Sixth Review Under the Three-Year Arrangement Under the Poverty Reduction and Growth Facility (PRGF), Requests for Waivers of Nonobservance of Performance Criteria, and Request for a New Three-Year Arrangement Under the PRGF

Policy implementation has improved since the completion point. Recent economic performance has been encouraging, but sustaining it will require an acceleration of reforms by the government. Rwanda’s medium-term program aligned with the Poverty Reduction Strategy Program priorities provides an appropriate framework, but further refinements are needed. The main macroeconomic challenge is managing the domestic demand impact of fiscal policies. Executive Directors welcome the scope of structural reforms aimed at private sector development, improving public service delivery, and productivity-enhancing strategies in key sectors.

Abstract

Policy implementation has improved since the completion point. Recent economic performance has been encouraging, but sustaining it will require an acceleration of reforms by the government. Rwanda’s medium-term program aligned with the Poverty Reduction Strategy Program priorities provides an appropriate framework, but further refinements are needed. The main macroeconomic challenge is managing the domestic demand impact of fiscal policies. Executive Directors welcome the scope of structural reforms aimed at private sector development, improving public service delivery, and productivity-enhancing strategies in key sectors.

I. Overview

1. Rwanda should seize the opportunity to graduate as a mature stabilizer from the proposed new PRGF arrangement. The EPA noted that the Fund’s engagement had supported the country’s emergence from a post conflict situation. However, Rwanda remained vulnerable to external shocks and dependent on external aid, with its structural reform agenda still incomplete. The objective of the new PRGF arrangement is to help set the stage for stronger growth and faster poverty alleviation, including through an acceleration of key structural reforms.

2. The Great Lakes region remains unstable, although violence has decreased. While the main rebel group in the East Democratic Republic of Congo (DRC) declared in March 2005 that it would disarm and return to Rwanda, only few combatants have handed in their weapons. To accelerate demobilization, the DRC army has sporadically attacked smaller groups with logistical support from UN troops.

II. Recent Developments and Program Performance

3. Macroeconomic performance was strong in 2005, despite temporary slippages in policy implementation. Real growth accelerated to 6 percent, reflecting an expansion in the manufacturing, financial, and communication sectors, as well as a good harvest after two years of largely stagnant agricultural output. Owing to increased food supplies, inflation declined as programmed to slightly below 6 percent at end- 2005. Overall, tighter fiscal policies combined with buoyant exports resulted in a lower-than-programmed current account deficit. However, the program went temporarily off track in the third quarter, mostly owing to difficulties in managing the backloaded fiscal program.

uA01fig01

Food and Energy Prices were the Key Determinants for Price Developments during 2001–06

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

4. The fiscal stance was tighter than projected, but mainly in the first half of the year. Buoyed by higher growth and efficiency gains in tax administration, the domestic deficit was below the target for the year as a whole.1 Until mid-year, spending was restrained, which, together with temporary absorptive constraints, caused priority spending to fall short of the end-June performance criterion (by 0.6 percent of GDP). When spending was accelerated in the third quarter, the program temporarily went off track as domestic arrears were accumulated as the central bank was reluctant to mop up the excess liquidity stemming from the additional spending. Moreover, the end-September target on priority spending was not met owing to continuing capacity constraints, and the authorities incurred new nonconcessional debt related to the electricity crisis.2 These slippages were corrected in the last quarter and the December targets were met as the arrears were cleared and absorptive constraints for priority spending removed by progress with the decentralization process.

5. The reserve money program mostly stayed on track. Broad money at end-2005 was substantially higher than programmed, allowing a sizable expansion in private sector credit. This reflected higher money demand and real growth, but also strong seasonality toward end-year. As a result of large fluctuations in the money multiplier,3 all end-quarter reserve money targets were met. While overall the tighter fiscal stance helped to keep reserve money on track, the National Bank of Rwanda (NBR) resorted to issuing short-term domestic debt to mop up the liquidity from the accelerated fiscal spending in the last quarter rather than selling foreign exchange.

uA01fig02

Both broad money and the money multiplier have been fluctuating substantially

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

uA01fig03

Interest rates remained unchanged…

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

uA01fig04

…but domestic debt sharply increased in the last quarter of 2005.

(In percent of GDP)

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

6. The current account deficit was lower than programmed. Exports were substantially above program projections, reflecting a supply response to a sharp increase in mineral prices, but also the move into high-value niche coffee markets. While imports for construction materials and medical supplies grew substantially, the tighter fiscal stance restrained the growth of other imports. This, combined with higher-than projected project disbursements and heavier reliance on domestic sterilization, led to net foreign assets of the NBR exceeding the program target.4 In line with these developments, the nominal and real effective exchange rates appreciated by 9 percent and 13 percent, respectively.

uA01fig05

Despite the substantial appreciation since end-2003, the real effective exchange rate remains well below its level at the end 90s

(1997 average=100; foreign currency per Rwandan Franc)

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

7. Overall progress on the structural side was mixed, but adherence to conditionality was poor ( Memorandum of Economic and Financial Policies (MEFP), para. 6). While the performance criterion (PC) on implementing zero-balance accounts for ministries and autonomous agencies was met ahead of schedule, the end-September PC on the publication of Prime Holdings’ audit was not met (but implemented in December).5 Moreover, out of five end-2005 benchmarks, three related to expenditure management were not observed, mostly due to the restructuring of administrative zones6 and delays in parliamentary approval of the Organic Budget Law (OBL); the implementation of these benchmarks is pending. The end-2005 benchmark on export promotion also was not met as two involved agencies signed the required memoranda of understanding only in April 2006. On the positive side, three tea estates were recently privatized. Moreover, banking supervision has been strengthened as evidenced by the submission of draft amendments to the banking law to stakeholders at end-2005 (benchmark) and an agreement on a restructuring plan with a problem bank. In the energy sector, the Lake Kivu methane gas project is advancing as scheduled.

8. Despite temporary slippages, performance under the PRGF program was broadly satisfactory. All but two performance criteria (on priority spending and the Prime Holdings’ audit) for the sixth review were observed. While difficulties in managing the fiscal program led to slippages in the third-quarter, these were subsequently rectified and all quantitative end-2005 targets were met.

III. Policy Discussions

9. Discussions centered on medium-term challenges, to ensure that policies in 2006 were designed to enhance Rwanda’s prospects for the future There was strong agreement that, in order to make inroads in poverty reduction, it was necessary to identify the growth-critical issues and remove the constraints to private investment and productivity growth.

A. Medium-Term Challenges

10. Poverty remains widespread. Substantial progress has been made toward some MDGs, particularly in education and gender issues. However, Rwanda is lagging behind in health care indicators and poverty remains substantially higher than before the genocide.

11. To improve Rwanda’s prospects, the EPA highlighted three challenges (MEFP, paras. 9–14):7

  • increasing growth rates and making the economy resilient to shocks;

  • financing the development effort without jeopardizing debt sustainability; and

  • marshalling aid flows to their most productive use while preserving macroeconomic stability.

The new PRGF focuses on reforms that are key to meeting these challenges as outlined below. The program also places particular emphasis on capacity building in designing and monitoring policies in order to set the stage for a graduation from PRGF resources.

uA01fig06

Rwanda: Selected Millennium Development Goals for 2015 1/

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

Sources: World Bank; World Development Report1/ When data were not available for the specific year, the figures use data from the closest available year.

Increasing growth and coping with shocks

12. To boost Rwanda’s growth rates, improvements in the business climate are needed. With the catch-up effect after the genocide waning and population growth at about 3 percent, only sustained high growth rates can make a marked dent in poverty. There was consensus that factors critical to this were financial sector and legal reforms, investment in infrastructure, and actions to facilitate trade. Moreover, staff urged the authorities to focus on reducing the cost of doing business, particularly with respect to obstacles to cross-border trade, property registration, and access to credit—areas where Rwanda is lagging in comparison with SSA countries.

uA01fig07

Rwanda ranks 139th out of 155 countries overall in the World Bank’s Doing Business Index

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

Source: World Bank Cost of Doing Business.
uA01fig08

Compared to the EAC and SSA, weaknesses exist in particular in property registration, access to credit, and cross-border trade

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

13. Increasing agricultural yields is particularly important for reducing rural poverty. The authorities felt that raising the productivity of the agricultural sector8 was essential for breaking the vicious cycle of poverty, with food security freeing people to seek better education and also employment in other sectors. While sector-specific bottlenecks are addressed in the Strategy Plan for Agricultural Transformation (MEFP, para. 36),9 the sector should also benefit from improvements in the overall business climate.

14. High growth rates may not be sustainable without reducing the cost and increasing the supply of energy. Owing to a combination of rising energy demand and declining production of hydropower, Rwanda had to start in 2004 to rely increasingly on higher-cost thermal power. To cover the additional costs, electricity tariffs were increased by almost 200 percent in 2005 and are now substantially higher than in neighboring countries. The Lake Kivu methane gas project,10 as well as the restoration of reservoir levels of the hydropower plants, is expected to replace the expensive thermal power and mitigate pressures on the current account from oil imports. However, the authorities underscored the need to increase supply to allow the use of competitive technologies in the economy and conserve limited wood fuel. To this end, staff urged the authorities to liaise with the World Bank to estimate and cost Rwanda’s medium-term energy needs.

Financing development

15. Despite debt relief under the Enhanced HIPC and MDRI Initiatives, development efforts will have to be financed mostly by grants11 Staff projections point to a substantial reduction in the NPV of debt-to-export ratio to about 59 percent at end- 2005 to a substantial reduction in the NPV of debt-to-export ratio to about 59 percent at end-2005 from the debt relief. While the relief provides some room for additional borrowing toward achieving the MDGs, high and sustained levels of grants are still required, given the low export base. Moreover, with Rwanda’s continued vulnerability to exogenous shocks and the eventual onset of external borrowing by the private sector, staff cautioned the authorities to keep the debt level below the LIC debt sustainability thresholds. The program thus maintains a zero limit on nonconcessional borrowing and requires a grant element of at least 50 percent for new borrowing.

uA01fig09

If half of the grants projected for 2007–10 were to be replaced by concessional loans (with a 50 percent grant element), Rwanda’s NPV of debt-to-exports ratio would exceed the HIPC threshold of 150 percent by 2010

Citation: IMF Staff Country Reports 2006, 245; 10.5089/9781451833409.002.A001

16. At the same time, domestic revenue will have to be mobilized to prepare for an eventual reduction in donor funding. The authorities agreed, but emphasized that the revenue ratio relative to monetized GDP was high and that bringing the nonmonetized sector into the tax net would take time. Staff concurred that it was not prudent to squeeze the existing tax base, but urged the authorities to broaden the tax net. Moreover, collection of minor local taxes could be simplified or integrated into the tax system.

Managing the development process

17. The authorities should prioritize development needs more clearly, while managing the likely pressures for a real exchange rate appreciation. Discussions focused on the macroeconomic impact of aid flows, in particular with respect to increased demand for nontradables. While some officials expressed a preference to spend an increase in aid on capital goods imports to avoid a real appreciation, there was broad consensus that social spending should not be neglected, despite its larger domestic component. Moreover, there was consensus that any persistent pressure for a real appreciation in line with fundamentals be accommodated through a nominal appreciation to maintain low inflation while raising absorption. Although the main elements of the development strategy appear to have been identified, staff urged the authorities to carefully consider the sequencing of reforms, ensure the productivity of spending, and to strengthen the linkages between policies and achieving the MDGs.

18. A further strengthening of expenditure management is warranted to assure development partners that resources are leveraged to their most productive use. The authorities expressed regret about the missed end-2005 structural conditionality. They stressed that this did not reflect a lack of attention, but rather the decision to first streamline administration toward more decentralization before improving the current system. Staff emphasized that strengthening expenditure management was even more important with the devolution of responsibilities to local governments, given their weak administrative capacity.

19. Improving the delivery of public services is a prerequisite for successful program implementation. There was agreement that more emphasis had to be placed on the training and retention of civil servants since limitations in administrative capacity had weakened program implementation in the past. In this context, staff urged the authorities to accelerate the training for the macroeconomic unit in the Ministry of Finance (MEFP, para. 37).

Macroeconomic framework

20. The medium-term framework assumes a gradual improvement in macroeconomic conditions, broadly in line with the projections presented at the completion point. The authorities maintained that public investment could generate significant returns within a short time period, especially in agriculture. However, acknowledging the pitfalls of overoptimistic projections, they agreed to base macroeconomic policies on conservative estimates. The framework is based on a real growth rate of 5 percent to 7 percent, 12 continued low inflation and a narrowing of the current account deficit. Fiscal policies would support preserving macroeconomic stability while increasing the revenue ratio and allowing an increase in spending priorities.

B. Program for 2006

21. Late rains are clouding the outlook for economic activity in 2006 (MEFP, para. 21). A January Crop and Food Assessment Survey projected a food deficit of about ½ percent of GDP (US$8.5 million) for the first half of 2006. Moreover, the mid-year harvest could be smaller than normal. To prepare for possible food shortages, the program sets aside US$8 million for food imports and includes an adjuster allowing a reserve drawdown should additional imports be needed.

22. The principal objective of the program is to advance the medium-term agenda, while preserving macroeconomic stability. The program envisages a real growth rate of 3 percent to 5 percent, inflation of 5 percent, and a level of international reserves of at least 4 months of imports. Macroeconomic and structural policies will aim at laying the foundations for moving Rwanda onto a higher growth trajectory.

Rwanda: Medium-Term Framework, 2003–09

article image
Sources: Rwandan authorities; and staff estimates and projections

Macroeconomic policies

23. Macroeconomic policies will continue to be geared toward managing the domestic demand impact of fiscal spending. Pressures for a real appreciation are expected to continue, with a widening of the fiscal deficit and the likely drawdown of project account balances.13 Moreover, the authorities intend to repay some of the domestic debt issued in the last quarter of 2005. Thus, safeguards have been put in place to ensure that fiscal policy is not crowding out investment.

  • To protect the monetary program, spending of about one percent of GDP (equally split between priorities and nonpriorities) has been made contingent on grants coming in as programmed and the monetary program remaining on track. The program also contains an explicit understanding on the central bank’s foreign exchange intervention strategy (para. 30).

  • To facilitate monetary management, a monitoring system for large project accounts will be set up (August 2006 PC on issuing the first report).

  • To evenly distribute the domestic component of fiscal policies, the quarterly path of expenditure has been smoothened.

Fiscal policy

24. The program allows an expansion in the domestic fiscal deficit14 by more than one percent of GDP, mostly stemming from a loss in revenue (MEFP, para. 17). Revenue is expected to drop by one percent of GDP, reflecting one-off revenue in 200515, as well as losses in import-related taxes, which are partly offset by measures to strengthen the revenue base. Expenditure will increase slightly with a strong shift toward priorities, financed by savings from the car fleet reform, the draw down of PRSC II disbursed at end-2005, and MDRI flow relief in 2006 (Box 1).

MDRI Relief

MDRI relief is being gradually incorporated into the program. Food imports1/ and Lake Kivu related spending broadly correspond to the resources freed-up by the MDRI relief in 2006 (0.6 percent of GDP). Depending on absorptive constraints and the domestic demand impact of fiscal policies, further priority spending could be accommodated in the context of the first PRGF review. Over the next three years, annual MDRI flow relief amounts to about ½ percent of GDP.

1/ If not needed for food, this could be used for other poverty-reducing spending.

25.Measures have been implemented to stabilize the revenue ratio at its 2005 level (excluding one-off revenue)(MEFP, para. 18). Staff welcomed the increase in revenue from fees and charges as well as the small increase in the petroleum excise (the reduction in the latter had resulted in revenue losses of about 2 percent of GDP in 2005).16 Further increases might be warranted should there be additional revenue shortfalls, but be accompanied by targeted measures to mitigate the impact on the poor and taking into account the prices in neighboring countries.17 While acknowledging the substantial improvement in the revenue ratio (particularly considering the losses from the petroleum excise), staff encouraged the authorities to explore other measures. Moreover, preparatory work for an eventual accession into the East African Community (EAC) (expected at end-2006) should be started.

26. The increase in priority spending reflects the authorities’ medium-term agenda. (MEFP, para. 19; text table). Particularly noteworthy is the elimination of school fees for the first three years of secondary education. While budgetary health spending does not increase substantially, health indicators are expected to improve from Rwanda being a major beneficiary of the Global Fund. Staff also welcomed the initiative to start monitoring a narrower definition of priority spending to complement the current aggregate as a first step to establish a more direct link between fiscal spending and poverty reduction.

Priority Spending, 2005–06

(In percent of GDP)

article image
Sources: Rwandese authorities; and IMF staff estimates.

27. Efforts in reducing nonpriority spending are ongoing. The authorities noted that “nonpriority spending” was a misnomer since it included many spending categories that were vital for running the government or ensuring social cohesion. As reclassifying such spending as priority would break the comparability of the time series, understandings were reached that any increases in specific nonpriority categories, if justifiable, would be clearly indicated (MEFP, para. 19, text table). With respect to specific categories, staff welcomed in particular the savings from the car fleet scheme, but suggested a continued monitoring of these accounts. It also noted the authorities’ proactive efforts to publish an audit report of 2005 and account for future peace-keeping activities as a clear signal that such activities did not imply a back-door increase in military spending (MEFP, para. 20).

Monetary and exchange rate policies

28. The reserve money program has served Rwanda well. Given the instability in the money multiplier, the authorities indicated their preference for a more flexible program design. Staff responded that moving to a NDA/NIR design would be premature as a nominal anchor was needed to prevent a rekindling of inflation, especially if aid inflows remained strong. If warranted, reserve money targets could be adjusted in the context of reviews.

29. While last year’s increase in private sector credit as such is welcome, the NBR concurred that a close monitoring of developments was needed (MEFP, para. 24). The NBR is stepping up its supervision, complemented by commercial banks’ efforts to tighten credit extension criteria, as well as loan classification and provisioning with a view to complying with the forthcoming banking law amendments. Staff urged the authorities to also make longer-term domestic papers more attractive relative to highly liquid instruments, which were closer substitutes for money and hence less effective as sterilization instruments.

30. It is crucial for the program’s success that the NBR follows through with its foreign exchange intervention policy (MEFP, para. 25). Since the private sector lists high interest rates as one of the main obstacles to growth,18 there was a broad agreement that sterilization through domestic debt was not an option as it would further drive up rates, negatively impact investment, and increase the losses of the NBR. However, some officials felt that any nominal exchange rate appreciation should be gradual in order to prevent exchange rate overshooting given the volatility of aid inflows. They also expressed a preference for a higher reserve coverage in the long run for the same reasons, while others noted that, given developments needs, increasing the reserve coverage would not seem appropriate. Staff concurred with a gradual approach for exchange rate adjustment, but emphasized that this would require close coordination with fiscal policies, including the smoothing of spending. Regarding reserve coverage, staff considered an import cover of four months as sufficient. To address the governor’s concern about the NBR’s revaluation losses from exchange rate appreciation, the treasury transferred 0.3 percent of GDP in March 2006 to cover the 2005 losses. The authorities also intend to revamp the foreign exchange auction system and revive the interbank market.

External sector

31. The current account is expected to widen in 2006 (MEFP, para. 26). Exports are projected to further increase with higher coffee receipts more than offsetting a decline in minerals receipts. Imports would surge, reflecting an increase in the oil bill,19 Lake Kivu project-related capital imports, as well as a drawdown of project accounts. With these import pressures expected to abate, the current account is projected to narrow starting in 2007.

Structural reforms

32. Structural policies will aim at moving forward the medium-term agenda (Box 2).

Structural Conditionality

Under the PRGF arrangement, structural conditionality (MEFP, Table 3) will focus on:

  • Monitoring project accounts to enhance the coordination of macroeconomic policies (para. 23).

  • Strengthening expenditure and debt management (paras. 33–34).

  • Undertaking a review of the wage structure (para. 35) as an interim step to making the civil service more efficient.

  • Improving the business climate for the private sector through financial sector and legal reforms (paras. 36–38).

  • Raising productivity in the agricultural sector (para. 39).

Under World Bank lending, structural conditionality is reflected in the Poverty Reduction Strategy Credit II, which focuses on: (i) the investment climate; (ii) improving public service delivery; and (iii) expenditure management and governance.

Table 1

Rwanda: Selected Economic and Financial Indicators, 2002–08

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

Data up to 2005 based on current exchange rates, for 2006 based on program exchange rate of RF 553.7/US$.

As a percent of the beginning-of-period stock of broad money.

Revenue excluding grants; minus current expenditure except interest due and exceptional expenditure; minus domestically financed capital expenditure.

Revenue excluding grants; minus current expenditure (excluding external interest), minus domestically financed capital expenditure and net lending.

After rescheduling, including arrears and new debt (the latter includes assumed project and budgetary disbursements for the period 2006–08).

Based on assumptions about expected new borrowing.

Table 2

Rwanda: Operations of the Central Government, 2004–08

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

The transactions related to the privatization of BCR and BACAR totaling RF 4.3 bn in the fourth quarter of 2004 did not impact the fiscal balance as privatization receipts (recorded as negative net lending) were offset by a transfer of the same amount (80 percent as a requited transfers for recapitalization of the banks and 20 percent as an unrequited capital transfers for the government purchasing a minority stake of 20 percent in the banks). Net lending in the fourth quarter of 2005 incorporates RF2.9 billion in receipts from the privatization of Rwandatel.

Definition excludes exceptional expenditures; defined as total revenue (excluding privatization proceeds) minus noninterest current expenditure (excluding exceptional expenditure) minus domestically financed capital expenditure

Revenue excluding grants; minus current expenditure, domestically financed capital expenditure and net lending; excluding external interest. In 2005, includes RF2 bn for recapilization of a housing bank, which are included below the line under nonbank financing.

Actual data for 2005 exclude Rf8.5 billion expenditures for peacekeeping operations, which were not covered by grants. The domestic deficit excluding grant-financed peace keeping would have been 5 percent of GDP in 2005. For 2006, peacekeeping activities not covered by grants is estimated at Rf 2.3 billlion and the domestic fiscal deficit excluding grant-financed peace keeping activities is estimated at 6.3 percent of GDP.

A negative sign indicates a reduction. Arrears are shown here in a fiscal accounting sense which may deviate from the definition of the TMU used for benchmarks and performance criteria.

The net credit to government at end-2004 has been corrected by the part of the PRSC disbursement that remained in the nontreasury accounts of the central government

CSR = Caisse Sociale du Rwanda.

6/ A negative number implies a discrepancy that is consistent with underestimation of financing.

Although energy was identified as a priority sector in the 2002 PRSP, spending on this sector was not included in priority expenditure until 2005. These expenditures amounted to RF11.4 bn and RF6.9 bn in 2004 and 2005, respectively.

Table 3

Rwanda: Monetary Survey, 2004–06

article image
article image
Source: National Bank of Rwanda (NBR); and Fund staff estimates and projections.

Converted at the actual exchange rate

Converted at the program exchange rate of RF 553.7/US$ in 2006.

The definition of reserve money as performance criterion or structural benchmark differs from the definition in the monetary program in that is excludes the deposits of a defunct savings bank, import deposits, and dormant accounts. Excess reserves were negative at end-2005.

From end-2005 onward, includes Caisse Hypothecaire du Rwanda (BHR) deposits (RF 1 billion at end-2005).