular, the country's dependence on food imports. As shown in the cointegration equation for food prices (e + wfp) - 0.93*fp+ 7.63 , in the long run, a 1 percent increase in international food prices increases domestic food prices by 0.93 percent.

In table B (see appendix), the coefficient of the long-run food equation in the food inflation equation indicates that about 3.3 percent of the deviations in equilibrium food prices between the world and domestic markets are absorbed each month in the domestic food market.

The speed of adjustment is 1.4 percent for the general CPI. International food prices have short-term effects on domestic food prices, although they are not significant. A 1% increase in (imported) food prices leads to a 0.04% increase in domestic food prices in the short run.

The producer price index for Chinese manufactured goods, used as a proxy for the price of Burundi's non-food imports, turns out to be a long-term determinant of Burundi's non-food prices. As shown in the cointegration equation for the price of non-food products (e+ ppi_chn) - 0.76*nfp + 5.53 , holding the exchange rate constant, a 1 percent increase in the price of Chinese manufactured goods increases domestic non-food prices by 0.76 percent.

This effect is as large as the effect of world food prices on domestic food prices cited above.As shown by the coefficients of the EC term in Table B (Appendix), the results suggest that 0.2 percent of the deviations from equilibrium in the non-food market are corrected each month.

• Exchange rate effects on inflation

The official exchange rate is controlled by the government and its value against foreign currencies such as the U.S. dollar remains fixed for relatively long periods; in other words, the official exchange rate is not market-determined. The parallel exchange market is not official. Therefore, the parallel exchange rate does not fully reflect a freely functioning foreign exchange market. The advantage of using the premium is that, being a linear combination of the two rates, it represents the developments in both exchange markets.

Changes in the exchange rate premium have statistically significant effects on food prices, non-food prices, and the overall CPI. Empirical results indicate that when the premium increases by 10 percent (e.g., due to an overvaluation of the official exchange rate), overall inflation increases by 0.04 percent, food inflation by 0.12 percent, and non-food inflation by 0.4 percent.

In addition to using the parallel market premium as a determinant of money demand, the premium also affects individual prices indirectly through the exchange rate used to convert foreign prices into domestic prices. The retail prices of fuel and other products on the so-called "priority" list for foreign exchange allocation depend on the official exchange rate, while the prices of other products not on the list are likely to be influenced by the parallel market rate.

Although oil prices are regulated by the government because of their impact on the overall economy through production, higher energy prices lead to higher prices for other products, particularly food, which ultimately drives up overall inflation.

• Effects of inflation inertia

The results of the econometric analysis indicate that inflation has significant persistence, with past changes in domestic prices affecting current prices with different lags. In the case of food prices, 10 percent of the current price is explained by the price change eight months earlier. It takes six months for a change in non-food prices to materialize in the domestic price of non-food products. This delay is not surprising-such changes can be driven by shocks in international markets, and their transmission to the domestic economy can be slowed by policy interventions or other factors⁵.

For the total CPI, lagged effects are tested by including lags of its components, and the results show that lagged food and nonfood price inflation have a positive and significant effect on total inflation. A 1% monthly increase in food inflation leads to a 0.1% monthly increase in the overall price level nine months later and a 1% monthly increase in non-food inflation leads to a 0.38% monthly increase in the overall CPI seven months later.

b. Empirical results with the alternative model (including the fiscal sector)

The empirical results of the alternative inflation model incorporating the fiscal sector in place of the monetary sector are presented in Table C (see Appendix). The results indicate that inflation is significantly affected by the disequilibrium in the fiscal sector in the long run and by increases in public sector borrowing in the short run.

As indicated by the estimated cointegrating equation for the fiscal sector (m1-p)- 1.04*credpub-0.35*def, in the long run, a 1 percent increase in the deficit increases real balances by 1.04 percent, and a 1 percent increase in public sector credit increases real balances by 0.35 percent. Compared to the results obtained for the monetary sector in the baseline model (table B), disequilibrium in the fiscal sector is absorbed more slowly over time except for the aggregate CPI. In the case of food prices, 0.3 percent of the imbalance in the fiscal sector is eliminated each month, compared to 2 percent for imbalances in the monetary sector. The rate of adjustment is even slower for non-food prices (0.1% versus 0.6%).

In the short run, changes in public sector borrowing affect all components of inflation. An increase in public sector borrowing leads to an acceleration of food, non-food, and total inflation with a lag of five, seven, and one months, respectively.A one percent increase in public sector borrowing leads to a 0.06 percent increase in food price inflation five months later and 0.07 percent six months later, with a total effect of 0.13 percent after six months. The corresponding effects on nonfood inflation are 0.02 percent after seven months and 0.03 percent after ten months, with a total effect of 0.05 percent. Growth in public sector borrowing has a positive and significant effect on headline inflation with a lag of one and six months, with a total effect of 0.03%.

On the domestic side, food and non-food inflation shows some persistence as shown by the positive and significant coefficients in the short run (Table C, Appendix).In the case of food prices, 10% of the current price is explained by the price change of 2.3% and 1.8% eight months and twelve months earlier. It takes six months for a non-food price change to materialize in the domestic non-food price. A 1 percent increase in nonfood prices results in a 0.18 percent increase after 6 months.

Overall, the results of the inflation models including the fiscal sector confirm the existence of statistically significant long-run effects of imbalances in the fiscal sector on all measures of inflation, with the exception of non-food prices. In addition, government borrowing has short-run inflationary effects on all measures of inflation. This implies that fiscal imbalances and the way in which fiscal deficits are financed must be taken into account in understanding the determinants of inflation in Burundi⁶.

This study examined the fundamental factors driving inflation in Burundi over the period from January 2007 to July 2022, including monetary, external and fiscal factors. A single error correction model in which the determinants of food and non-food components of inflation were incorporated into the overall inflation equation.

The empirical results showed that the long-run sources of headline inflation are determined in the external sector, where the exchange rate premium and world food and non-food prices influence domestic prices in Burundi. In the short run, exchange rate movements, imported inflation (non-food items) as well as domestic factors (food, non-food inflation and real balances) influence overall inflation.

The results of this study underscore the importance of structural supply-side factors, including developments in the food sector, in understanding inflation dynamics in Burundi. In the short run, shocks to food production, such as droughts, lead to higher food prices, which in turn affect overall inflation. In the long run, imbalances in relative food prices in Burundi compared to international prices lead to inflation in the country. These results reflect the large share of food in household consumption, the exposure of production to climate shocks, and the dependence on imported products.

Unlike some previous studies (N. Vénuste, 2017), this paper finds significant long-run effects of imbalances in the monetary sector, with a relatively slow adjustment mechanism. The results suggest that monetary factors have both short- and long-run effects on inflation in Burundi. The fiscal sector, which has not been considered in previous studies, has both a long run and short-run impact on inflation in Burundi. In the long run, imbalances in the fiscal sector are absorbed comparatively more slowly than adjustments in the monetary sector. Public sector borrowing has strong short-run effects on all measures of inflation. The results also suggest that there is some inertia in inflation, particularly in the case of food and non-food inflation. Inertia appears strong and significant Persistence appears strong and significant. The latter is probably due to agents' anticipations of price changes after major agricultural shocks, such as drought or sharp increases in food prices on the international market, or to currency shortages or a problem with the supply of imported products.

Overall, non-food inflation appears to behave differently from other measures, with a relatively slow speed of adjustment, which explains the difficulty in unpacking its determinants, as illustrated by the (small) number of significant determinants and the low R-squared. These results underline the importance of a disaggregated approach to inflation modeling, which allows us to identify the behavior of the fundamental characteristics of each sub-component.

The results of the study confirm that the Burundian economy is highly integrated with the global economy. Its prices are strongly influenced by relative prices in the food and non-food sectors in the long run, and changes in world non-food and oil prices in the short run that are transmitted to domestic prices. In addition, the parallel market exchange premium has a positive effect on food, non-food, and headline inflation. Since the exchange rate premium incorporates the official exchange rate and the parallel market rate, the results imply that exchange rate policy that keeps the rate outside the market rate has an impact on inflation.

Indeed, because of the small size of its market relative to the world market, Burundi has practically no influence on the prices of imported goods, both inputs and finished goods. Nevertheless, the short-term effects of exchange rate changes and the speed with which they are transmitted to consumer prices depend on several factors. The share of imported goods in the consumer basket is likely to remain the most important determinant. This underscores the need for policy actions by the authorities to curb excessive depreciation of the local currency.

In addition, exchange rate stabilization requires large foreign exchange reserves to intervene in the foreign exchange market. The problem is that the Central Bank does not have sufficient reserves.

The imbalances in the trade balance, due to the high prices of fuel and consumer and capital goods, have caused the level of reserves to fall to 1.6 months of imports at the end of June 2022, compared with 2.2 months at the end of 2021 (IMF Press Release No. 22/328, 2022).

The role of supply shocks, evident in the impact of food prices on inflation, suggests the need for significant reforms in the agricultural sector to boost productivity through the use of modern techniques such as irrigation to reduce dependence on weather conditions. In addition, monetary policy must become more effective. A better functioning financial sector and more effective control of the money supply could lead to better management of price shocks.

A. Honoré, B. Edouard Normand et N. Ephrem.2017. Impact des chocs exogènes sur l’efficacité de la politique monétaire au Burundi. BRB, REM Vol 01, n°01
A. Kabundi. 2012. Dynamics of Inflation in Uganda. African Development Bank Working Paper
A. NKURUNZIZA et A. Ntwari.2022. Analyse du pass-through du taux de change sur l’inflation au Burundi. BRB, REM Volume 04.
Alesina, A. Drazen, A. (1991). Why are stabilizations delayed?. The American Economic Review, 81(5), 1170-1188.
BARARUZUNZA, F. 2008.Fondements réels et monétaires de l’inflation au Burundi, Ridec, Volume 3 n°1.
BRB, Rapport Annuel, 2008,2013 et 2017.
BRB, Rapport de Stabilité financière, 2019.
BRB, Rapport du Comité de Politique monétaire N°01, 2013-2022.
Calvo, G., and C. Végh. 1999. Inflation stabilization and bop crises in developing countries. In Taylor, J., and M. Woodford (Eds.). Handbook of Macroeconomics, Vol. C., pp. 1531–1614. Amsterdam: North-Holland.
Catão, L.A.V., et M.E. Terrones. 2005. Fiscal deficits and Inflation. Journal of Monetary Economics.52 (3):529-54
Dureval, D., J.L.Loening et A.Y.Birru.2013.Inflation dynamics and food prices in Ethiopia. Journal of Development Economics 104 (September):89-106.
Gordon, D.B et E.M.Leeper.2006.The price level, the quantity theory of money, and the fiscal theory of the price level. Scottish Journal of Political Economy 53(1): 4-27.
Helmy, Omneia (2008). The Impact of Budget Deficit on Inflation in Egypt. The Egyptian Center for Economic studies Working Paper No. 141, pp. 1-32.
J. Mawejje & M. Mayanja Lwanga. 2016. Inflation dynamics and agricultural supply shocks in Uganda. African Journal of Economic and Management Studies, Emerald Group Publishing Limited, vol. 7(4), pages 547-567, December.
J.H. Olivera. 1964. On Structural Inflation and Latin America Structuralism. Oxford Economic Papers, Volume 16, Issue 3, Pages 321-332.
Khieu, H.V. (2021).Budget deficits, money growth and inflation: empirical evidence from Vietnam. Fulbright Review of Economics and Policy, Vol. 1 No. 1, pp. 61-78.
Kurayish Ssebulime & Bbaale Edward, 2019. Budget deficit and inflation nexus in Uganda 1980–2016: a cointegration and error correction modeling approach. Journal of Economic Structures, Springer;Pan-Pacific Association of Input-Output Studies (PAPAIOS), vol. 8(1), pages 1-14, December.
Léonce.N, Janvier Désiré.N. 2021. Monetary, Fiscal, and structural drivers of inflation in Ethiopia. World Bank, Policy Research Working Paper 9881.
M.Chipili. 2021.Inflation dynamics in Zambia. AERC Research paper 484, Bank of Zambia
M.Diouf, 2007. Modeling Inflation for Mali. IMF Working Papers 2007/295, International Monetary Fund.
Makochekanwa, A. (2008). Impact of budget deficit on inflation in Zimbabwe. MPRA Paper.
Mishkin, Frederic S. 2006. The economics of money, banking, and financial markets. Seventh edition update, Pearson International Edition.
N. Vénuste. 2017. Les causes de l’inflation au Burundi. BRB, REM Vol 01 n°01.
Oladipo, S., & Akinbobola, J. (2011). Budget deficit and inflation in Nigeria: A causal relationship, Journal of Emerging Trends in Economics and Management Sciences, 2(1), 1-8.
Sargent, Thomas J., and Neil Wallace. 1987. Inflation and the Government Budget Constraint. In Economic Policy in Theory and Practice, 170–207.
Solomon, M and W.A. de Wet. (2004). The Effect of a Budget Deficit on Inflation: The Case of Tanzania.South African Journal of Economic and Management (SAJEMS), 7(1).

The coffee price volatility on the international market affects foreign exchange reserves to the extent that coffee represents a significant share of Burundi's exports despite its decline in recent years. According to foreign trade statistics, on average annually, coffee accounted for 40% of export earnings between 2012 and 2016 (BRB, 2017).
Seigniorage generally refers to the income earned from money creation. For central banks, seigniorage income can be defined as the flow of interest from assets held against money in circulation (or, more generally, from the money base).
The real effective exchange rate was used instead of the official exchange rate because it measures the evolution of a country's exchange rate vis-à-vis its trading partners. This indicator allows a broad interpretation of a country's price competitiveness. To obtain the trend, the HP (Hodrick-Prescott) filter was used to remove the non-stationary component of the real exchange rate.
Between May 2012 and March 2013, the Government exempted 13 food products from tax, which contributed, to a slightly extend, to a decrease in inflation during this period (ministerial order n°540/678 of 15/05/2012).
In order to adjust the impact of high inflation on the economy and maintain macroeconomic stability, the Bank of the Republic of Burundi (BRB) has also responded appropriately to mitigate inflation. In 2012, the BRB tightened its monetary policy by slightly increasing its key interest rate, thereby bringing down apparent inflation from 25 percent in March-April to 11.8 percent by the end of the year. (Word Bank Document, "Burundi Poverty Assessment," November 2016)
The increase in the fiscal deficit led to an increased use of the domestic debt in 2016.This increased from 21.9% of GDP in 2015 to 30% of GDP in 2016 (World Bank, 2018).
In turn, the rise in domestic debt materialized in a sharp increase in the volume of the state-owned bonds in the total assets of banks with an increase of 73% in 2016 and 47% in 2017 while credit to the economy remained very low with an increase of only 3% in 2017 and a negative growth two years before (BRB, 2018). This led to the crowding out effect vis-à-vis the private credit sector. The increase in monetary financing of the fiscal deficit therefore contributed to inflationary pressures.

The authors declare no competing interests.

appendix.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Assessing the Exchange-rate pass-through to inflation in Burundi: A RE -EXAMINATION

Status:

Version 1

Abstract

Figures

1. Introduction

2. Brief description of inflation in Burundi

1.1. Context and Inflation

1.2. Inflation and exchange rate pass-through

1.3. Budget Deficits, Money Supply and Inflation

3. Related literature on inflation

a. Demand-side factors of inflation

a.1. Monetary factors

a.2. Budget deficits and Inflation

b. Structural factors of inflation

c. External factors of inflation

4. Data sources and stylized facts from the data

5. Modeling inflation in Burundi

6. Discussion of empirical results

7. Conclusion

References

Footnotes

Additional Declarations

Supplementary Files

Status:

Version 1