A. Alternative Scenarios

In this first approach, the impact on the public debt-to-GDP ratio in 2021—in both nominal and present-value (PV) terms—is simulated in response to the following shocks: (1) 2 percent lower real growth than in the baseline scenario; (2) 350 basis points higher nominal interest rates on domestic public debt; (3) 200 basis points higher nominal interest rates on external public debt; and (4) 4 percent more depreciation of the nominal exchange rate of the domestic currency vis-à-vis the U.S. dollar each year (that is, 28 percent cumulatively over 2014–21).⁸ As an extreme case, a combined shock of (1) to (4) is also considered. This approach isolates the impact of well-specified, individual shocks by computing the impact of each of the shocks, assuming that the other key assumptions remain as in the baseline scenario. This is, of course, a crude approach, as it does not take into account any possible policy reaction to the shock or the fact that the shock itself could have implications—negative or positive—for other variables. The results are summarized in Table 1.

^{Table 1.1}.

EAC: Sensitivity Analysis for Key Indicators of Public Debt, 2021¹

Sources: Country authorities; and IMF staff calculations and projections.

¹ The baselines correspond to DSAs conducted in 2014, which did not necessarily reflect the recent re-basing in national accounts in the region.

^{Table 1.1}.

EAC: Sensitivity Analysis for Key Indicators of Public Debt, 2021¹

	Burundi	Kenya	Rwanda	Tanzania	Uganda
PV of Debt-to-GDP Ratio
Baseline	17	36	18	33	30
A. Alternative scenarios
A1. Real GDP growth slows by 2.0 percent each year	29	56	33	48	45
A2. Interest rates on domestic debt are 350 basis points higher	19	41	19	35	33
A3. Interest rates on foreign debt are 200 basis points higher	18	38	21	37	33
A4. Exchange rate depreciates by 4 percent more each year	20	38	23	38	33
A5. Combined A1-A4	39	71	44	65	59
Debt-to-GDP Ratio
Baseline	19	41	28	44	38
A. Alternative scenarios
A1. Real GDP growth slows by 2.0 percent each year	32	62	44	61	55
A2. Interest rates on domestic debt are 350 basis points higher	22	46	29	47	41
A3. Interest rates on foreign debt are 200 basis points higher	21	43	31	48	41
A4. Exchange rate depreciates by 4 percent more each year	23	46	38	54	45
A5. Combined A1-A4	44	79	60	83	73
Baseline assumptions: averages for 2015–21
Real growth	5.3	6.9	7.4	7.0	7.1
Interest rate on domestic debt	7.3	9.1	0.5	7.0	11.0
Interest rate on external debt	1.1	2.4	2.0	2.2	2.2
Annual nominal depreciation (+)	2.5	0.3	3.0	2.7	3.6

Sources: Country authorities; and IMF staff calculations and projections.

¹ The baselines correspond to DSAs conducted in 2014, which did not necessarily reflect the recent re-basing in national accounts in the region.

View Table

^{Table 1.1}.

EAC: Sensitivity Analysis for Key Indicators of Public Debt, 2021¹

	Burundi	Kenya	Rwanda	Tanzania	Uganda
PV of Debt-to-GDP Ratio
Baseline	17	36	18	33	30
A. Alternative scenarios
A1. Real GDP growth slows by 2.0 percent each year	29	56	33	48	45
A2. Interest rates on domestic debt are 350 basis points higher	19	41	19	35	33
A3. Interest rates on foreign debt are 200 basis points higher	18	38	21	37	33
A4. Exchange rate depreciates by 4 percent more each year	20	38	23	38	33
A5. Combined A1-A4	39	71	44	65	59
Debt-to-GDP Ratio
Baseline	19	41	28	44	38
A. Alternative scenarios
A1. Real GDP growth slows by 2.0 percent each year	32	62	44	61	55
A2. Interest rates on domestic debt are 350 basis points higher	22	46	29	47	41
A3. Interest rates on foreign debt are 200 basis points higher	21	43	31	48	41
A4. Exchange rate depreciates by 4 percent more each year	23	46	38	54	45
A5. Combined A1-A4	44	79	60	83	73
Baseline assumptions: averages for 2015–21
Real growth	5.3	6.9	7.4	7.0	7.1
Interest rate on domestic debt	7.3	9.1	0.5	7.0	11.0
Interest rate on external debt	1.1	2.4	2.0	2.2	2.2
Annual nominal depreciation (+)	2.5	0.3	3.0	2.7	3.6

Sources: Country authorities; and IMF staff calculations and projections.

¹ The baselines correspond to DSAs conducted in 2014, which did not necessarily reflect the recent re-basing in national accounts in the region.

Changes in the key macroeconomic variables compared with the baseline projections could cause the public debt-to-GDP to come close to or exceed the EAMU threshold of 50 percent of GDP, especially when all adverse shocks occur simultaneously (combined adverse shock scenario). In the simulations, the shocks with the largest impact individually are a slowdown in real GDP growth and an exchange-rate depreciation.

While this first approach provides clear estimates of the impact of individual shocks, it does not incorporate information on the likelihood that these shocks would occur simultaneously or affect each other. To do so, it is necessary to analyze the whole distribution of macroeconomic shocks.

B. Complete Probability Distribution of Shocks

In this second approach, past data are used to estimate not only the volatility of each key macroeconomic variable (growth, interest rates, exchange rate) but also the correlations among them (variance-covariance matrix), to take into consideration the extent to which shocks to one variable are associated with shocks to the others. Specifically, pooled quarterly data for 1993–2013 (subject to availability) for the EAC countries are used to estimate the variance-covariance matrix for these variables (reported in Appendix I, Tables 1–2).⁹ Thousands of future paths are then generated by random draws of each of the variables from the estimated joint normal distribution to represent shocks, whose implications are then traced starting from the baseline projections. The fan charts reported in Figure 1.1 represent the distributions of projected debt paths for each EAC partner state.

Fan Charts of PV of Debt-to-GDP Paths for EAC Countries, through 2023

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Note: The solid dark blue line shows the staff’s baseline projection, and the solid red line denotes the EAMU convergence criterion ceiling on the PV of debt-to-GDP. The cone in light yellow in the center is the 20-percent standard error interval around the median projection. The outer edge of the dark pink cone marks the 99-percent confidence interval.Sources: EAC authorities; and IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Fan Charts of PV of Debt-to-GDP Paths for EAC Countries, through 2023

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Note: The solid dark blue line shows the staff’s baseline projection, and the solid red line denotes the EAMU convergence criterion ceiling on the PV of debt-to-GDP. The cone in light yellow in the center is the 20-percent standard error interval around the median projection. The outer edge of the dark pink cone marks the 99-percent confidence interval.Sources: EAC authorities; and IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

Fan Charts of PV of Debt-to-GDP Paths for EAC Countries, through 2023

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Note: The solid dark blue line shows the staff’s baseline projection, and the solid red line denotes the EAMU convergence criterion ceiling on the PV of debt-to-GDP. The cone in light yellow in the center is the 20-percent standard error interval around the median projection. The outer edge of the dark pink cone marks the 99-percent confidence interval.Sources: EAC authorities; and IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

Based on this approach, macroeconomic shocks alone seem unlikely to push the present value of government debt as a share of GDP of all EAC partner states beyond the 50-percent threshold in the early 2020s. The probability exceeds 5 percent only in Kenya and Tanzania; it is less than 1 percent in Burundi, Rwanda, and Uganda.

Based on these exercises, macroeconomic shocks are potentially a significant source of fiscal risks; however, given the relatively low public-debt ratios from which EAC partner states are starting and the relatively benign baseline scenarios, such shocks themselves do not cause a high likelihood of a breach of the EAMU debt threshold in 2021. It is important to recall, though, that there are several others sources of risk, some of which are analyzed in the next sections. Moreover, some of these additional risk sources are positively correlated with adverse macroeconomic shocks. For example, low economic growth is correlated with weaker financial performance of public enterprises, likely engendering the need for the central government to provide additional fiscal support.

A. Public Enterprises

Historically, public enterprises have played an important role in the economies of EAC partner states.¹⁰ The quality, timeliness, and coverage of information on the scale and financial performance of the public enterprises, both individually and in aggregate, needs to improve; however, available information makes clear that public enterprises are sufficiently important to pose significant fiscal risks—most require fiscal transfers and there have been many cases of bail-outs—and that the information gaps tend to heighten such risks. Lack of consolidated data covering the full public sector, including public enterprises, can prevent proper monitoring and risk management. Indeed, strengthened monitoring and oversight would help to keep risks in check.

Financial Performance

Collectively, public enterprises in the region pose macroeconomically relevant fiscal risks. They periodically require sizable transfers from the central government to cover losses. There is, however, considerable variation among the public enterprises with respect to financial performance. Some are chronic loss-makers; others’ performance is subject to significant variation, depending on commodity prices, weather conditions, and so forth; and some public enterprises reliably generate profits. This last group should not be overlooked from the perspective of fiscal risks: it is important to ensure that public enterprises operating in favorable conditions make adequate profits and that they remit profits to the treasury through taxes and dividends. Indeed, a significant source of fiscal risks relates to public enterprises that fail to generate adequate profits despite having a privileged position (monopoly, control of natural resources, and so forth) or that fail to remit profits to the treasury. As natural resources gain importance, the financial situation of public enterprises involved in the sector will merit even further disclosure of information and appropriate management.

In Tanzania and Uganda, the aggregate profit and loss account (after transfers) of the public enterprises has been broadly in balance, excluding the losses incurred by the national electricity companies (Table 2.1). In Kenya, the fiscal position has been the reverse, with the national electricity company generating profits and sizable losses for the public-enterprise sector outside electricity. In Rwanda, the budget position for utilities was in balance, with a slight aggregate deficit for the other public enterprises. In Burundi, the electricity company made profits in the most recent year for which information is available (2012) but remains in a weak position, with ongoing support from the World Bank; information for the rest of the public-enterprise sector suggests a small aggregate surplus, though there is no data on the extent of public subsidies received by public enterprises. Box 2.1 provides further information on the financial position (net income) of public enterprises in individual countries.

^{Table 2.1}.

EAC: Indicators for Public Enterprises

Sources: Country authorities.

¹ Commercial parastatals

² For Uganda, financial position is 2010/2011.

³ For Burundi, financial position is calendar year 2012.

^{Table 2.1}.

EAC: Indicators for Public Enterprises

	Number of companies	Public enterprise employment		Financial position in 12/13
		(percent of total employment)	1,000s	(in percent of GDP)
				Total	Total (excluding Utilities)
Kenya1	21	1.5	289	−2.3	−1.6
Rwanda	16	3.4	200	−0.6	−0.6
Tanzania	26	0.3	60	−0.4	0.4
Uganda2	29			0.1	1.7
Burundi3	16			0.4	0.3

Sources: Country authorities.

¹ Commercial parastatals

² For Uganda, financial position is 2010/2011.

³ For Burundi, financial position is calendar year 2012.

View Table

^{Table 2.1}.

EAC: Indicators for Public Enterprises

	Number of companies	Public enterprise employment		Financial position in 12/13
		(percent of total employment)	1,000s	(in percent of GDP)
				Total	Total (excluding Utilities)
Kenya1	21	1.5	289	−2.3	−1.6
Rwanda	16	3.4	200	−0.6	−0.6
Tanzania	26	0.3	60	−0.4	0.4
Uganda2	29			0.1	1.7
Burundi3	16			0.4	0.3

Sources: Country authorities.

¹ Commercial parastatals

² For Uganda, financial position is 2010/2011.

³ For Burundi, financial position is calendar year 2012.

^{Box 2.1}.

Available Information on Financial Position of Public Enterprises in EAC Partner States

Consistency of information on the financial position of public enterprises is difficult to ensure: therefore this box focuses on only a few indicators that can be compared across countries. These include the level of employment and the financial position of the public-enterprise sector as a whole (including and excluding utilities). In general, public-enterprise employees represent between 1 and 3 percent of the total number of employees, with the high number in Rwanda probably a reflection of better data.

Public enterprises have generally been making losses, with only Uganda and Burundi showing a positive aggregate balance (in 2010/11). Public utilities tend to make significant losses (for example, 0.7 percent of GDP in Kenya and 1.6 percent of GDP in Uganda). These losses materialize despite public subsidies. For country specifics, see the following:

Burundi. Following years of losses, the Burundi water and electricity company made a slight profit after tax of $2.4 million in 2012 (0.1 percent of GDP), following increases in water and electricity tariffs in September 2011 and March 2012. Small profits are projected through 2018.

Kenya. Transfers to public enterprises were equivalent to 5 percent of GDP between 2005 and 2010. In FY2011/12, 11 state corporations made losses.

Rwanda. Transfers to parastatals amounted to slightly less than 1 percent of GDP per year over the past five years. Rwandair, the national airline, continues to make losses after these transfers.

Tanzania. The public enterprises had aggregate net income close to balance in FY2012/13, but this masks wide differences among individual majority and wholly owned public enterprises, as well as large government subventions paid to some parastatals. Particularly strong financial performances have been reported by the National Housing Corporation and the Tanzania Ports Authority, which in recent years have generated significant financial surpluses and have not relied on government financial support. On the other end of the spectrum, the largest losses have been incurred by power utility TANESCO. Other public enterprises incurring large financial losses include Air Tanzania, water utilities, Tanzania Broadcasting Corporation Tanzania Telecommunication Corporation Limited (a joint venture with Airtel), the Tanzania-Zambia Railways Authority (jointly owned by the two governments), and public universities and colleges.

Uganda. Except for the electricity sector, subsidies to parastatals have decreased considerably in recent years to about 0.4 percent of GDP. The main loss-making public enterprise in Uganda is the telecom company, which made losses of more than 0.3 percent of GDP in FY2010/11 (most recent data point).

Oversight

Although rules for effective oversight are in place, the practical implementation of financial reporting in EAC partner states needs to be improved, as the lack of an effective system for monitoring and oversight can entail significant fiscal risks. Too few annual reports of public enterprises are publicly available. Reporting to national parliaments of sector-wide financial performance and operational developments is limited. Availability of timely information to ministries of finance also seems to be hampered by delays and data weaknesses in several countries. All of the EAC countries received Public Expenditure and Financial Accountability (PEFA) scores in the area of monitoring public enterprises that are about or below the global average (Figure 2.1). Box 2.2 summarizes the state of play in this area for individual countries.

EAC Country PEFA Scores on Financial Controls on Public Enterprises

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

EAC Country PEFA Scores on Financial Controls on Public Enterprises

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

EAC Country PEFA Scores on Financial Controls on Public Enterprises

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

^{Box 2.2}.

Oversight of Public Enterprises—Framework and Implementation in Individual EAC Partner States

Burundi. A 1998 decree states that all public enterprises must report on all of their activities to the Service Chargé des Entreprises Publiques; the Audit Court (Cour des Comptes) is authorized to audit the public enterprises. However, implementation of the regulatory framework has lapsed, and the World Bank is providing assistance to revamp it.

Kenya. The Department of Government Investment and Public Enterprises oversees public enterprises, but such monitoring could be enhanced. Since 2005 management performance contracts have been set with the parent ministry, and the Permanent Secretary of the Ministry is a member of the managing board. The Public Audit Act 2003 states that public enterprises must submit their financial accounts to the parent ministry and to the Ministry of Finance within three months of the end of the year, though information on compliance is limited. A recent effort has outlined clear reform plans for the government-owned entities (GOEs). Proposals include reducing the number of GOEs by 28 percent and grouping the remaining ones into five components: commercial state corporations with strategic functions, commercial state corporations without strategic functions, executive agencies, independent regulatory agencies, and research institutions. The shareholding role for commercial entities will be exercised by the treasury through a holding company called the Government Investment Corporation (GIC). The size of the boards of directors of the GOEs will be restricted to seven to nine members, and they will be vested with the authority to appoint the chief executive and top management. A revamped performance management framework will be set up and anchored in the overarching new law. The GIC will monitor performance of the public enterprises.

Rwanda. Public enterprises are monitored by the Government Portfolio Monitoring Unit. Their accounts are audited by external auditors, and the audits are normally presented to parliament within 10 months of the end of the accounting period. In practice, the audited reports are not comprehensive because of unavailability of data from some of the public enterprises. For example, the auditor general was not able to sign off on the annual financial report for Energy, Water and Sanitation Authority (EWSA), the electricity and water company, because the report was incomplete and not supported by underlying accounting records.

Tanzania. Each individual public enterprise is accountable to a board of directors, with oversight carried out by the parent ministry. The water and electricity utilities are regulated by an independent regulator. The Treasury Registrar (TR) has an overarching role in monitoring public enterprises (as well as numerous extrabudgetary agencies). TR prepares an annual report to parliament that includes the list of public enterprises and other government agencies, information on government equity holdings, government loans and guarantees, transfers from the central government, loan repayments and arrears, and dividend and other payments to the budget. TR also compiles summary information on the financial performance of public enterprises and extrabudgetary agencies, which was first published in November 2014.

Uganda. Most public enterprises are monitored by the Parastatal Monitoring Unit. Public enterprises must submit their financial accounts to the parent ministry and to the auditor general within three months of the end of the year. The accounts are generally audited with delays, in part because of workload issues.

B. Large Infrastructure Projects

Consistent with the need to bridge the infrastructure gap and foster economic growth, all EAC partner states envisage a significant scaling-up of infrastructure investment, including several large-scale projects mostly in energy and transportation. The combined value of all projects under consideration (to be implemented over the next few years) is in excess of US$90 billion, or about 70 percent of the combined GDP of the EAC countries in 2014.¹² This is, of course, a rough estimate; these projects are at different planning stages, and some may well not come to fruition, owing to such factors as lack of financing or operational challenges. Nevertheless, the sheer scale of the projects suggests that the fiscal implications are potentially major.

Some of the key projects in the pipeline are regional projects, undertaken in a coordinated way by some EAC countries. The main current ones include (1) the construction of the standard gauge railway (with about 2,900 kilometers expected to link Mombasa, Nairobi, Kampala, and Kigali, and be operational by December 2018, with an eventual extension to Juba in South Sudan), and (2) oil-related infrastructure, including a refinery in Uganda and a crude oil pipeline—to link the refinery in Hoima with the Kenyan coast—and a refined products pipeline—linking Mombasa, Nairobi, and Eldoret in Kenya with Uganda and, eventually, Rwanda.¹³

The magnitude of the envisaged investment varies considerably across countries (Figure 2.2). The time frame for individual projects also varies significantly, with some projects having started in the late 2000s and with completion dates ranging from 2014 to beyond 2020. Similarly, while some projects’ financing has been finalized, others’ funds are still being negotiated or planned, with most countries planning to have recourse to nonconcessional borrowing. Box 2.3 describes the major projects in the pipeline, derived from information available to IMF staff in the context of discussions with the authorities.

Envisaged Investment Projects in EAC, with or without Financing

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Envisaged Investment Projects in EAC, with or without Financing

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

Envisaged Investment Projects in EAC, with or without Financing

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

While these investment projects have the potential to address growth bottlenecks, they also carry fiscal risks whose extent depends on a range of factors. If the obligations stemming from a project listed in Box 2.3 are already included in the authorities’ fiscal framework (for example, the budget, the medium-term expenditure framework, and the DSAs), then fiscal risks would only be associated with cost overruns, delays in the finalization of construction, or the realization of contingencies that had not been included in the baseline. If some projects have not yet been included in the fiscal framework—as seems to be the case for a number of them—their large scale has the potential to significantly add to the fiscal deficits and debt burden, unless savings are made elsewhere in future budgets or these projects are financed by the private sector. Other economic issues could arise, such as the risk of overheating or running into absorption-capacity constraints.

C. Public-Private Partnerships

Given the fiscal constraints in most countries in the region, PPPs¹⁴ have received increasing attention in discussions on potential financing modalities. EAC partner states are taking active steps to improve their PPP frameworks, including their specific PPP legislation (Box 2.4 and Figure 2.3).

Contract Value of PPP Commitments over the Past Decade

(percent of 2012 GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Source: World Bank.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Contract Value of PPP Commitments over the Past Decade

(percent of 2012 GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Source: World Bank.

• Download Figure
• Download figure as PowerPoint slide

Contract Value of PPP Commitments over the Past Decade

(percent of 2012 GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Source: World Bank.

• Download Figure
• Download figure as PowerPoint slide

Although PPPs can increase efficiency, they also lead to fiscal risks for two broad sets of reasons. The first is that PPPs allow for deferred spending compared with public investment projects undertaken through standard public procurement, as the expenditures implied by PPPs tend to be recorded later—often at project completion rather than as the investment is carried out. The second relates to fiscal risks in their stricter sense of unexpected expenditures. Indeed, the PPP contract signed by the government (or the state-owned entity with the service provider/investment company) is usually difficult to design and implement, and includes several elements whose ultimate costs are uncertain, including: (1) the cost of the project in terms of the loans’ amortization; (2) interest payments; (3) the equity return for the investors; (4) operational costs and maintenance of the plants, including insurance; and (5) guarantees provided by the government. The main risks emanate from delays in construction, litigation, and increases in financing costs. Moreover, the assumptions regarding some projects are often optimistic and can result in guarantees being called on the government by the service provider. Additionally, some of the state-owned entities may be ill-equipped to manage such large projects or meet their PPP contract obligations. This may, in turn, result in a materialization of contingent liabilities for the government.

A number of preconditions must be met for effective management of fiscal risks. Firstly, project selection and appraisal need to be strong in order to ensure that only good projects get implemented. Secondly, a sound institutional, legal, and regulatory framework is necessary for a proper assessment of risks. Finally, a good budgeting, accounting, and reporting system must be in place to ensure full and transparent disclosure of risks.

D. Pensions

Public pension spending varies significantly across EAC countries (Table 2.2). This largely reflects differences in the coverage of the pension systems and their generosity. Only Kenya has a universal pension payment for the elderly older than age 65. In other EAC countries, pension payments by the government are mostly directed to retired civil servants and cover old age and disability. Most government schemes are operated on a defined benefit basis. The systems in Burundi, Rwanda, and Tanzania include an element of contribution from both the employee and the government. The ones in Kenya and Uganda are noncontributory, thus representing a bigger strain on government finances. The architecture of the public pension system varies across the EAC countries (Box 2.5). All of the EAC countries also have privately managed pension funds and other voluntary schemes.

^{Table 2.2}.

Pensions Expenditure in EAC

(percent of GDP)

Source: IMF staff estimates based on authorities’ data.

^{Table 2.2}.

Pensions Expenditure in EAC

(percent of GDP)

	2010–11	2011–12	2012–13	2013–14
Burundi	0.28	0.35	0.34	0.36
Kenya	0.46	0.54	0.55	0.60
Rwanda	0.22	0.23	0.23	0.25
Tanzania	0.19	0.20	0.17	0.15
Uganda	0.58	0.51	0.52	0.41

Source: IMF staff estimates based on authorities’ data.

View Table

^{Table 2.2}.

Pensions Expenditure in EAC

(percent of GDP)

	2010–11	2011–12	2012–13	2013–14
Burundi	0.28	0.35	0.34	0.36
Kenya	0.46	0.54	0.55	0.60
Rwanda	0.22	0.23	0.23	0.25
Tanzania	0.19	0.20	0.17	0.15
Uganda	0.58	0.51	0.52	0.41

Source: IMF staff estimates based on authorities’ data.

Public pension plans will represent an increasing strain on government finances. Pension spending is indeed bound to grow because of such demographic developments as an increase in the number of retirees and improvements in life expectancy, resulting in pensioners having longer retirement. These trends will directly affect the budget in the case of noncontributory schemes, but they could also indirectly affect the budget in contributory schemes if the contribution rates or benefits are not adjusted.

A broader issue is the potential impact of low coverage in a context of increasing life expectancy. The demand for pension coverage was low when life expectancy was low; however, as life expectancy improves, one can expect an increase in such demand, particularly for an old-age scheme where not available (all countries but Kenya), with possible fiscal implications.

Another risk is lower-than-expected returns on pension fund assets. Most of the current portfolios are invested domestically, mostly in government bonds. There is also generally limited investment expertise within the pension schemes. As a result, most of the schemes have experienced low investment returns over the years. There have also been cases of mismanagement, which have impacted negatively the long-term financial health of some schemes.

A. Quality, Timeliness, and Coverage of Reported Fiscal Data

When financial data quality is low, governments risk making inappropriate fiscal decisions. A related problem is that timely data coverage is often restricted to central government, with little available data on the wider public sector, such as subnational government and state-owned enterprises (see Chapter 2.A on the latter). This creates the risk that fiscal pressures can develop in these sectors without the knowledge of policymakers, which limits their ability to address them early.

Figure 3.1 presents the results of an analysis of data and reporting practices in each EAC country, undertaken by IMF staff in 2013. Current practices are then compared with benchmarks and classified in three color-coded categories: green representing that the benchmark is met, red indicating significant weaknesses, and yellow in between. This analysis points to risks related to both data quality and coverage in EAC countries, in particular the following:

Data Quality and Reporting Practices in EAC Countries, 2013

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Source: IMF staff analysis

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Data Quality and Reporting Practices in EAC Countries, 2013

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Source: IMF staff analysis

• Download Figure
• Download figure as PowerPoint slide

Data Quality and Reporting Practices in EAC Countries, 2013

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

Source: IMF staff analysis

• Download Figure
• Download figure as PowerPoint slide

• None of the EAC countries meet the benchmarks related to independent production and audit of financial data.

• In none of the countries are financial data produced by an independent statistics agency; only in Tanzania are data audited to international standards by an independent auditor.

• Only in Rwanda are data presented on a general government basis in the budget.

B. Expenditure Allocation and Control

If expenditure is greater than that envisaged by the budget, then—other things equal—the deficit will be higher than forecast. This may occur because of unforeseen external events—for example, a natural disaster that creates additional demand for public expenditure compared with budget allocations. However, expenditure overruns may also stem simply from weak internal controls.

Figure 3.2 shows the average difference between medium-term expenditure forecasts and outturns in Kenya, Uganda, and Tanzania over the 2000–12 period, with South Africa also presented for comparison (a comparative data set for Rwanda and Burundi is not available). This shows that both the current year and the medium-term expenditure projections have not provided a strong guide to outturns in these countries. Actual expenditure exceeded the one-year and two-year ahead forecasts over this period by a significant degree, particularly in Tanzania.¹⁶ This increases the risk that deficits will be higher than forecast and also makes it difficult for agencies to plan higher-than-expected levels of public expenditure, increasing the likelihood of wasting resources on inefficient expenditure or allocating it to nonpriority areas—thereby reducing efficiency and effectiveness. By contrast, the budget-year estimates of expenditure in these countries have, on average, been higher than outturns, in particular in Uganda and Tanzania. This outcome suggests that in-year expenditure plans are not realistic and that budget execution is weak. By comparison, in South Africa the initial budget estimate is, on average, a good guide to actual expenditure.

Medium-Term Expenditure Forecasts versus Outturn

(average 2000–12)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Medium-Term Expenditure Forecasts versus Outturn

(average 2000–12)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

Medium-Term Expenditure Forecasts versus Outturn

(average 2000–12)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

PEFA assessments provide useful indications of areas of institutional weaknesses across countries.¹⁷ Figure 3.3 shows the results from each EAC country’s most recent PEFA assessment of the items related to expenditure control. For comparison, this is shown alongside the average worldwide PEFA country score for each item. This figure shows that Uganda and Burundi have particularly low assessments against most areas of institutional expenditure control.

EAC Country PEFA Scores on Expenditure Control

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

EAC Country PEFA Scores on Expenditure Control

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

EAC Country PEFA Scores on Expenditure Control

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

Weak expenditure controls can also lead to the buildup of arrears, which occurs if expenditures are incurred but there is no available financing from revenues or from borrowing. Figure 3.4 illustrates a significant problem of arrears in Burundi and Tanzania, based on available data from 2010–11.

Arrears in EAC Countries, 2010–11

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Arrears in EAC Countries, 2010–11

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

Arrears in EAC Countries, 2010–11

(percent of GDP)

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

C. Oversight of Local Governments—Devolution

Fiscal devolution—whereby subnational authorities are given the authority to raise revenue, spend, and borrow—is significant in a number of EAC countries; it is expected to increase substantially in Kenya (Box 3.1). In principle, devolution helps better reflect the preferences of the citizens in the decision-making process and thus allows public spending to better address needs. Devolution may also increase flexibility in policymaking. For these reasons, devolution is expected to increase the efficiency and effectiveness of government expenditure, including on social services, and to enhance social cohesion and accountability.

In Kenya, Rwanda, and Uganda, local governments execute about 17 percent of general government expenditure on a releases basis; in Tanzania this share exceeds 20 percent (Figure 3.5).

Composition of General Government Expenditure, FY2013/14

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Composition of General Government Expenditure, FY2013/14

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

Composition of General Government Expenditure, FY2013/14

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

The decentralization process is complex and poses a number of challenges. It is therefore essential to ensure that it is properly sequenced and implemented, with due attention to institutional arrangements to limit fiscal risks. To be successful, the process requires all levels of government to be effectively committed to fiscal discipline in order to avoid the “common pool” problem—whereby incentives to control local spending are diminished because resources are raised nationally—and moral hazard—the perception that the central government will bail out the local administrations in trouble—as these problems might lead to overspending and overborrowing. Another risk is the duplication of functions.

International experience has shown that there are different approaches to controlling these risks. Sound institutions and procedures should ensure sufficient coordination and consultation, clarity of roles and responsibilities, adequate accounting and reporting, monitoring, auditing, and strong accountability arrangements. To prevent overborrowing, borrowing limits can be set through negotiated agreements, national legislation, or administrative controls. Other ways to help control these risks include: establishing an adequate framework for subnational borrowing to limit indebtedness, moving toward a unified treasury single account, and involving local governments in the scope of the PFM legal and regulatory framework, including the expansion of the fiscal reports to cover all of the public sector.

D. Revenue Forecasting and Administration

As noted earlier, an important source of fiscal risks relates to the possibility that government revenues will fall short of forecasts if underlying economic activity turns out to be weaker than expected. In some circumstances, weak revenue administration can also represent a related institutional source of fiscal risks, implying that revenues collected turn out to be a lower-than-expected proportion of the available tax base. In many circumstances, weaknesses in revenue administration are well known when the forecasts are made; however, these weaknesses can become a significant source of risk when a new tax or other reform is introduced to the existing tax regime. In some instances, there is also a systematic bias toward basing the budget on overly optimistic assumptions, which can be seen as another institutional source of fiscal risks.

Regarding revenue administration more specifically, Figure 3.6 shows the results from each EAC country’s most recent PEFA assessment, compared with the full set of PEFA scores. This shows that Kenya and Rwanda had particularly low assessments in the effectiveness of tax payments collection, whereas Tanzania received low scores on the effectiveness of taxpayer registration and assessment.

EAC Country PEFA Scores on Revenue Administration

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

EAC Country PEFA Scores on Revenue Administration

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

EAC Country PEFA Scores on Revenue Administration

Citation: Departmental Papers 2015, 009; 10.5089/9781513551265.087.A001

• Download Figure
• Download figure as PowerPoint slide

E. Risks Related to Natural Resource Wealth Management

Some EAC countries have recently made natural resource discoveries, mostly oil and gas (Box 3.2). The revenues associated with these natural resources, if they materialize, could be transformational for the concerned economies. Given that these potential revenues are not included in fiscal framework, they clearly constitute positive fiscal risks.

However, natural resources are also a source of macroeconomic and fiscal risks, and need to be properly managed even before production starts to ensure fiscal sustainability. For example, the fiscal position could deteriorate if borrowing increases are based on expectations for natural resource revenues and those revenues do not materialize. Given the large fluctuations in their prices, natural resources introduce a significant source of volatility in revenues once production starts, and that volatility can lead to boom-and-bust cycles. Furthermore, natural resources are exhaustible, and there are important trade-offs as to how much should be spent now or saved for future generations. Risks also include Dutch disease or potential diversion of resources.

These risks can be addressed by establishing a sound and transparent policy framework that constrains discretion and promotes fiscal discipline while isolating the budget from price volatility. These mechanisms need to work within a general context of sound public financial management. Institutions mechanisms could include medium-term fiscal rules, resource funds, and fiscal responsibility laws.¹⁸

In addition to creating challenges for macroeconomic management, the arrival of significant natural resource revenues will also test the institutional budget framework. In particular:

• The volume of additional revenues can erode aggregate budget discipline without a strong medium-term expenditure framework and enforceable prohibitions on expenditure in excess of appropriated amounts.

• The volatility of that revenue poses a problem of ensuring that the related expenditure is planned and spent efficiently and productively.

• The many different streams of revenue create challenges around ensuring that resources are all spent in line with the government’s strategic objectives, with approval and oversight from parliament.

A first step in preparing for these challenges is to ensure that a robust medium-term budget framework is in place before any revenues arrive. This reinforces the importance of reforms that aim to strengthen budget implementation, execution, and reporting. The volatility and uncertainty associated with natural resource revenues also underscores the need to develop a medium-term approach to budgeting, with forward expenditure ceilings linked to the medium-term fiscal policy objectives.

A second step would be to prepare the framework specifically to deal with significant new revenue flows; a key principle should be that the flows are fully integrated into the budget. This will ensure the integrity of resource revenue flows and protect the budget’s role as the mechanism for setting expenditure priorities and allocating public resources. All spending from resource revenues should be subject to the same scrutiny and accountability, including parliamentary oversight, as any other public spending. The use of revenues should be governed by a well-designed macro-fiscal framework based on a set of long-term fiscal policy principles. For example, such principles could emphasize the primary aim of using resource revenues for the development of the economy, while maintaining fiscal sustainability.¹⁹ In addition, they could stress the need for fiscal policy to protect against volatility and to effectively guide the choice between consumption, investment, and saving of resource revenues