(Kabbashi M. Suliman, Department of Economics - Faculty of Economic and Social Studies - University of Khartoum) - Sudan experienced economic instability
during the 1970s through the mid 1990s characterized by a high and volatile rate of inflation. The period between 1971 and 1990 is dominated by the occurrence of double-digit inflation averaging about 29%. A series of reforms were introduced during this period including devaluations, price and import controls and rationing. However, these efforts were largely unsuccessful, and triple-digit inflation persisted to the 1991–1996 period. Sudan is included by Fischer, Sahay and Végh (2002) among the 25 countries in the world that experienced episodes of very high inflation crossing 100% growth rate per annum.
The huge deprecation of the parallel market exchange rate indicates a basic disequilibrium in the foreign exchange market and trading system (see Elbadawi, 1994). It could be argued that the high inflation during the period is both a cause and result of this process. High inflation may cause the relatively fixed official exchange rate to be overvalued, which expands the size of the parallel market and the premium. In turn, the rise in the premium increases demand for money due to an increase in the domestic currency cost of foreign exchange, which feeds inflation.
More important, it has been shown that the dual exchange system can indirectly influence price movements through the quasi-fiscal deficit of the central bank operations. The direction of change depends on whether the Central Bank is a net buyer of foreign exchange, that is, collector of implicit tax or a net seller, that is, provider of implicit subsidy (see Pinto, 1991; Morris, 1995). These linkages are empirically evaluated in this paper. In 1992 the government liberalized the foreign exchange by allowing the official exchange to reflect its market value. As a result, real export reverted its declining trend over 1992–1996. However, other institutional failures remained, for example, export licensing and the Central Bank surrender requirement. The overall balance of payments continued to be in deficit; the ratio of the current account deficit to GDP averaged 6% over the period due to expansionary domestic policies; and the negative shock to the terms of trade declined by about 4%. Other factors also contributed to this external imbalance.
First, the drastic drop in gum Arabic production, which is a key export, due to reintroduction of government monopoly of export marketing in 1992 and the sharp decline of livestock export in 1993, due to an outbreak of Rift Valley Fever, (RVF).
Second, the withdrawal of IMF balance of payments support and the loss of the IMF seal of approval. The combined effects of these internal and external shocks caused inflation to grow by phenomenal records over 1988–1996, negatively affecting economic growth.
As seen in Table 1, real output growth was negative and volatile, with a recorded average decline of -3.9% and a coefficient of variation of 2.53.
The third episode of inflation development followed the reform initiated in 1997 which focused on reducing inflation; lifting the remaining imports and prices controls; fiscal consolidation; and streamlining the banking system. IMF started to monitor this programme in 1998. Because liberalization of trade and foreign exchange are essential components of the reform, the Central Bank of Sudan significantly reduced the surrender requirements, and in some cases eliminated the surrender applicable to selected exports.
The reform initiatives coincided with a substantial inflow of FDI associated with the commercial exploitation of oil and the increase of oil money as an important source of revenue. The combined effects of these developments provided an enabling macroeconomic environment. Real export including oil grew, on average, by 23.3% during the period. The growth in M2 (Table 1) declined by about 55 percentage points compared with the previous period and inflation receded to a single digit as from 2000. The real output grew by 6.8%, on average, over 1997–2002 . Fiscal consolidation was at the centre of this reform effort: the large cuts in public spending had improved the fiscal stance and significantly curbed escalation of inflation.
The preceding discussion suggests that from the 1970s through the 1980s Sudan experienced a combination of shocks relating to the upsurge of energy prices, the negative consequences of the failed FDI boom in post mid 1970s and the outbreak of the civil war in 1983, all of which negatively affected the internal and the external balances.
In particular, the fiscal deficit financing appeared as a key element determining the environment for monetary policy. Despite the introduction of Islamic financing model in 1984, the financial market remained relatively thin and characterized by entrenched guidelines for credit allocation. The effects of the continued deficit financing on price stability was a concern for the Central Bank and became a cause of worry by the early 1990s. A series of monetary reforms was introduced in 1992 and 1997 which enabled the Central Bank to engage effectively in dismantling controls, liberalizing the financial system and improving supervision and regulation of banks. Nonetheless, during most of the reform period the monetary conditions were tight; this largely reflects more the structural bottlenecks in the financial system than the efforts of the Central Bank.
In 1998 the Central Bank introduced the Central Bank Musharaka Certificate (CBMC) to streamline its long standing cost-free lending to the government and commercial banks.
The certificates are short-run liquidity papers which the Central Bank trades with banks to smooth short-term fluctuations in liquidity and hence finish a base for open market operations. A similar Government Musharaka Certificate (GMC) was introduced in 1999.
Both CBMCs and GMCs were only recently instituted and their combined size available for open market operations is too small to fully neutralize the effects of deposit changes on reserves money. Still, under the Islamic finance model, without ex ante interest rates, the Central Bank may use the monetary aggregates; change the reserve requirements; and pursue moral suasion and selective credit controls to achieve its objectives. Yet, to a great extent, the monetary policy in Sudan is based on the management of the exchange rate and the control of the monetary aggregates. Selecting an appropriate anchor within a comprehensive monetary framework relevant for the bank’s objectives of sustained output growth and low inflation presents a challenge for the policy mix in the period ahead.
This is crucial, especially because the recent rise of inflows of oil revenues may lead to money supply growth and hence inflation if the oil money is not appropriately sterilized following the developments in the real sector of the economy. Therefore, it is important to highlight the main channels through which the behaviour of the monetary aggregates affects inflation in a small open economy like that of Sudan with an underdeveloped financial sector and instruments of monetary policy.
Inflation is an important macroeconomic indicator and is widely discussed in the literature. Both demand and supply side factors including policy variables and expectations affect inflation. Numerous studies on inflation in developing countries draw from the monetarist and structuralist approaches to provide an explanation for inflation. According to the monetarist view, given stable demand for money, inflation is a purely monetary phenomenon and can be controlled by curbing excessive growth of money supply. The structuralist approach distinguishes between basic or structural inflationary pressures and the propagating mechanisms that transmit such pressures.
The identified key structural sources of inflation in these studies include distortionary government policies; foreign exchange bottlenecks; inelastic supply of food; the government budget constraint; and sectional disequilibria.
Early application of the monetarist model to explain the dynamics of inflation is found in Harberger (1963) for Chile. The monetarist hypothesis has also been tested in other less developed countries.
Many studies developed models augmenting the monetarist model with structuralist features. For example, Aghevli and Khan (1978) modelled the fiscal deficit as initiator and propagator of a cyclical process of money supply and inflation. Other studies examined the inertial and the supply side factors including cost-push elements and sectoral disequilibria
Studies on the sources of inflation in Sudan differ in their empirical models, sample period, modelled macroeconomic variables and hence their main results. However, there is a broad agreement on the following key factors affecting the rate of inflation: money growth, income growth and exchange developments.
Safi-Eldin (1976) implemented a version of the monetarist model using step-wise regression and annual data over the period 1960–1977. He points out that the observed variation in the rate of inflation is significantly explained by money supply and its lagged value, the real growth of GDP, and lagged inflation, indicating that the statistical evidence supports the hypothesis that inflation in Sudan over the period is a monetary phenomenon.
Hussain (1986) experimented with a hybrid model that includes typical monetarist variables augmented with wages and productivity as indexes measuring the cost-push elements. The model is estimated over the period 1967–1975 using annual data. The main result reveals that inflation in Sudan is of a demand-pull variety. However; as acknowledged, failure to obtain satisfactory results on the structural proxies may partly reflect data deficiencies on these proxies, and may also reflect a more wage control policy regime during that period aiming to ease rather than trigger inflation. Suliman (1989) uses the Aghevli and Khan (1978) type of model to examine the existence of a process of self-generating inflation in Sudan.
The model allows for feedback from inflation rate to the real fiscal deficit. The model is estimated by three-stage least squares method using annual data over the period 1960–1982. Significant evidence of a symbiotic relationship between government deficit, base money and inflation is found.
Abdella (1994) applies a version of the “ideal stock of money” type of model over the period 1989–1994 using annual data. He points out that, “for the period under study, the growth rate of the nominal GDP is higher than the growth rate of the money supply except for year 1992 where the nominal GDP has grown by 108% corresponding to 143% for the stock of money” (Abdella, 1994:17). The main result is that inflation in Sudan during the sample period was not a pure monetary phenomenon. Mahran and Gangi (1996) implemented a simultaneous model using two equations.
The first equation related the growth rate of inflation to the parallel market exchange rate — denoted as domestic currency in US dollars — the rate of world inflation, the rate of government borrowing from the banking system and the rate of growth of GDP.
The second equation related the parallel market exchange rate to the official exchange rate, the banks credit to the private sector, the rate of inflation, and the growth rate of the government expenditure. The model was estimated by two-stage least squares method using annual data over the period 1971–1991.
The results reveal that government borrowing from the banking sector and the imported inflation contribute significantly to domestic price growth. In addition, the continuous depreciation in the free exchange rate is the most significant single variable contributing to inflation in Sudan. Furthermore, Mahran and Gangi (1996:20) indicate: “the free exchange rate itself is significantly influenced by the expansion in credit made available to the private sector, which is mostly used for speculative purposes in the foreign exchange market.”
Abdel-Rahman (1997) uses a variant of the autoregressive distributed lag (ADL) model for money (M1), price (CPI) and real GDP, along with a set of deterministic variables—these are impulse and interaction dummies. The results of the estimation of the model over the period 1970–1994 using annual data in rates of growth indicate that nominal money correlates positively with inflation in an almost one-to-one basis, whereas real output serves to depress inflation. However, the diagnostic tests indicate the presence of the ARCH effect, and the cumulative sum of residuals shows a number of possible breakpoints. The re-estimation of the model, with interaction dummies included, in error correction form improves the fit and confirms the depressing effect of income regardless of the breaks. While nominal money was found to be insignificant before 1982, its effect has increased sharply since then.
Motivated by the need to understand the impact of the livestock export ban of 2000, due to an outbreak of RFV, on Sudanese inflation, Ramcharan (2002) applies multivariate cointegration analysis to food and non-food CPIs, M2, the first difference of the official exchange rate, livestock export revenues as an income scale variable, and the ratio of dollars to M2, as an index of currency substitution. The data is observed quarterly during 1994- 2001. The standard cointegration tests are applied to two subsets of data with the eigenvectors normalized with respect to food and non-food CPIs respectively. The
basic idea is to investigate the differential impact of the livestock export ban on these two prices separately. The observed impact depends on the information set on which livestock revenues are conditioned; in the case of non-food inflation, it would be expected to act as an income variable, exhibiting positive long-run relationship with non-food inflation.
In contrast, the reverse holds for food inflation. The intuition is: “the ensuing… decline in overall aggregate demand and the substitution away from the nonfood goods would then lower their prices, leading to large decline in inflation” (Ramcharan, 2002:9). Aside from the acknowledged missing variables bias, the results reveal that the dollarization index is not a significant determinant of food and non-food inflation in the long run while the rest of the variables are correctly signed and enter significantly in each sub-model.