Democratic Republic of the Congo
Selected Issues and Statistical Appendix
Author: International Monetary Fund
Publication Date:
14 Oct 2005
eISBN:
9781451808384
ISBN:
9781451808384
Language:
English
Keywords:
ISCR; CR; wage bill; foreign currency; broad money; monetary policy; pay scale; government Economic Program; Congolese authorities; authorities accommodation; narrow money; real GDP
Download PDF (4.5 MB)

This Selected Issues paper analyzes recent developments in the financial system of the Democratic Republic of the Congo (DRC). It describes the monetary developments in 2003–05 and reform of the financial system. The paper examines the impact of monetary and exchange rate policies on inflation, including the nature of the lags involved. It reviews developments in the DRC’s mining sector and mining’s contribution to the economy from a fiscal perspective, showing a low effective level of taxation. The paper also discusses restructuring of the mining sector.

Abstract

This Selected Issues paper analyzes recent developments in the financial system of the Democratic Republic of the Congo (DRC). It describes the monetary developments in 2003–05 and reform of the financial system. The paper examines the impact of monetary and exchange rate policies on inflation, including the nature of the lags involved. It reviews developments in the DRC’s mining sector and mining’s contribution to the economy from a fiscal perspective, showing a low effective level of taxation. The paper also discusses restructuring of the mining sector.

I. Recent Development in the Financial System1

A. Background

1. Over the past few years, the authorities started addressing many of the weaknesses facing the banking system of the Democratic Republic of the Congo (DRC) in order to deepen financial intermediation. These weaknesses include a lack of confidence in banks owing to their financial fragility and the recent experience with the nonfungibility between bank money and currency; high and rising dollarization, with dollar deposits accounting for more than 85 percent of bank deposits; the lack of financial infrastructure (for example, the payment system, telecommunications, and transportation); shortcomings in the judicial and legal framework; and the tax administration’s frequent garnishments of bank accounts. In addition, the independence of the Central Bank of the Congo (BCC) has been undermined by large operational losses and weak institutional and operational capacities.

2. With technical assistance from the Fund and bilateral donors, and financial support from the World Bank, the authorities and the BCC have focused on establishing the legal framework for the financial system, strengthening the BCC’s organization, and restructuring commercial banks. The reform of the BCC has consisted in (i) improving its capacity to conduct monetary operations and liquidity management, (ii) enhancing foreign exchange policy and reserve management, (iii) strengthening its accounting and information system, (iv) improving banking supervision, and (v) setting up a mechanism to combat money laundering and the financing of terrorism. The reform program at the central bank has led to the drawing up of an action plan, which is a key aspect of the Government Economic Program (PEG) supported by the IMF with an arrangement under the Poverty Reduction and Growth Facility (PRGF). In addition, consistent with these reforms, the BCC published, in August 2003 a broad strategy for its development and for that of the financial system.

3. During 2001 and 2002, the authorities adopted a legal framework for the financial system.2 Key measures included the adoption of (i) a central bank law that provides for its independence; (ii) a banking law that gives the BCC full responsibility for the supervision of the financial sector and establishes frameworks for bank licensing and liquidation and for the operations of credit unions; and (iii) a new legal framework for the restructuring of the banking system. In addition, for the first time in many years, external auditors conducted an audit of the BCC accounts for 2000-01.

4. The BCC also put in place the key components of a framework to strengthen monetary management. In particular, it designed a new system to improve the monitoring of the main items of its balance sheet and the different sources of money creation—thereby strengthening its capacity to forecast the autonomous factors affecting liquidity. To identify and contain the impact of central bank losses on inflation, the BCC consolidated the net treasury deficit in the government’s net indebtedness to the central bank. And, in September 2002, the BCC created the Consultative Group on Monetary Policy, comprising members from the BCC and the Ministry of Finance, to facilitate the coordination of fiscal and monetary policies. In addition, with a view to developing a market-based approach to monetary policy, the BCC adopted a more flexible interest rate policy, which reflects price developments, and unified the multiple exchange rates in the foreign exchange market by floating the Congo franc, beginning in May 2001.

5. The strengthening of monetary management has allowed the BCC to pursue a monetary policy designed to achieve price stability within the framework of a floating exchange system. The measures it has implemented since 2001 have been essential in breaking the vicious circle of depreciation and hyperinflation, restoring macroeconomic stability, and accelerating growth.

6. The rest of this chapter analyzes monetary developments in 2003-05, and describes the reform of the financial system.

B. Monetary and Credit Developments in 2003-05

7. Monetary and credit developments since 2002 have been marked by a rapid growth in base and broad money, including dollar deposits, which until mid-2004, was accompanied by a decline in inflation and relative stability of the exchange rate of the Congo franc against the U.S. dollar (Tables I.1 and I.2). In the second half of 2004, however, further money expansion accompanied by increased political and security tensions led to a sharp depreciation in the value of the Congo franc.

Table I.1.

Democratic Republic of the Congo: Monetary Survey, 2001-05

article image
Sources: Congolese authorities; and IMF staff estimates and projections.
Table II.2.

Democratic Republic of the Congo: Accounts of the Central Bank of the Congo, 2001-05

article image
Sources: Congolese authorities; and IMF staff estimates and projections.

8. In 2003, broad money grew by 32 percent, or much more than nominal GDP. This reflected a strengthening of the demand for Congo francs in the reunified provinces, a resumption of economic growth, and newly achieved macroeconomic and political stability, in particular after reunification in mid-2003. The pickup in money demand led to a decline in income velocity from 20.4 at end-2002 to 18.5 at end-2003. The improved macroeconomic situation also provided room for the central bank to reduce its refinancing rate, which had been raised from 12 to 27 percent in January 2003 to dampen inflationary pressures at end-2002. By end-2003, the rate had been cut to 8 percent. The commercial bank lending rates fell in line with the refinancing rate but remained at about 25 percent at end-2003.

9. In spite of the improvement in the macroeconomic situation, the dollarization of the economy increased in 2003, with dollar deposits growing slightly faster than currency in circulation and their share in total deposits rising from 81 percent to 85 percent. This reflected the still-limited confidence in the Congo franc and the banking system and, to some extent, a lack of large-denomination currency notes, even though the BCC started issuing CGF 200 and CGF 500 (US$1.30) notes to supplement its CGF 100 note. Furthermore, the BCC was using currency rationing as a tool of monetary policy (see next section). Hence, dollar currency notes and deposits continued to be used for relatively large transactions.

10. The main factor contributing to the growth in broad money was the increase in net central bank credit to the government, which, relative to beginning-of-period broad money, grew twice as fast as currency in circulation, in part because the government was covering the BCC’s cash-operating losses. Bank credit to the private sector grew by close to 50 percent in 2003 but had a limited impact on broad money growth given the very low base. More than 80 percent of bank credit to the private sector is denominated in U.S. dollars.

11. In 2004, broad money expanded by 73 percent, with foreign currency deposits increasing at a faster rate than currency in circulation. As a result, the share of dollar deposits in total deposits rose to 86 percent. The expansion reflects an improvement in the net foreign assets of the banking sector (measured at constant exchange rate) and increases in net bank credit to the private sector and in other items net of the central bank (excluding revaluation accounts). The latter was due to the payments of central bank domestic debt and movements in suspense and liaison accounts. However, given the depreciation of the Congo franc during the year and the large net foreign liabilities of the central bank, the valuation of foreign currency accounts at end-of-period exchange rate resulted in a deterioration in the net foreign assets of the banking system and in a large increase in the banking system’s other items net.

12. While net bank credit to the government declined slightly in 2004, the fall was much smaller than had been expected. This was because part of the foreign financing of the budget was spent domestically rather than being used to reduce net credit to the government and increase the BCC’s international reserves as anticipated. This contributed to the-faster-than anticipated increase in base money.

13. In spite of the rapid increase in broad money in the first half of 2004, the BCC did not tighten monetary policy immediately. Rather it reduced the refinancing rate to 6 percent in August 2004, largely because the exchange rate was stable and inflation was below 5 percent until June 2004. Nevertheless, with security and political tensions rising, especially in the eastern provinces, government spending started rising and market confidence in government policies weakened. As a result, the BCC was unable to prevent the Congo franc from depreciating and inflation from rising despite successive increases in interest rates to 14 percent by end-2004 and limited sales of foreign exchange. Furthermore, the central bank measures were made somewhat ineffective as the government increased spending further toward the end of the year, partly in response to rising security threats on the eastern border. Hence, in the second half of the year, the Congo franc depreciated by 12 percent and 12-month inflation rose from 4 percent to 9 percent.

14. These trends continued in the first months of 2005, as the Congo franc weakened by an additional 15 percent in the first four months of the year and 12-month inflation rose to a peak of 27 percent in May. Nevertheless, the situation improved anew in June when the currency appreciated substantially and the consumer price index dropped by 7 percent. The improvement resulted from a tightening of monetary policy, including an increase in the financing rate to 65 percent, and the government improved control over expenditures, which helped the central bank contain the growth in base money.

C. Reform of the Banking System in 2003-04

15. The banking system comprises the Central Bank of the Congo, nine commercial banks—including five implementing restructuring plans agreed with the BCC—and nine commercial banks being liquidated. The nonbank financial sector consists of five public financial institutions3 and numerous credit cooperatives and microfinance institutions. The BCC intends to assess the importance of microfinance institutions in the economy and establish the legal framework for their supervision.

Reform of the central bank

Monetary operations and liquidity management

16. The banking system in the DRC is still at an early stage of development. The conduct of monetary policy is severely hampered by the low level of financial intermediation in domestic currency and the high level of dollarization, weak monetary programming tools, and a dearth of policy instruments.

17. Commercial banks still mostly act as financial agents of the government for which they receive tax revenue and effect payments. They mostly receive foreign currency deposits, from the nongovernment sector, which they lend out in dollars or keep with correspondent banks abroad. Hence, commercial banks’ contribution to money creation is small as they extend a limited amount of credit and rarely use the central bank refinancing window. As a result, the sources of broad money growth are central bank financing of the overall fiscal deficit, including the financing of the deficit of the central bank, and the accumulation of net foreign assets by the banking system.

18. Until 2002, to contain money growth and stabilize the macroeconomic situation, the central bank used two main instruments: foreign exchange sales and, at times, currency rationing achieved by staggering the payments of government checks based on the availability of currency at the central bank. As a result, banks had controlled access to their excess reserves at the BCC, and Congo franc bank money traded at a discount (“décote”) vis-à-vis Congo franc currency. The discount was a major impediment to bank intermediation and an incentive for dollarization. However, the practice has largely stopped since end-2002.

19. To improve its monetary policy framework and overcome the difficulties described above, the BCC took measures to improve its liquidity forecasting and management capacity. It started by strengthening the programming of currency issues and streamlining its instruments for conducting monetary policy.

Technical assistance and the reform of the banking system

Since 2001, the Fund has assisted the BCC in strengthening its institutional and operational capacity by providing extensive short-term and long-term technical assistance. The principal areas of assistance are: (i) monetary operations and liquidity management; (ii) foreign exchange operations; (iii) accounting, internal audit, and information systems; (iv) banking supervision; and (v) combating money laundering and the financing of terrorism.

The central bank also received bilateral assistance to (i) strengthen its organizational structure and human resource management, and (ii) prepare its recapitalization.

The monetary authorities are working to ensure transparency in their operations and functioning, in conformity with the IMF Code of Good Practices on Transparency in Monetary and Financial Policies.

20. The programming of currency issues has largely involved better coordinating of how the government cash-flow plan and the BCC’s liquidity forecasting exercise are executed. To that effect, the BCC has progressively improved the monitoring of currency creation resulting from its operations, and those of the treasury and commercial banks. Nevertheless, the central bank and the Ministry of Finance must do more to improve their ability to forecast and program currency issues resulting from government operations.

21. Regarding monetary instruments, the BCC introduced central bank bills with maturities of 7 to 28 days in December 2002 to absorb excess liquidity in the banking system (and avoid unsustainable foreign exchange sales and recourse to currency rationing). The BCC fixes the interest rates on the bills each week, largely to ensure that the real interest rate, taking into account recent trends in inflation, remains positive. Banks then inform the central bank of the amount and the maturity of the bills they will purchase.

22. Banks, however, have been reluctant to increase their claims on the central bank, beyond their holding of excess reserves, in part because of the difficulties of converting their central bank deposits into currency. In particular, they have not sought to attract nonbank deposits to invest in central bank bills. Also, nonbanks (mostly large enterprises) have until recently not bought BCC bills directly partly because of their high denomination and a lack of information.4 Consequently, the interest rate differentials between time deposits and central bank bills have, at times, widened sharply, as they did in March 2005, when banks were slow to raise interest rates on deposits and loans following the sharp increase in interest rates announced by the central bank.

23. The commercial banks’ limited response to the central bank’s interest rate policy represents a major impediment to efforts to contain currency growth. To address this issue and improve the effectiveness of interest rate policy, the central bank announced in May 2005 that bank deposits were fully fungible. In addition, the central bank is encouraging enterprises to invest surplus funds in central bank bills. For those measures to be effective, however, the central bank must gradually regain its independence and have a large enough stock of currency notes to meet commercial bank demand for liquefaction at all times.

A01fig01

Outstanding stock of bills and interest rate on 7-day bills, 2003-05

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

24. The BCC’s lack of high denomination currency notes has held back demand for Congo francs, and the high cost of printing notes has kept the BCC’s net seignoriage low, thus contributing to its large operating losses. Over the coming months, the BCC plans to issue higher denomination currency notes progressively. To dispel any notion that it is easing monetary policy—which happened in the past when new notes were issued during periods of increasing or hyperinflation—the BCC will first launch an information campaign. The issuance of high-denomination notes will help reduce the dollarization of the economy in two ways. First, it will facilitate large payments to be effected in Congo francs. Second, with bank deposits being fungible, enterprises will be more willing to use bank checks.

25. Nevertheless, for bank intermediation to increase and dollarization to fall, confidence in the banking system and central bank policy will have to improve markedly. This will necessitate further improvement in the operations of the central bank, including the following: (i) greater transparency in the conduct of monetary policy so as to increase its credibility; (ii) a strengthening of the payments system to ensure the supply of effective payment services in Congo francs throughout the country; (iii) a strengthening of the financial situation of commercial banks; (iv) limits on the discretionary powers of the tax administration on taxpayers’ bank deposits; and (v) a judicial and legal framework favorable to banking activity and to the enforcement of property rights and contracts. However, given that progress in those areas is likely to be slow, dollarization will probably remain high over the medium term.

26. The central bank’s lack of financial autonomy is also a major impediment to its monetary operations. The BCC has recorded losses for many years, and in 2003, its operating loss amounted to CGF 49 billion, including a cash loss of CGF 18.4 billion, with the remainder due to the reevaluation of foreign exchange accounts and provisions and amortization. At end-2004, the central bank’s net worth is estimated at minus CGF 408 billion (US$900 million). Besides the reevaluation of foreign exchange accounts, the main reasons for the central bank’s losses are a low level of earning assets, high personnel and administrative costs, and limited net seignoriage because of the low denomination of currency notes and high production costs.

27. The central bank is studying ways to restructure its balance sheet and rebuild its capital base with the assistance of the National Bank of Belgium. It is also studying measures to streamline and strengthen its internal organization, including by meeting the staffing and training needs required to implement its ambitious reform plans. The key issue is to reduce the central bank cash operating losses at a minimum cost to the government budget. To that end, the BCC must first implement a strict cash-flow plan, which will require making its operations more transparent and strengthening its business practices, including by adhering to transparent bidding procedures and better preparing and monitoring its projects.

Foreign exchange operations and reserves management

28. The DRC adopted a floating foreign exchange system in 2001; the objective of foreign exchange operations has been to meet the international reserves target set in the Government Economic Program. Although intervention should be limited to smoothing out short-term movements, the authorities have intervened over the past two years to ease pressures on the exchange rate and acquire currency notes to make payments for the government without issuing new currency. As a result, the BCC, while trying to meet its inflation and money growth targets—in the face of excess expansion of net credit to the government and other items net—allowed the net foreign asset target to be missed (Table 1.3).

Table I.3.

Democratic Republic of the Congo: Foreign exchange Rate Intervention

(In millions of Congo francs)

article image
Source: Congolese authorities.

29. The BCC used several variations of Dutch auctions to sell foreign exchange to commercial banks. To some extent, those variations aimed at achieving specific monetary policy objectives rather than to improve the functioning of the foreign exchange market. Until recently, no mechanism existed through which the central bank could purchase foreign currency, and the few times it did so, the exchange rate was agreed on a bilateral basis and set close to the selling rate for that day. Repeated changes in the mechanism used for selling foreign exchange as well as the lack of a transparent regulatory framework added to the volatility of the exchange rate.

30. To improve the situation, the central bank issued a new circular in February 2005 that clarified the main feature of the Dutch auction that it would use to sell foreign exchange. In particular, banks can now purchase foreign currency with cash or with a bank check in Kinshasa or at the branches of the central bank in the provinces;5 and no one bank can buy more than 40 percent of the amount to be sold. Similar guidelines for the BCC’s purchase of foreign exchange were issued in June 2005. The BCC is also working on strengthening the departments responsible for managing foreign exchange reserves, including through the creation of a computerized trading room and back offices and the drafting of a manual of procedures. To facilitate the management of international reserves, the BCC has reduced its foreign exchange deposits with local commercial banks, given that the cost of transactions is high and such deposits are generally not remunerated.

Accounting, information system, and internal audit

31. Taking into account recommendations from the Fund safeguards mission conducted in October 2002, the BCC has started an ambitious program to raise its accounting system to international standards. The objective is to have the BCC’s accounts approved by external auditors without reservation. To that end, the BCC is working with the Permanent Congolese Council on Accounting (PCCA) and a charter defines the roles and responsibilities of the BCC’s departments in preparing the accounts. In addition, senior management selected an accounting software that would allow all of the BCC’s offices and branches to have direct access to the integrated accounting application that will be implemented by late 2005. This will greatly facilitate the reconciliation of transactions between headquarters and branches and hence reduce the amount in suspense accounts, a major source of problems.

32. A key component of the central bank action plan is to modernize its computerized information system, including the infrastructure, operational systems, and procedures. Guided by the Information Technology (IT) Committee established in April 2004, the BCC has selected computer applications in the areas of accounting and foreign exchange transactions and has started the bidding process to acquire equipment and software. In addition, with technical assistance from the Fund and bilateral donors, the BCC has prepared an ambitious IT plan for all its departments and branches for which the authorities are seeking financing.

Banking supervision

33. The strengthening of banking supervision has involved the computerization of supervision (Bank Supervision Application), which is soon to be completed. To that end, the central bank is working on a new system to generate periodic statements reflecting the reform of banks’ charts of accounts. In this regard, with the assistance of the PCCA, the BCC has prepared a draft charter of accounts, which is expected to be implemented starting in January 2006. The BCC has also assigned staff to either off-site or on-site supervision to improve the professionalism and level of competence of its inspectors. To ensure continuous monitoring of banks, it has formalized the operating mechanism for off-site supervision and implemented a framework for permanent files. The BCC has strengthened on-site supervision by providing it with remote assistance.

Combating money laundering and the financing of terrorism

34. The law concerning money laundering and the financing of terrorism (ML/FT), promulgated in July 2004, conforms to international norms: it specifies activities considered to be crimes; preventive measures to be taken by financial institutions; monitoring to be exercised by bank supervisory authorities; the seizure of items resulting from illegal activities; actions to be taken to facilitate international cooperation; and the creation of a national unit responsible for collecting financial information (CENAREF, Cellule national de renseignement financier).

35. To implement the law requires an instruction from the governor of the BCC specifying cases when payments of more than US$10,000 can be made in cash, a decree creating the CENAREF, and a circular from the ministers of justice and of finance stipulating the rules to be applied for businesses operating in the informal sectors or in localities without financial institutions. These documents are expected to be issued by September 2005. To strengthen the national law, the DRC must also ratify international conventions concerning money laundering and the financing of terrorism.6 Implementing the new law and regulations and creating the new institutions responsible for combating ML/FT will require extensive staff training, which could be provided by multilateral institutions (IMF, the World Bank, the United Nations, and the European Union) and bilateral donors.

Reform of commercial banks

36. The BCC and the authorities have taken steps to strengthen the commercial banks with a view to increasing financial intermediation, and thereby, the effectiveness of monetary policy. All banks submitted their restructuring plans, which were based on audits completed in 2003, to the BCC in January 2004. From those plans, the BCC drew up a list of nine banks to be liquidated or restructured and launched a call for bids to select the liquidators. The social plans for the three public banks in liquidation were finalized in September 2004, and the World Bank has committed financing. However, liquidation is awaiting a decision on the benefit package to be granted laid off employees. The BCC also approved the restructuring plans of five other banks, which are currently being implemented.

37. The BCC has also introduced a number of measures to strengthen commercial banks’ operations: instructions on the banks’ prudential ratios in July 2003, adoption of a formal framework for on-site and off-site audits, reform of the tax and legal framework for commercial banks to make it more transparent and more in line with good international practice and adoption, in early 2004, of a decree amending the chart of accounts to facilitate the reconstitution of equity capital. In addition, draft laws introducing tax deductibility of statutory loan-loss provisions, the reduction of the personal property tax (from 6 percent to 1 percent), and a more attractive investment regime for the financial sector were submitted to the National Assembly in July 2004 but have not yet been discussed. Finally, with the 2005 Finance Law, the taxation of money transfers has been streamlined, ensuring that commercial banks can compete with nonbank financial institutions operating in that sector.

II. The Relationship Between Money, the Exchange Rate, and Inflation7

38. The relationship between money, the exchange rate and inflation has been at the center of recent developments in the DRC. Inflation has fallen from 511 percent in 2000 to 4.8 percent at end-2003, before rising to 9.3 percent by end-2004. During the same period, the Congo franc was relatively stable against the U.S. dollar, depreciating gradually from CGF 312 per US$1 at end-2001 to CGF 373 at end-2003. Since June 2004, however, the currency has become more volatile, with the exchange rate rising to CGF 444.1 by end-2004 and CGF 515 by April 2005, before recovering to about CGF 490 by end-June.

39. Monetary policy has played a key role in the developments described above. During 2002-03, broad money grew by an average of 34 percent a year, while nominal GDP growth averaged 28 percent and inflation 10 percent. In 2004, however, broad money grew by 73 percent much faster than nominal GDP (13 percent). The main reason behind the surge in money growth was the easing of fiscal policy as a result of increased military expenditure associated with the situation in the eastern provinces and increased outlays related to political institutions, and an expansion in the BCC’s other items net. The monetary authorities accommodation of these expenditures led to an increase in the monetary base of 63 percent during 2004. This rapid monetary expansion led to instability in the exchange rate, as noted above.8

40. This chapter examines the impact of monetary and exchange rate policies on inflation, including the nature of the lags involved. In particular, it seeks answers to the following questions:

  • Does monetary policy affect the exchange rate and, if so, by how much? What is the lag between a change in the monetary base and a change in the exchange rate?

  • Do exchange rate movements affect inflation? How strong is the exchange rate pass-through; that is, by how much will inflation increase in response to a 1 percent depreciation of the exchange rate? How long does it take for inflation to respond to the exchange rate depreciation?

  • Does an increase in money growth lead to an increase in inflation, or does money growth merely reflect changes in inflation (endogenous money)? By how much will inflation rise in response to a 1 percent increase in money growth? How long does it take for inflation to respond to an acceleration of the money supply?

A. The Data

41. The study, covering the period 2002-04, uses monthly data for the monetary base, the dollar bilateral exchange rate, and the consumer price index (CPI). The data are from the IMF’s International Financial Statistics with staff updates as required for the most recent months.

42. The sample period was chosen because it is considered to be representative of the policy environment in which the authorities currently operate. A longer period would include episodes of conflict and hyperinflation, which would introduce a bias into the analysis. The study uses the monetary base as a measure of money because it allows us to better gauge the role that monetary policy plays in controlling the exchange rate and inflation.9 The monetary base variable is defined as currency in circulation (held by banks and outside banks), bank deposits, public enterprise deposits, and private deposits at the central bank.10

43. The study uses the U.S. dollar bilateral exchange rate (as opposed to the nominal effective exchange rate) because of the prominence of the U.S. dollar in the Congolese economy11 and the fact that the central bank uses the U.S. dollar for foreign exchange intervention.

44. Finally, we used the end-of-month consumer price index (CPI) as a measure of the price level because the monetary authorities, in practice, aim to control CPI inflation. Although CPI inflation attracts the most attention and is widely understood, it would have also been desirable to consider a measure of domestic inflation, which would better assess the impact of monetary policy on prices.12 However, the authorities do not produce such a measure.

B. A Few Stylized Facts

45. For the sample period, the average rate of growth of money is 3.3 percent a month with a standard deviation of 5.1.13 The average rate of change of the exchange rate is 0.9 percent a month, with a standard deviation of 4.1. The average rate of change of inflation is 0.7 percent a month with a standard deviation of 1.2. More detailed statistics are presented in Tables II.1 and II.2.14

Table II.1.

Democratic Republic of the Congo: Summary Descriptive Statistics, 2002-04

(Monthly growth rates)

article image
Table II.2.

Democratic Republic of the Congo: Summary Descriptive Statistics, 2002-04

(Monthly indices)

article image

46. Looking at the data series (see Figures II.1 and II.2), one can observe the following:

  • An increase in the monetary base is associated with a depreciation of the exchange rate. Until the middle of 2003, the rate of depreciation of the exchange rate appears to be in line with the rate of growth of the monetary base. Beginning in mid-2003, however, the depreciation of the exchange rate was considerably less rapid than the increase in the monetary base.

  • A faster growth in the monetary base is associated with slightly higher inflation. However, the relationship between the two variables is considerably weaker beginning in mid-2003, when rapid increases in the monetary base do not seem to have led to a considerable increase in inflation.

  • A faster depreciation of the exchange rate is associated with higher inflation. The relationship between the two variables appears to be rather close. The lag between exchange rate movements and movements in the price level appears to be short, with the evidence suggesting that exchange rate movements precede changes in the inflation by less than two months.

Figure II.1.
Figure II.1.

Democratic Republic of the Congo: Money, Exchange Rate, and Price Level, 2002-04, January 2002 = 100

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

Figure II.2.
Figure II.2.

Democratic Republic of the Congo: Inflation, Exchange Rate, and Monetary Base, 2002-04

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

47. Examination of the contemporaneous correlation matrix among the three variables (Table II.3.) shows that there is a strong correlation between the price level and the monetary base, on the one hand, and the price level and the exchange rate, on the other. In contrast, the correlation between the exchange rate and the monetary base is weaker. The contemporaneous correlation among the rates of change of the three variables (i.e., growth of (Table II.4) base money, rate of depreciation, and the rate of increase of the price level) does not appear to be substantial only in the case of the exchange rate and the price level.15 This may suggest that, whereas changes in the exchange rate and the price level are simultaneous, the impact of money growth on inflation is felt with a lag.

Table II.3.

Democratic Republic of the Congo: Sample Correlation Matrix Among Indices

article image
Table II.4.

Democratic Republic of the Congo: Sample Correlation Matrix Among Growth Rates of Indices

article image

48. Pairwise Granger causality tests (based on F-tests) also show that the null hypothesis of no Granger causality fails to be rejected (that is, it is accepted) for all variables in most tests conducted.16 On a few occasions, however, there is some evidence that money and, to a lesser extent, the exchange rate “cause” the price level.17 When causation involving the rate of change of variables (growth rate of money, exchange rate depreciation, inflation rate) is examined, the null hypothesis of no Granger causality fails to be rejected (that is, it is accepted) except in the case of the rate of depreciation causing an increase in inflation.18

49. Overall, the data give the impression of a strong relationship between the exchange rate and the price level and a weaker relationship between the monetary base and the price level, on the one hand, and the monetary base and the exchange rate, on the other.

C. Vector Autoregression Analysis

50. The study estimates the dynamic structure underlying the relationship between money, the exchange rate and prices using vector autoregression analysis (VAR). The use of VAR will provide some insight into the direction of causality between the variables and the appropriate lag structure. Once the model is specified, it is used to analyze the dynamic impact of random disturbances on the system of variables.

51. The study is motivated by McCarthy’s (2000) work on the exchange rate pass-through in a sample of nine developed countries. Analyzing the impact of exchange rate changes and import prices on producer and consumer prices, McCarthy finds that the exchange rate has a rather limited and insignificant effect on consumer prices, and a stronger effect on import prices. In general, the pass-through appears to be endogenous to different regimes and tends to be smaller when inflation is lower. The pass-through in transition economies tends to be larger, in part because monetary policy may suffer from lack of credibility and domestic firms may act as price takers in international markets.19

52. The underlying model used in this paper is based on a simplified version of McCarthy (2000) reflecting data constraints. In the model, aggregate demand shocks act as exogenous to the exchange rate shock in period t. This exchange rate shock then feeds into domestic inflation.

53. Money shocks are considered to be the main source of fluctuations in aggregate demand. Thus, any disequilibrium between the demand for and supply of money is assumed to feed into the exchange rate and prices.20 Under a floating exchange rate system, shocks in aggregate demand are manifested primarily in the exchange rate and less in volatility of reserves and the monetary base. In principle, the money stock should reflect the central bank’s behavior as well as private sector decisions (given the degree of dollarization in the DRC). However, because the purpose of the paper is to draw implications for monetary policy, and given that money shocks emanate mostly from the central bank’s behavior, the study uses the monetary base as a proxy for money.21 Unlike McCarthy (2000), this study does not explicitly model the behavior of the central bank. Therefore, an interest rate variable is not included in the model (similar to a central bank “rule”) since the DRC does not have a well-functioning money market (interest rate data do not reflect market behavior). In effect, the equilibrium stock of money is determined by the expectations of the previous period and an error term.

54. The VAR used in this study is thus based on three endogenous variables: the monetary base, the exchange rate, and inflation. Details on the model specification are provided in the appendix.22

D. Results

55. The estimated VAR is used to simulate impulse response functions that, over time, trace the effect of a onetime temporary shock to one variable (that is, more specifically, a shock to one of the innovations associated with a given variable) on current and future values of the endogenous variables in the VAR. A shock to a given variable not only directly affects that variable, but it is transmitted through the lag structure of the VAR to all other endogenous variables. For stationary VARs, the impulse responses (calculated using the Cholesky orthogonalization) should die out to zero.

56. The analysis of the impulse response functions shows that the response of variables to shocks affecting other variables is in most cases, rather modest, although the speed of adjustment is rapid. In most cases the bulk of the adjustment is complete within three to four months. In particular, the following results are obtained (Figures II.3 and II.4):

  • A one- standard- deviation innovation in the monetary base (equivalent to an increase in the monetary base of about 4.4 percent) leads to a depreciation of the exchange rate of 1.6 percent. The adjustment is complete in five months, with most of the adjustment (about 85 percent) completed within two months.

  • A one-standard-deviation innovation in the exchange rate (equivalent to a depreciation of the exchange rate of about 4.1 percent) leads to an increase in inflation of 1.4 percent. The adjustment is complete in six months, with most of the adjustment (about 85 percent) occurring within two months.

  • A one-standard-deviation innovation in the monetary base (equivalent to an increase in the monetary base of about 4.4 percent) leads to an increase in inflation of 0.3 percent. The adjustment is complete in less than six months, with most of the adjustment (about 88 percent) completed within three months.

Figure II.3.
Figure II.3.

Democratic Republic of the Congo Response to Cholesky One Standard Deviation Innovations ± 2 Standard Error 1/

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

1/ D(LNMB) = Natural log Monetary Base; D(LNE_EOP) = Natural Log Exchange Rate; LNINFL_EOP = Natural log Inflation.
Figure II.4.
Figure II.4.

Democratic Republic of the Congo: Accumulated Response to Cholesky One Standard Deviation Innovations ± 2 Standard Errors 1/

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

1/ D(LNMB) = Natural log Monetary Base; D(LNE_EOP) = Natural Log Exchange Rate; LNINFL_EOP = Natural log Inflation.

57. It is worth noting the following:

  • Although changes in the monetary base tend to have a substantial impact on the exchange rate, the reverse is not true. Specifically, a one-standard-deviation innovation in the exchange rate (equivalent to a depreciation of the exchange rate of about 4.1 percent) leads to a an increase in the monetary base of only 0.1 percent. Most of the adjustment occurs during the first two to three months and dies out subsequently.

  • Changes in the exchange rate tend to affect inflation, but the reverse is not true. Specifically, a one-standard-deviation innovation in the inflation rate (equivalent to an increase in the rate of inflation of 0.95 percent) leads to no change in the monetary base. The lack of response is evident from the outset.

  • Although changes in the monetary base have only a rather limited effect on inflation, the response of the monetary base to changes in inflation seems to be stronger. Specifically, a one-standard-deviation innovation in the inflation rate (equivalent to an increase in the rate of inflation by 0.95 percent) leads to a reduction in the monetary base of 1 percent. The adjustment is complete in about five months, with a stronger adjustment occurring at the beginning (within two months).

58. Although the above relationships are not formal proofs of causality, they do provide tentative indications of the following directional relationships:23

  • The monetary base affects the exchange rate but the exchange rate does not affect the monetary base.

  • The exchange rate seems to affect inflation, but inflation does not have an impact on the exchange rate.

  • Inflation affects the monetary base, but the monetary base does not significantly affect inflation.

59. In addition to the impulse response functions, the study analyzes the variance decomposition of the endogenous variables in response to the innovation shocks (Figure II.5). In particular, the study separates the variation of an endogenous variable into the component shocks to the VAR, thus permitting conclusions about the relative importance of shocks in explaining the variation of a given variable.

Figure II.5.
Figure II.5.

Democratic Republic of the Congo: Variance Decomposition 1/

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

1/ D(LNMB) = Natural log Monetary Base; D(LNE_EOP) = Natural Log Exchange Rate; LNINFL_EOP = Natural log Inflation.

60. Regarding inflation, the study finds that the most important factor explaining the variation in inflation24 is the depreciation of the exchange rate (it accounts for about 54 percent of the total variation of inflation). Of the rest, 38.5 percent is explained by disturbances affecting inflation, and 7.5 percent by changes in the monetary base.

61. The variation in the exchange rate is almost entirely due to innovations affecting the exchange rate itself (94.9 percent).25 Innovations affecting the monetary base account for only 4.8 percent of the variation in the exchange rate and innovations affecting inflation account for 0.3 percent.

62. Finally, the variation in the monetary base is, for the most part, due to innovations affecting the monetary base (about 83.6 percent), with innovations affecting inflation accounting for about 11.8 percent, and innovations affecting the exchange rate about 4.6 percent.

E. Policy Implications and Conclusions

63. The main conclusions that can be drawn from this study are the following:

  • The transmission mechanism for monetary policy seems to work primarily through the exchange rate.26

  • More specifically, an acceleration in the rate of growth of the monetary base leads to a depreciation of the currency (that is, a change in the price of tradable relative to non-tradable), which then leads to higher inflation. The apparently weak relationship between the monetary base and inflation is consistent with the dollarization of the Congolese economy. In particular, in the presence of dollarization27 the effectiveness of monetary policy is limited. Furthermore, monetary policy is further limited by the lack of banking intermediation, which is partly due to the difficulties faced by the banking system.28

  • The exchange rate pass-through in the DRC is quite rapid.

  • This result is consistent with the widespread dollarization of the economy. In particular, the substantial and rapid response of inflation to the exchange rate seems to imply that many local currency-denominated prices are first set in dollars. Hence, those local currency prices adjust soon after there is a change in the exchange rate.

64. A number of other policy implications can be drawn from the study: First, monetary policy does not seem to be responsive to changes in the exchange rate. This result is consistent with a floating exchange rate regime under which the monetary authorities do not target a particular exchange rate.29 Second, there is some evidence of money endogeneity, in that the monetary authorities seem to respond to inflation developments (by reducing money growth) rather than by actively controlling inflation. Finally, the nonresponsiveness of the exchange rate to changes in inflation could be consistent with the view that the nominal exchange rate is not influenced by considerations other than market fundamentals, such as the maintenance of external competitiveness.

65. The above results should be interpreted cautiously. The time period covered by the study is too short to permit the drawing of accurate conclusions about the effects of monetary policy on the exchange rate and inflation. For example, the rather small sample may reduce the power of estimation results. Furthermore, data limitations (for example, monthly data on output and the interest rate) could bias the results presented in this study. It is possible, for example, that monetary policy might have a substantial and independent effect on prices through an interest rate channel that is not captured by the model.

III. Wages and Civil Service Reform30

A. Background

66. Consolidating the gains of the stabilization program that began in 2001 requires a civil service that is well trained and capable to lay the groundwork for the reform agenda. To rebuild administrative capacity, the authorities have started to implement an ambitious reform of public administration including the civil service. The reforms aim at improving the quality and capacity of service delivery, payroll management, and organizational structure. As first steps, the authorities conducted audits of ministries and started a census of civil and military personnel to streamline the civil service, better reward qualified personnel, and improve transparency in budget execution. Preliminary census results implemented in Kinshasa seem to indicate that the payroll could be reduced by more than 20 percent by removing ghost workers.

67. The key challenges for the civil service reform are to keep the wage bill at a level consistent with macroeconomic stability while finding ways to achieve the following:

  • Raise salaries, which, at the bottom of the scale, are about US$25 a month in Kinshasa and US$4 a month in the provinces.

  • Widen the salary compression ratios between the highest and lowest paid, which are less than 2 in Kinshasa and about 5 in the provinces.

  • Hire additional personnel in the social sectors to achieve the Millennium Development Goals (MDGs).

  • Strengthen civil service management, including improving control over the payroll, to ensure that the wage bill remains at a level compatible with macroeconomic stability, and that staffing reflects the government’s priorities.

68. The paper assesses the scope for increasing salaries, the salary compression ratio, and recruitment for the social sectors while keeping the wage bill at a manageable level.

B. Level of the Wage Bill

69. The wage bill as a share of GDP in the DRC is low relative to that in other African countries, including postconflict countries. It amounted to 3.6 percent of GDP in 2004, even though it has doubled since 2001 in line with steps taken toward establishing a fully functioning administration in DRC (Table III.I).31 Per capita salaries are also low compared with those of other African countries, averaging only 3.7 times per capita GDP in 2004 (Table III.2 to Table III.4). Specific steps taken in the past few years include the regular payment of wages to the civil service, military, and police, the creation of new political institutions, increases in pay scales, and the granting of new allowances. Also, the wage bill (including for the military) as a share of GDP is much lower in the DRC than in neighboring countries.32

Table III.1.

Democratic Republic of the Congo: Composition of Government Personnel and Remuneration

article image
Source: Fund staff estimates and Congolese authorities.
Table III.2.

Democratic Republic of the Congo: Government Wages and Salaries in Selected African Countries

(In percent of GDP)

article image
Sources: Fund staff estimates.

The small states comprise Botswana, Lesotho, Namibia, and Swaziland.

Table III.3.

Democratic Republic of the Congo: Government Wages and Salaries

(In percent of total government expenditure)

article image

Small states include Botswana, Lesotho, Namibia and Swaziland.

Table III.4.

Democratic Republic of the Congo: Comparison of Government Wages and Salaries in Selected African Countries

(In percent of revenue excluding grants)

article image
Sources: Fund staff estimates.

The small states comprise Botswana, Lesotho, Namibia, and Swaziland.

70. In spite of its low level, the wage bill has absorbed an increasing proportion of total government revenue. The wage bill has increased relative to revenue from 26 percent to 38 percent since 2001. However, the ratio of the wage bill to government revenue is close to the average for the region.

uA01fig02

Postconflict Countries: Wages to GDP Ratio

(In percent)

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

71. As a proportion of recurrent expenditures, the wage bill has risen slowly—accounting for less than 30 percent—because other current expenditures, which include transfers and spending on goods and services, have risen faster since 2000.

72. Restructuring of the police and armed forces has also played a role in increasing the size of the wage bill. The Global and All Inclusive Agreement on Transition and Memorandum on Military and Security Issues of March 6, 2003, defined the framework for the formation of a national, integrated, and restructured armed forces. As a result, the size of the military and police is estimated to have increased by 83 percent since 2001 to 410,000 by end-2004, and their payroll has more than doubled, accounting for one-third of the wage bill.

uA01fig03

Postconflict Countries: Wages as a Percent of Total Revenues

Citation: IMF Staff Country Reports 2005, 373; 10.5089/9781451808384.002.A001

73. Another issue is that the salary compression ratio for the civil service has fallen sharply in recent years. Before to April 2004, salaries consisted of a base pay with a ratio between the highest and lowest salary of 10:1 and lump-sum transport allowances in Kinshasa. Hence, the overall compression ratios were 2.2:1 in Kinshasa and 10:1 in the provinces. In response to pressures from the civil service unions, the Mbudi Accord of February 2004 was reached to find ways to increase civil service pay.33 As a first step, in 2004, the authorities granted civil servants in Kinshasa and Lubumbashi a doubling of the lump-sum transportation allowance while keeping the scale for base salary unchanged (Table III.5).34 In addition, in the 2005 budget, civil servants in Kinshasa were granted a lump-sum housing allowance and those outside Kinshasa were granted an additional lump sum for transport. As a result, the compression ratios fell to 1.6:1 in Kinshasa and to 5:1 in the provinces. Nevertheless, there remain large disparities between the salaries of civil servants and those in political positions, with parliamentarians for instance, receiving as much as US$1,500 a month.

Table III.5.

Democratic Republic of the Congo: Monthly Wage Bill by Category of Worker Based on Proposed Compression Ratio

article image
Sources: Congolese authorities; and Fund staff estimates and projections.