Botswana
Technical Assistance Report-Public Investment Management Assessment

This Technical Assistance Report presents an assessment of public investment management in Botswana. Botswana’s public investment has been consistently high for the past 25 years coupled with prudent fiscal policy and moderate debt at about 10 percent of GDP. Relatively high public investment spending has contributed to steady accumulation of capital stock—almost three times more per capita than peers and emerging market averages. More than 60 percent of investment is spent on economic affairs. While Botswana performed better than peer and emerging market averages up to 2010, more recently, indicators of infrastructure quality suggest significant bottlenecks, particularly in access to electricity supply and railways.

Abstract

This Technical Assistance Report presents an assessment of public investment management in Botswana. Botswana’s public investment has been consistently high for the past 25 years coupled with prudent fiscal policy and moderate debt at about 10 percent of GDP. Relatively high public investment spending has contributed to steady accumulation of capital stock—almost three times more per capita than peers and emerging market averages. More than 60 percent of investment is spent on economic affairs. While Botswana performed better than peer and emerging market averages up to 2010, more recently, indicators of infrastructure quality suggest significant bottlenecks, particularly in access to electricity supply and railways.

I. Trends in Public Investment

A. Trends in Total Public Investment and Capital Stock

1. Public investment has been a key element of Botswana’s fiscal policy for the last three decades (Figure 1). From 1990 to 2006, public investment averaged 11.4 percent of GDP, almost doubling the average real GDP growth rate during those years (4.6 percent). Public investment peaked in 2009 at 17 percent of GDP, on account of economic stimulus measures, and gradually declined to 8.4 percent of GDP by 2014. Despite strong public investment during this period, GDP remained highly volatile driven mostly by the impact of the mining industry. In the second half of 2015, a new economic stimulus program including a significant scale-up of public investment was implemented to counteract the economic slowdown resulting from lower mineral revenues.

Figure 1.
Figure 1.

Botswana: Public Investment and GDP Growth

(In percent of GDP and percent change)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Public investment coverage corresponds to general government.

2. Despite a high dependency on mineral revenues, Botswana has managed to successfully implement countercyclical fiscal policy, while keeping low public debt ratios.2 Prudent fiscal policy combined with formal debt limits allowed the government to build fiscal buffers during good times and spend them judiciously during bad times.3 As a result, between 1990 and 2006 the government fiscal surplus averaged 4.8 percent of GDP, with public debt ratios at 10 percent of GDP, on average (Figure 2). Since 2007, the fiscal position deteriorated reaching a fiscal deficit of −14.3 percent of GDP by 2009, in part due to fiscal stimulus measures in response to the 2008 global financial crisis. Public debt peaked at 20.3 percent of GDP in 2012, and has declined to 15.6 percent of GDP by end-2016.

Figure 2.
Figure 2.

Botswana: Public Debt and Deficit

(In percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Public investment coverage corresponds to general government.

3. By 2015, public spending ratios in Botswana were significantly high—both current and capital—even when compared to emerging market economies (Figure 3). Notably, the share of current spending has risen consistently since 1990 showing a positive trend during almost the whole period, except when large economic stimulus plans (ESP) were implemented (Figure 4).

Figure 3.
Figure 3.

Current vs. Capital Public Spending, Comparison to Peers, 2015

(In percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Coverage of the public sector corresponds to general government for all countries.
Figure 4.
Figure 4.

Botswana: Composition of Public Spending, 1990–2015

(In percent of total)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Coverage of the public sector corresponds to general government for all countries.

4. Consistent with relatively high public investment spending over the past three decades, Botswana’s estimated public capital stock, as a percent of GDP, is high relative to peer countries.4 Up to 2010, Botswana spent on average about twice as much on public investment compared to the average of emerging market economies (EMEs) and peer countries, even though this differential has subsequently narrowed (Figure 5). As a result, Botswana’s public capital stock, estimated at 125 percent of GDP by end-2015, stands above the average of its peers (Figure 6). Similar results are observed on a per capita basis, where Botswana has almost doubled its public capital stock per capita since 1990 (Figures 7 and 8).5

Figure 5.
Figure 5.

Public Investment Trends: Comparison with Peers

(2005 PPP$-adjusted, percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Public sector coverage corresponds to general government for all countries.
Figure 6.
Figure 6.

Public Capital Stock, Comparison with Peers, 2015

(2005 PPP$-adjusted, percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Public sector coverage corresponds to general government for all countries.
Figure 7.
Figure 7.

Botswana: Public Capital Stock per Capita

(2005 PPP$-adjusted, percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Public sector coverage corresponds to general government for all countries.
Figure 8.
Figure 8.

Public Capital Stock per Capita: Comparison with Peers, 2015

(2005 PPP$-adjusted, percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Public sector coverage corresponds to general government for all countries.

5. Botswana’s public investment is quite volatile compared to peers, driven mostly by the impact of the mining industry (Figure 9). Moreover, since 2006, the execution rate of the capital budget ranged between 73 to 97 percent of the revised budget (Figure 10). Typically, relatively low execution, as a percentage of the budget, is a sign of poor budget planning at the project level, but might not necessarily affect public investment efficiency per se. On the other hand, while in Botswana budget allocations for public investment are often revised upward during the year, these revisions might allow some projects that are ready to go forward, contributing to reducing the overall under-execution rate (i.e., avoiding that projects that are ready, delayed, or cancelled due to insufficient budget appropriations).

Figure 9.
Figure 9.

Comparison of Volatility of Public Investment1

(Average 2010-13)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Coverage of public sector corresponds to general government for all countries.1 Volatility is calculated as the standard deviation of year-on-year growth of investment to GDP ratios (real, deflated, and ppp=adjusted).
Figure 10.
Figure 10.

Botswana: Execution of Capital Expenditure

(Percentage executed, relative to revised budget)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data. Coverage of public sector corresponds to general government for all countries.

B. Composition of Public Investment

6. Almost two thirds of Botswana’s public investment is devoted to economic infrastructure, compared to broadly half in EMEs (Figures 11 and 12). Between 2010–15 the allocation of public investment to economic infrastructure (e.g., roads, ports, airports, electricity) was much higher in Botswana relative to EMEs. On the other hand, social infrastructure (e.g., housing, hospitals, schools, recreation, social protection) accounts for a similar share in both cases, representing approximately one fourth of total capital expenditure.

Figure 11.
Figure 11.

Botswana: General Government Investment by Function, Average 2010–15

(In percent of total general government investment)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data.Note: In line with COFOG classification methodology. Economic infrastructure is proxied by economic affairs and includes public investment for transportation infrastructure, among other components. Social comprises public investment in education, health, housing, social protection, and recreation and culture. Other includes public investment for general public services, safety and public order, and environment.
Figure 12.
Figure 12.

Emerging Market Economies: General Government Investment by Function, Average 2010–15

(In percent of total general government investment)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data.Note: In line with COFOG classification methodology. Economic infrastructure is proxied by economic affairs and includes public investment for transportation infrastructure, among other components. Social comprises public investment in education, health, housing, social protection, and recreation and culture. Other includes public investment for general public services, safety and public order, and environment.

7. Besides central government, SOEs—including parastatals—account for a sizeable share of public sector investment in Botswana. Between 2010–15, the central government provided about 65 percent of public sector investment, with SOEs and parastatals contributing about one-fourth, and subnational governments (SNGs) playing only a limited role at 10 percent (Figure 13).6 SOEs and parastatals provide much of the public investment in infrastructure assets, notably in the transport, energy, communication, water, and waste management sectors.

Figure 13.
Figure 13.

Botswana: Composition of Public Investment by Level of Government

(In percent of total public investment)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data.Note: In line with COFOG classification methodology. Economic infrastructure is proxied by economic affairs and includes public investment for transportation infrastructure, among other components. Social comprises public investment in education, health, housing, social protection, and recreation and culture. Other includes public investment for general public services, safety and public order, and environment.

8. Public investment in Botswana is mostly funded through domestic sources (Figure 14). In 2014, domestic financing of public investment represented 95 percent of the total. Foreign loans, both concessional and non-concessional, comprised the bulk of external financing sources, while external grants accounted for 20 percent of total external funding sources for public investment, on average.

Figure 14.
Figure 14.

Botswana: General Government Investment by Sources of Funding, 1990–2014

(In percent of total general government investment)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: WEO and staff estimates based on official data.Note: In line with COFOG classification methodology. Economic infrastructure is proxied by economic affairs and includes public investment for transportation infrastructure, among other components. Social comprises public investment in education, health, housing, social protection, and recreation and culture. Other includes public investment for general public services, safety and public order, and environment.

9. Botswana has not yet made extensive use of PPPs, but many projects are in the pipeline (Figure 15). In the 2002/03 Budget Speech and the NDP 9, the government announced that: “PPPs would be used extensively as a form of procuring and financing infrastructure projects in the public sector to ensure sustainable investment in infrastructure, as well as to restore soundness in public finances and bring down the budget deficit to a sustainable level.”7 Despite these policy intentions, the government started using PPPs later than many peer countries, with the first PPPs implemented only in 2011.8 By the end of 2015, the estimated stock of PPPs represented 0.5 percent of GDP, much lower than in most comparator countries. Existing PPP projects in Botswana are used for a public office and power purchase agreements in the electricity sector. More recently, the government has taken actions to promote PPPs in many of the key economic sectors.

Figure 15.
Figure 15.

Public-Private Partnerships Capital Stock, Comparison with Peers, 2015

(2005 PPP$-adjusted, percent of GDP)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: Staff estimates based on official data.

II. Efficiency and Impact of Public Investment

10. Botswana’s overall quality of infrastructure has fallen sharply in recent years. Up to 2010, survey-based indicators of perception of infrastructure quality suggested that Botswana compared favorably to regional peers and EMEs in the perception of overall infrastructure quality (Figure 16). Later on, indicators suggest that structural bottlenecks in electricity, and railroads are the drivers of the sharp deterioration in the perception infrastructure quality observed in the last five years (Figure 17).

Figure 16.
Figure 16.

Perception of Infrastructure Quality, 2006–15

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Figure 16: World Economic Forum (2015).
Figure 17.
Figure 17.

Botswana: Infrastructure Quality, Decomposition, 2006–15

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Figure 17: World Development Indicators (2015).

11. Similarly, physical indicators of access to infrastructure and service delivery show large variations in the provision of economic and social infrastructure by end-2015 (Figure 18). While Botswana generally performs better than its peers in access to social infrastructure (i.e., schools, water, and to a lesser extent hospitals), service delivery of economic infrastructure in the electricity sector is quite poor (Figure 19). By contrast, Botswana scores much better than peers, and even EMEs, in the delivery of roads.

Figure 18.
Figure 18.

Measures of Infrastructure Access and Service Delivery, 2015

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Figure 18: Staff estimates based on official figures.Note: Figure 18. Units vary to fit scale. Left hand axis: Public education infrastructure is measured as secondary teachers per 1,000 persons; electricity production per capita as thousands of KWh per person; total road network as km per 1,000 persons; and public heath infrastructure as hospital beds per 1,000 persons. Right axis: Access to treated water is measured as percent of population.
Figure 19.
Figure 19.

Power Generation Capacity, Regional Comparison, 2006

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Figure 19: Africa’s Power Infrastructure: Investment, Integration, Efficiency, World Bank, 2011.

12. Indicators of infrastructure access and service delivery point to inefficiencies in the provision of electricity and, to a lesser extent, in railways. In particular:

  • Botswana remains energy import dependent, even after significant public investment in the sector, which underscores the need for further investment in electricity generation. The energy sector is extremely vulnerable, despite significant recent improvements. In 2012 Botswana imported 93 percent of electricity from neighboring countries, decreasing to 39 percent by 2015. The small size of the domestic market and the dispersion of the sparse population reduces potential economies of scale, and raises the cost of supplying electricity. To tackle low levels of domestic power generation (Figure 20), the BPC has undertaken several infrastructure projects since 2006, including the promotion of renewable energy. Nevertheless, it has experienced significant challenges, including limited procurement and project management capacity. As of 2015, electricity production capacity remains below peer countries, while only 66 percent of the total population had access to electricity, dropping to 54 percent in rural areas.

  • Railways construction and maintenance services are provided by Botswana Railways, which operates a de facto monopoly under the Ministry of Transport. Despite a substantial increase in the volume of services, sustained lack of proper maintenance resulted in accelerated depreciation of fixed assets, hampering financial performance of the company. In addition, the existing tariff system is quite rigid, further raising the risks of future underinvestment and inadequate maintenance.

  • Although the indicator used in estimating the public investment efficiency frontier does not suggest bottlenecks in the water sector (i.e., access to treated water as percent of total population), significant challenges exist in terms of coverage and pricing. Access to irrigation water in rural areas is limited to farmers and isolated population, while connection charges are considered too high-limiting user connection to pipeline water systems. Moreover, Botswana has recently faced significant reduction in volume of water in its dams, partially due to climate changes.9

13. Despite relatively high investment levels, Botswana has a significant public investment efficiency gap, underscoring the need to improve public investment management. Botswana scores below the most efficient countries with comparable levels of public capital stock per capita.10 Public investment efficiency is lower than the average in EMEs and the world. Based on the methodology outlined in the IMF Board Paper, the efficiency gap for Botswana is calculated at 37 percent compared to an average of 27 percent for the world and EMEs (Figures 20 and 21). These results suggest that about one third of Botswana’s public capital stock did not result in the maximum level or quality of infrastructure assets or service delivery (i.e., efficiency loss). Therefore, there is significant scope for improving public sector investment efficiency to take full advantage of its impact on economic growth.

Figure 20.
Figure 20.

Efficiency Frontier, Quality Indicator, 2008–15

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Staff estimates.Note: Figure 20: Methodology is indicated in IMF (2015). Public sector coverage corresponds to general government.
Figure 21.
Figure 21.

Efficiency Gap, Quality Indicator, 2008–14

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Staff estimates.Note: Figure 21: The box shows the median as well as the 25th and 75th percentiles, while the whiskers show the maximum and minimum values. The black square shows the average.

14. Under the current fiscal policy framework, improving public investment efficiency could potentially create additional fiscal space for addressing the existing infrastructure bottlenecks, without jeopardizing debt sustainability. A targeted approach to address existing infrastructure weaknesses would reduce pressures on the government balance. Going forward, achieving the authorities’ infrastructure goals and tackling bottlenecks in the implementation of public investment projects will require addressing key drivers of public sector inefficiencies. Developing a robust public investment framework, and strengthening the institutions will be crucial in addressing these inefficiencies, as discussed later in this report.

III. Public Investment Management Institutions

A. Overall Assessment

15. Botswana’s institutions for managing public investment compare reasonably well to other EMEs. Figure 22 compares Botswana’s ratings to other countries. National and sectoral planning as well as fiscal rules are stronger for Botswana than its comparators, but PPPs, company regulation—and to a lesser extent—investment by subnational governments score less well than comparators. The development budget is reasonably comprehensive and includes all donor-funds, but not PPP transactions. Project appraisal, project selection, and integration of capital and recurrent budgets require further work. Although the scores for availability of funds and protection of investment exceed comparators, project management is weaker than in comparator countries.

Figure 22.
Figure 22.

Strength of Public Investment Management Institutions

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Sources: Staff estimates based PIMA assessement for Botswana, and on the database developed in line with the IMF’s Board Paper “Making Public Investment More Efficient,” IMF, 2015.

B. Investment Planning

I.1. Fiscal rules (Strength –Medium; Effectiveness –Medium)

16. Botswana has a debt rule and expenditure principles in place to safeguard fiscal sustainability and guide internal decision making. The debt rule caps total domestic and foreign debt each at 20 percent of GDP. There is no provision in the law for rules to be suspended or readjustment mechanisms if exceeded.11 The expenditure principle includes a 40 percent of GDP government spending cap introduced in NDP 9 and a target reduction of government spending to 30 percent of GDP by 2016 stated in NDP 10.12 Within the total expenditure limit, as a guideline, 30 percent of expenditure is targeted for capital in efforts to prioritize capital investments. These principles are used to frame discussions during the budget but, ultimately, spending needs dictate decisions and this is not intended to be binding.

17. The debt rule and expenditure principles have served their purpose although the capital/recurrent ratio has not generally been followed. The debt rule has been adhered to since enactment in 2005 with total debt peaking at 20.3 percent of GDP in FY2012/13 driven predominantly by external borrowing, which accounted for 19.3percent.13 The same is true for the 40 percent expenditure principle, although this was breached over a three-year period from 2008–10, due to the implementation of a stimulus package to limit the impact of the global financial crisis. The target to reduce total spending to 30 percent of GDP was not met by the end of NDP 10 (Figure 24). The capital/recurrent spending ratio of 30:70 has only been met twice since implementation of NDP 9 (Figure 23).14

Figure 23.
Figure 23.

Botswana: Total Spending to GDP

(percent)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Botswana Statistics and Ministry of Finance and Economic Development.
Figure 24.
Figure 24.

Botswana: Capital Budget Shares

(percent of total budget and outturn)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: Botswana Statistics and Ministry of Finance and Economic Development.

I.2. National and sectoral planning (Strength –Good; Effectiveness –High)

18. A comprehensive NDP is published along with the Public Investment Plan (PIP), which provides six year projections for the estimated Total Economic Cost (TEC) by project. Sector and subsector strategies are integrated within in the NDP and are linked to broader policy goals (thematic policy areas). Plans are underway to link project outputs in the PIP to these thematic policy areas, which have outcome targets for the duration of the NDP.

19. The inclusion of multi-year projections in the NDP provides an effective platform for investment funding and implementation. The NDP framework includes project codes, names and their estimated TEC, which provides a tangible bridge to the annual and multi-year budget and execution processes. This link does not exist amongst regional peers, where it is common for development planning and budgeting to be quite separate processes.

20. Whilst the mechanics of the planning process are in place, the credibility of cost estimates could be improved in the NDP through better project screening and costing. Adequate first level screening could be more thorough by determining whether projects or better suited as recurrent activities. It was estimated that 40 percent of the PIP is recurrent in nature and should not form part of the project pipeline.15 With regards to costing, an assessment of six of the largest infrastructure projects in NDP 10 show variations between the projected TEC for the first three years of the project and the actual spending (Figure 25).

Figure 25 .
Figure 25 .

Variances in Major Infrastructure Projects, 2009–10, 2011–12

(percent)

Citation: IMF Staff Country Reports 2017, 188; 10.5089/9781484307489.002.A001

Source: NDP 10 volume 2 and annual financial statements.

I.3. Central-local coordination (Strength –Medium; Effectiveness –Medium)

21. Local government development spending is coordinated by the Ministry for Local Government and Rural Development and comprises less than 10 percent of total government spending (Figure 12).16 Development spending priorities include construction of primary school facilities, teachers colleges, customary courts, rural administrative centers, and recreational parks. Operational costs, including maintenance spending is financed through the revenue support grant, which finances about 75 percent of priorities—own revenue amounts to 25 percent on average (depending on the council size and location).

22. There are some limits on local government borrowing and external borrowing is minimal. Local governments are allowed to incur overdrafts not exceeding one third of previous year’s revenue (excluding) grants with approval of Minister of Local Government and Rural Development (MLGRD) after consultation with the MFED. The outstanding loans for local urban councils amount to P 18.311 million as at end-2016 fiscal year. Urban councils used to receive funds under the Public Debt Service Fund (PDSF), and some councils are yet to repay the funds that have previously been advanced.17 The MLGRD does not monitor the performance of loans which could lead to future fiscal risks.

23. Transfers are published for each local authority per transfer before the start of the fiscal year based on NDP priorities. The budget book provides transfers to each district based on submissions of district plans, linked to the NDP.18 District priorities are based on community consultations and included in discussions of Technical Working Groups (TWGs) before funding decisions are made. District Development Plans (DDP) are aligned to the NDP. Local funding is predictable and investment is generally protected, albeit insignificant in overall spending.

I.4. Public-private partnerships (Strength –Medium; Effectiveness –Low)

24. In 2009, the government adopted a PPP strategy together with standard criteria for entering into PPP arrangements. The 2009 PPP Policy and Implementation Framework (the PPP policy) is broadly in line with good international practices, albeit with some weaknesses. The PPP policy provides clear guidelines and procedures for identifying, appraising, and selecting projects to be procured as PPPs. It requires that projects undertaken by central and subnational government entities be subject to value for money and budget affordability analysis by the PPP unit under the MFED, and that the tendering of projects be conducted on a competitive basis through the PPADB. It also prescribes the institutional framework governing PPPs, including the roles and responsibilities of all public entities involved in PPP contracts. Notably, the MFED has an approval role from early stages of the PPP cycle, with the capacity to terminate those projects that are deemed too risky or unaffordable. Moreover, the PPP policy requires that the MFED recognizes the financial commitments to pay unitary charges for the duration of the PPP project life cycle, although it does not prescribe any form of reporting.19

25. The lack of a dedicated PPP legal framework, gives rise to regulatory uncertainty, and might expose the government to excessive risks.20 Botswana still has no dedicated PPP act or regulations. A few PPPs have been procured in the last five years using the existing PPAD Act, which makes no explicit reference to PPP projects. Given existing challenges in implementing the PPP policy, the lack of a dedicated legal and regulatory framework for PPPs is particularly important. Main challenges generating regulatory uncertainty relate to the institutional coverage of the PPP policy, that is to say, which public entities are covered by the policy. Currently, compliance with the PPP policy is mandatory for central and subnational governments, while SOEs are simply encouraged to contract PPPs in line with the policy guidelines, but are not necessarily required to consult the PPP unit in the MFED.21

26. In practice, recent PPPs have been contracted without any involvement of the MFED’s PPP unit, even when they required government guarantees.22 As a result, the role of the MFED as gatekeeper of public finance prescribed by the PPP policy is significantly diluted, given that value for money and budget affordability are not guaranteed or evaluated using harmonized criteria. Also, the PPP unit is inadequately staffed to cope with a growing PPP portfolio (notably, there is no dedicated lawyer assisting the unit).23 In addition, fiscal risks of PPPs are neither systematically assessed nor reported by the MFED, despite being explicitly required by the PPP policy. PPPs are reflected in the budget or the MTFF only to the extent that they require government payments, with no additional information included in the budget documentation. Finally, the lack of ceilings for the accumulation of explicit or contingent liabilities from PPPs might leave the Government excessively exposed to risks from a growing PPP portfolio.

27. While Botswana’s PPP capital stock is low compared with peers in the region, several projects are in the pipeline, underscoring the need for strengthening their legal and institutional framework. The government has not yet made extensive use of PPPs, at least at the central government level. Capital stock of PPPs accounted for 0.5 percent of GDP at the end of 2015. Only two small government-funded projects were executed, namely a 10 year PPP for an office complex (Plot 21), and a 17 year PPP for the Southern African Development Community (SADC) headquarters. There has not been PPPs in the water and sanitation, ICT, and transport sectors. In 2007, the government amended the Electricity Supply Act to allow independent power producers (IPPs) to participate in the investment and supply of electricity in collaboration with the BPC. So far there is only one IPP in operation,24 and one IPP recently signed that will become operational by 2020.25 However, several projects are being developed, in line with NDP 10 (including cooperation with neighboring countries to tap into undersea fiber optic cables and power purchase agreements). The government plans to amend the PPAD Act, instead of enacting a separate PPP law, to improve the legal and regulatory framework related to PPPs.

I.5. Regulation of infrastructure companies (Strength –Medium; Effectiveness –Low)

28. Domestic competition exists in some economic infrastructure markets despite the potential for contestable markets, domestic monopolies dominate. The rate of privatization has been slow in Botswana since the adoption of the 2000 policy. The telecommunications sector is the only fully liberalized sector that has an autonomous regulatory authority and legislation has been tabled to introduce water and energy regulators. Prices in the other sectors are set centrally through the parent ministry following Cabinet approval—but do not always cover the full cost of operation e.g., in the water sector and often require transfers from government to the water utility. Domestic monopolies exist in the energy, transport, and water sectors (Table 2). A consolidated financial SOE report exists for 15 SOEs, but does not cover investment plans and operational targets.

Table 2.

Botswana: Competition and Regulation in Infrastructure Sectors

article image
Source: Mission.

29. International surveys suggest that barriers to competition and market entry may exist. Botswana’s 2017 doing business survey whose ranking fell for 8 out 11 indicators compared to the 2016 survey.26

30. SOEs play a large and strategic role in the economy yet reporting on SOE investment plans and operational targets is limited. SOE investments financed through capital grants alone are estimated at 1.5 percent of GDP based on average spending between 2010 and 2015. The consolidated SOE report produced by the PEEPA currently only dedicates three paragraphs to operational performance and includes no details of planned investments nor a statement of corporate intent for major infrastructure SOEs.

C. Investment Allocation

I.6. Multiyear budgeting (Strength –Good; Effectiveness –Medium)

31. Estimates of the development budget by ministry and project are only published for the budget year, but the TEC provides a global ceiling for ministries to commit to contracts over a multi-year horizon. The TEC is derived from the six-year projections in the NDP, of which three-year projections are communicated at the ministry level through the budget call circular.27 The breakdown of full project cost details does not appear in the development budget estimates, but is used for internal decision making at the Cabinet level and to derive the MTFF, which is published in the budget in brief document.

32. Ceilings are indicative, and can provide flexibility for ministries to commit funds to different projects based on updated information, but may “squeeze” funding for other projects. Given the nature of the Botswanan planning and budgeting process, there are pros and cons to flexible ceilings. Analysis of project costing in the NDP28 shows that costing and provisional gatekeeping were not rigorous at the planning stage—yet this is essential when projects are selected for a multi-year period.29 In this context, the advantage of indicative ceilings allows for flexibility to adjust annual and multiyear ceilings at the time of budgeting based on better informed information on cost estimates.30 Disadvantages of this approach center on the potential risk of some projects receiving less funding during annual negotiations due to substantial cost increases of others.31

33. Publishing indicative medium-term projections by ministry and project would allow for more transparency on project selection to support the budget negotiation process. Without this information, it is unclear whether movements across financial years are due to efficiency improvements to support the completion of projects or for different reasons. Multiyear projections could support the budget negotiation phase with ministries and agencies having to justify movements from one financial year to another.

I.7. Budget Comprehensiveness (Strength –Medium; Effectiveness –Medium)

34. All development spending is appropriated in the budget. Although a number of special funds exist, only the Domestic Development Fund (DDF) and Tertiary Education Development Funds are used for development spending—all funds are appropriated in the budget. Parastatals are responsible for about 20 percent of development spending and their estimates are provided in separate reports by PEEPA. In addition, the budget estimates contain transfers to parastatals on development spending.

35. Externally funded projects are negligible in proportion to overall spending. Donor-funded projects comprise about three percent of overall development funding and donor-financed loans are included in the NDP. External resources include bilateral and multilateral loans, and miscellaneous and multilateral grants. PPP transactions for plot 21 and power purchase agreements are not accurately quantified in budget documents (refer to institution 4 and recommendation 6).

I.8. Budget Unity (Strength –Medium; Effectiveness –Low)

36. MFED has acknowledged inadequate coordination of the development and recurrent budget preparation, and has taken some steps to address it. The development and recurrent budgets are presented together but prepared by different units in the MFED. In an effort to improve coordination between the two budgets, MFED combined the Technical Assessment of Ongoing Projects and Baseline Budget Review meetings in the strategic phase of the budget process, which formerly were conducted separately. These joint meetings occurred for the first time in June/July 2016. In a similar manner, the Project Review and Expenditure Review meetings, which occur in October for review of the detailed budget submission, have been combined beginning in 2016. Such joint meetings are especially important given that the government aims for a ministry budget as a whole to be approximately 30 percent development and 70 percent recurrent. The economic classification, which applies only to the recurrent budget, contains some capital items in the recurrent portion of the classification, and vice versa. The budget classification does not include a program segment.

37. Coordination between capital and recurrent budgets is generally less effective than it would appear. First, the development budget includes expenditures that have a clear beginning and ending, and thus may be recurrent or capital in nature. Overall, the development budget is typically 40 percent recurrent and 60 percent capital.32 Since the development budget does not have an economic classification, it is not possible to easily measure the size of capital and recurrent planned expenditures, which impedes the monitoring of capital stock. Second, estimates of recurrent costs are required in the NDP, and in the project memoranda on which the development budget is built. However, such estimates are commonly prepared by Development Programs staff based on formulas, such as a fixed percent of the initial cost of a facility. Budget Administration, responsible for the recurrent budget, is not directly involved in estimating future recurrent costs in either case.

38. The separation of the development and recurrent budgets appears to be masking important underlying mismatches of resources. Because the budget classification does not allow for easy measurement of the two types of expenses, the size of the mismatch is not evident. It is clear, for example, that maintenance has been neglected, as special maintenance projects are being included in the development budget of NDP 11, a good example of a recurrent expense policy initiative appearing in the development budget.

I.9. Project appraisal (Strength –Good; Effectiveness –Low)

39. Project appraisal occurs through the project memorandum (PM), which is prepared after the project is approved in the NDP. Appraisals are guided by the Planning Officers Manual (POM), last updated in 1996. They are prepared by implementing agencies when the project is ready for budget funding. Chapter 6 of the PO Manual provides guidelines for conducting cost-benefit analysis, but these are not always undertaken rigorously. Even a large project, such as the P 1.4 billion Mohembo Bridget Construction, does not appear to have been subjected to rigorous cost-benefit analysis, as described in the PO Manual, although such analyses have been conducted for other large projects in the past. Some appraisals have been conducted by consultants. Project memoranda are reviewed by Development Programs staff in the MFED, but staff do not commonly provide central support to line ministries when project memoranda are being prepared. Neither project memoranda nor cost-benefit analyses are published. The PO Manual addresses risk, but generally focuses on financial risk, such as exchange rate risk. It does not address extensively non-financial risks such as engineering complexity and benefits derived from the project being less than expected. The PO Manual does not define a project.

40. Projects included in the NDP are commonly viewed as having been approved, which tends to undermine critical appraisal. This is especially true for relatively small projects, for which project memoranda are prepared by implementing agency staff. Rigorous appraisals are more likely for large projects when consultancies are used to conduct the appraisal before the project is submitted for budget funding. The outcome of such studies should be whether the project should move toward full feasibility study or implementation. The fact that appraisals are conducted when the project is ready to be submitted for budget funding means that there is not a pipeline of appraised but not selected projects, and it tends to shorten the period of time for the project memoranda to be prepared before the budget preparation process begins. The short preparation period is especially problematic for costing, which consistently results in actual costs exceeding estimates, as evidenced by the large number of project addenda to increase the TEC of many projects.

41. Poor costing undermines project selection and budget planning. If costs are not estimated accurately, the conclusion of cost-benefit analysis will be distorted, possibly resulting in inefficient projects going forward. Contingencies are typically not included in project costing. In addition, budget planning is affected if necessary spending is higher than planned. In particular, supplementary budgets may be required if funds are insufficient to meet the programmed scope of work.33 Inadequate costing of capital and recurrent costs of large projects can have significant fiscal consequences, both in the construction phase and after operations begin. The “African Property Construction Cost Guide 2016” issued annually by AECOM provides some guidance on standard infrastructure cost.

I.10. Project selection (Strength –Medium; Effectiveness –Low)

42. Projects are typically selected in the planning process, not the budget process. The process for including projects in the NDP is based primarily on project concepts, their adherence to policy, and the anticipated contribution to development strategies. They are not selected following a rigorous appraisal through which scope, risks, and cost are refined, and whether the project will realize the anticipated contribution to development. In effect, there is no pipeline of projects from which to select for budget funding other than projects approved in the NDP and their sub-projects. Implementing agencies are primarily responsible for proposing which NDP projects and sub-projects should proceed each fiscal year, although their proposals are reviewed by the MFED Development Programs’ staff and approved in the budget process. Proposed projects are rarely scrutinized by external experts prior to their inclusion in the budget. The PO Manual does not provide selection criteria separate from appraisal, but the MFED has issued guidelines on project selection which has not been followed. Some projects, not included in the NDP, are added in the budget preparation process for political or technical reasons, but these are a relatively small proportion of the total.34 For such projects, there is not sufficient time to prepare an appraisal, even if the outcome of that appraisal might be used to critically evaluate the proposed project.

43. It is difficult to deny budget funding for a project that is in the NDP. There are several reasons why a project in the NDP might be subject to critical review in the budget process. The projects may lack important pre-requisites such as a final feasibility study, land approvals during planning stage, and verification of supporting access roads, and water and electricity supply. Some high priority projects are added in the budget without having undergone any formal appraisal process, which could contribute to cost overruns and project implementation delays. Project memoranda are completed for projects, but information such as project expected program for construction work, monthly cash flows, and risk assessments are missing. Finally, conditions may have changed since the NDP was prepared that reduce the need for or benefits from the project. Strengthening the gate keeping role of the budget process could improve the efficiency of public investment. Appendix II provides an example of a gatekeeping role.

D. Investment Implementation

I.11. Protection of investment (Strength –Good; Effectiveness –Medium)

44. Procedures protect the total development budget for a ministry but do not ensure that individual projects will be implemented fully. NDPs include TEC for each project over the course of the plan period. Tranches, a mechanism that acts to commit funds on a multiyear basis provides assurance that funds will be available on which to sign contracts. However, a ministry, with the approval of the MFED, may shift TEC among projects, and thus tranches, so long as the ministry totals are not exceeded. Similarly, with regard to the annual budget, a ministry may establish warrants35 at amounts more or less than the project amount shown in the budget with the approval of the MFED, as long as the sum of warrants for the ministry does not exceed the ministry’s development budget. The Public Finance Management Act (PFMA) Section 36 (3) does not allow for transfers of funding between the approved development and recurrent budgets without the consent of Parliament. It also prohibits funding shifts between projects in the development budget without the consent of Parliament, but this provision is not enforced and such reallocations occur, within the scope of a single ministry, with the approval of the MFED. Thus, the development budget as a whole for a ministry is protected, but not the proposed spending for specific projects, either in the NDP or in the budget. Appropriations are made annually and there are no provisions for carry-forward of the unused balance of appropriations to the next fiscal year.

45. Substantial cost increases for some projects may reduce funding available for other projects. The updated TEC of development projects started is substantially higher than what was planned in NDP 10, with an overall increase of nearly 70 percent as of 2016. Poor costing in the NDP appears to be the principal reason for the increase, which means that projects that do not increase, effectively receive a cut. The cost of projects added in the annual budget process, and not in the NDP, is relatively small. Given that many of the NDP projects are a pool of smaller projects of the same type (e.g., Second Roads Construction) it is not clear how many of the smaller projects have been dropped.

Table 3.

Change in the Total Estimated Cost of Projects

article image
Note: i) percent changes are based on original NDP 10 data; ii) the Revised TEC is as shown in the 2016–17 Development Budget.Source: staff estimates.

I.12. Availability of funding (Strength –Good; Effectiveness –High)

46. Funding is reliably available to support expenditures primarily through a system of cash buffers. As noted earlier, tranches provide for multiyear commitment control, and warrants control payments. Most payments draw from the Remittance Account, which acts as a Treasury Single Account (TSA).36 The balance of the Remittance Account is maintained between 300 million and 700 million Pula. If needed, the Remittance Account is topped-up from the Liquidity Account, which holds enough money to cover 3 to 9 months of imports. If needed, the Liquidity Account is topped up from the Pula Account, which is the long-term national savings account. These accounts are held at the Bank of Botswana (BoB).37 Commodity-related revenue, primarily diamonds, upon which savings have accumulated, has declined in recent years. In anticipation of tighter fiscal policy in the future, the MFED introduced cash management reforms beginning in 2015, including creation of a six-person cash management unit and drafting a series of cash management procedures. Cash forecasts are prepared covering weekly, monthly, and quarterly cash flows. External financing from donors is channeled through the Remittance Account, and are tracked by means of Deposit Accounts established within the accounting system.

47. The cash management system is effective but expenditure control is experiencing some difficulties. There have been no instances in recent decades of cash not being available when needed. Expenditure control, however, is not complete and unauthorized expenditures have been noted in the Auditor General’s report,38 totaling approximately 3 percent of budgeted expenditures from the Consolidated Fund in FY 2014–15. Much of the over-expenditure occurs as a result of faulty integration of the Government Payroll, Pensions, and Passages System, and Government Accounting and Budgeting System (GABS), and affects the recurrent budget. The source of the problem is understood; the potential for substantial increase in unauthorized spending is low.

48. The system of cash management does not have high importance for new reform proposals from this mission. Capital project implementation is not impeded by lack of cash, and the system in place is robust. The Accountant General’s office has embarked on a cash management reform program, which is focused primarily on lowering the amount of idle cash as opposed to improving cash availability. Over-expenditures are not of a magnitude to threaten availability of cash to implement the development or recurrent budgets.

I.13 Transparency of execution (Strength –Medium; Effectiveness –Medium)

49. A regulatory framework for procurement incorporating competitive procedures is in place but does not apply to all SOEs. The PPAD Act of 2001 and associated regulations provide an open, competitive, and transparent legal framework for procurement. The Act requires open, competitive bidding procedures for all procurement transactions and applies these principles to all ministries, local authorities, and some parastatals. The PPADB is organized as a parastatal to support its independence. It adjudicates and awards tenders for central government, however it has devolved this mandate to the Tender Committee at ministry and district level depending on approved financial thresholds. PPADB handles the vetting of tender documents, and the adjudication and award of high value tenders beyond devolved amounts. It does not evaluate tenders as that is the mandate of the procuring entities. Under exceptional circumstances, the Board can appoint an Independent Evaluation Committee to evaluate tenders. A major part of its role is to oversee and strengthen the capacity of Ministerial Tender Committees, the tender committees of local authorities, and the contracting community, as well as monitor adherence to the PPAD Act. Many state-owned enterprises use their own procurement procedures, but they are expected to be aligned with the principles articulated in the PPAD Act. The PPADB publishes an annual report on its operations. The MFED appoints members of the Independent Complaints Review Committee that deals with procurement complaints if they are dissatisfied with PPADB resolution of disputes. The Committee is empowered to stop ongoing projects where it is clear that such projects will prejudice public interest.

50. Projects are monitored during the execution stage, but no formal reports are submitted. For all major projects, financial and physical progress are monitored during implementation by procuring entities. Financial monitoring was previously supported by the Development Project Management System (DPMS), but the system has not been operational since the transition from NDP 9 to NDP 10. In the absence of the DPMS, financial monitoring is enabled through GABS in a more limited fashion. Physical progress is monitored by each procuring entity responsible for the project, GICO with support from the Community-Based Monitoring (CBM) program, and implementing agencies. However, formal reports are not compiled to substantiate the observations of the monitoring entities, but site meetings are conducted for works projects. Such a report would typically identify risks (and courses of action to mitigate them), time constraints, and the sources of cost overruns. Appendix I contains a typical Project Management tool designed specifically for the monitoring of project execution.

51. Three types of ex-post audits are conducted. The focus of this review is to provide feedback to the procurement system and evaluate contractors’ results are not published. Second, ex-post audits are conducted by the OAG. Projects to be audited are selected on a risk basis. Entity audits by the OAG are published which incorporate project audits, but audits of individual projects are not published on a standalone basis. Third, ex-post audits are conducted by internal auditors of implementing agencies. These audits are intended to document lessons learned, reduce risk on similar future projects, and identify shortfalls in planning, budgeting, scoping, and execution stages. More systematic audits would assist to ascertain the value for money obtained from capital spending and also check the accuracy of the project justification.

I.14. Project management (Strength –Medium; Effectiveness –Low)

52. Project management can be improved in several important aspects—including a monitoring system, and identifying skilled staff. First, while the PM is the starting point for managing the project, it is followed up by the DPMS, a project management information system that covers the full range of project management activities, from NDP project identification, through procurement, to project closure and identification. Efforts are underway to re-establish the dysfunctional system, which the mission strongly supports. Second, implementing agencies do not systematically identify senior responsible officers for major projects prior to budget approval. Ideally, the project manager would be assigned at the time of preparing the PM and be involved in writing it. Project management is not only important during implementation— experienced project managers are key from inception to the ex-post completion stage—to ensure design and specifications are adequately addressed. Third, implementing agency project managers often leave government service as they gain experience, which reduces the availability of senior staff. This capacity constraint contributes to cost and time overruns, inefficient spending, and poor quality projects. The NSO and GICO have been created recently to provide project management support, which constitute an implicit acknowledgement that project management capacities need improvement. However, the working and reporting relations between these organizations, the MFED, and implementing agencies have yet to be fully clarified.

53. Project adjustments are guided by PPAD Regulation 95. This regulation provides standardized rules and procedures for project adjustments. However, it does not include the alternative, under certain circumstances, of a fundamental review and reappraisal of a project’s rationale, costs and expected output. Articles 6 and 7 of the regulation stipulate that an individual amendment should not exceed 15% of the original contract price, and that no contract should be amended by more than 25% of the original contract price. These limits appear not to be strictly adhered to—particularly for works related projects. More attention is required during the scoping of projects to eliminate the need for significant project adjustments.

54. Ex-post reviews focusing on project cost, deliverables and outputs are sometimes conducted. The only ex-post review and evaluation of projects that are completed is the End of Activity Report required by the PPADB—which is mainly used for disciplining underperforming contractors. But this report does not cover the full range of project management issues that ideally should be covered in an ex-post review. Such reviews are an important management tool aimed at assessing whether the project outputs and outcomes were successfully and efficiently achieved. The review should typically be conducted in two stages. The first stage focuses on whether the project has complied with the rules and procedures laid down in legislation; the second stage focuses on the quality of the services and facilities delivered. Ex-post reviews provide lessons learned and highlight project and management deficiencies to be avoided in planning, budgeting, and executing future projects. Box 1 provides an example from Chile’s experience.

Chile: Ex-Post Reviews

The ex-post evaluation procedures of the National Investment System (SNI) comprises two stages: evaluation of implementation and in-depth ex-post evaluation on the outcomes.

First stage:

A representative sample of investment initiatives (including pre-investment studies, projects, and investment programs) completed during the past two years are analyzed. The sample consists of 8–10% of all projects. Compliance with or divergence of project implementation from the ex-ante project specifications is analyzed concerning: total cost, outputs, and procedures and schedules.

The evaluation covers the role of all stakeholders, owners and sponsors of investment projects, the Ministry of Social Development (MDS), related agencies at sub-national levels, and sectoral ministries. The evaluation is then sent to the National Congress.

Second stage:

In-depth ex-post evaluations of specific projects are undertaken after the project has been operating for at least five years. It determines whether the anticipated benefits have been realized, and if not, an investigation of how and why the projections failed is undertaken. In these cases, all internal and external aspects are analyzed relating to the operations of the project.

These ex-post evaluations were initially applied to projects financed by the Regional Development Fund (FNDR) and executed by sub-national governments, and have gradually expanded to projects financed by sectoral ministries. These reports are published on the MDS website: http://sni.ministeriodesarrollosocial.gob.cl/.

Role of the Ministry of Finance:

Ministry of Finance is in charge of developing the framework for evaluating and monitoring public investment. MDS, the successor of the Ministry of Planning, is responsible for conducting the evaluations of public investment. External experts as well as staff are used for conducting the evaluations.

Source: Ministry of Social Development, Chile; IMF staff.

I.15. Assets accounting (Strength –Low; Effectiveness –Medium)

55. A central asset register is being rolled out in anticipation of introducing accrual accounting in 2019, but is not yet in widespread use. The Accountant General has developed a new information system, the IFAR, which was piloted in 2013 and is being introduced to line ministries in phases. IFAR is a module in the Oracle e-Business Suite, and is managed by a dedicated four-person unit in the Systems Support unit of the MFED. IFAR is intended to cover budgetary central government and to provide the basis for reporting non-financial assets and depreciation when accrual accounting standards are introduced in FY 2019–20. Once fully operational, IFAR will record asset purchases through a link with the procurement system, and ministries will subsequently identify initial location, changes in location, and retirement or disposal of an asset. IFAR is not designed to record information on the condition of assets, which will continue to be the responsibility of line ministries using separate systems without central guidelines or oversight. A policy framework for managing public assets has been drafted.39 Because IFAR is not yet being used by all ministries, the financial statements of the government do not yet report non-financial assets or depreciation.

56. Progress in establishing a centralized asset register and updating its accounting policies is clear but there is some work yet to be done. The MFED is committed to a system for monitoring public assets, demonstrated most clearly by using its own funds to develop IFAR and establishing the dedicated unit to implement the system. IFAR is working as intended and has the capability to handle the number of users and amount of data planned for it. Portions of five ministries are using the system now, and roll-out to an additional four ministries is ongoing. A number of policies and procedures need to be refined, including verification that all existing assets have been registered, both initially and periodically; how to use the asset information for non-accounting purposes, such as budgeting for maintenance and replacement, and determining the appropriate level of assets for the type of service being delivered. Accounting policies regarding the depreciation method to be used are yet to be determined. For the time being, there are no plans for periodic re-valuation of assets.

IV. Reform Priorities and Recommendations

A. All Projects

Issue 1: Expand the MTFF to include development projects

While the NDP includes economic and fiscal forecasts, an MTFF has been introduced to provide a medium-term perspective through which such forecasts are updated. The MTFF forecasts form a more reliable basis for estimating funding for development projects in the medium term, and determining the fiscal space for new projects after ongoing projects are funded. This enables staff to appraise and select a smaller set of projects in a timely manner well before the annual budget process begins, and to better plan for ongoing as well as new projects requiring multiyear funding.

Recommendations:

  • Select projects for inclusion in the third year of the development budget portion of the medium-term fiscal framework (MTFF). Selection should trigger preparation of the project memoranda, thus providing two years to scope and cost the project, among other tasks.

  • Publish indicative multiyear estimates in the development budget document for ongoing projects and those selected for inclusion in the forward years of the MTFF. This information should be drawn from project memoranda.

Issue 2: Improve coordination between capital and recurrent expenditures

Capital projects create the need for recurrent expenditures for the operation and maintenance of the new facility. Thus, approval of a capital project should occur only after it is estimated by the Budget Administration, the MFED unit responsible for preparing the recurrent budget, that sufficient recurrent funds are likely to be available to obtain the full set of benefits from the investment. While some progress has been made to coordinate the preparation of the development and recurrent budgets, there is scope for additional improvements.

Recommendation:

  • Extend the economic classification to the capital component of the development budget. This will enable measurement of capital versus recurrent expenditures in the development budget, and thus support coordination of expenditures, GFS reporting, and monitoring of changes in public capital stock.

B. Large and Mega Projects

Issue 3: Establish thresholds to define large and mega projects

Large and mega projects require sophisticated appraisal, and selection should not be made until the projects have been properly appraised. Consequently, projects classified as large and mega should be subject to different procedures and higher standards to ensure that the spending is efficient and fiscal risk is minimized.

Recommendation:

  • Define large and mega projects in terms of costs. For example, any project costing more than P 1 billion might be considered a mega project. Build appraisal, selection, and project management procedures based on this classification.

Issue 4: Clarify roles and responsibilities of all public entities involved in large and mega projects

Several organizations are currently, or planning to, participate in the planning, monitoring, management, and coordination of large and mega projects. These include the NSO, GICA, implementing agencies, and MFED. Overlapping roles and responsibilities have the potential to create confusion and diffuse accountability.

Recommendations:

  • Strengthen the role of the MFED in evaluating budget affordability for large projects regardless of how they are procured (e.g. traditional procurement or PPPs) or which public entity procures them (e.g. central, government sub-national governments, or parastatals). Specifically, this means strengthening:

    • ⚬ the PPP unit, with regard to oversight over establishment of PPPs; and

    • ⚬ a group of existing staff assigned to dedicate at least part of their time to reviewing funding request for large and mega projects, and thus to build up specialized knowledge and experience.

Issue 5: Improve costing, project preparation, and project management

Adequate project appraisal is essential to ensure that project scope, costs, and implementation schedules are realistic. Sophisticated appraisals are themselves expensive and time consuming, and should be conducted only for large or mega projects. Project selection should occur only after project appraisal is completed.

Recommendations:

  • Appraisal should follow a three-phase process covering pre-feasibility, feasibility, and independent review.

  • Pre-feasibility studies for a specific large or mega project should be identified as a project in the development budget. This ensures that there are sufficient funds to conduct a proper study and ensures at least one year of study before the project is selected for implementation funding.

  • Costing should be verified through the use of international unitary costs by type of project. For example, the Africa Property and Construction Cost Guide, published by AECOM.

  • Improve project management by establishing:

    • ⚬ A clear definition of the responsibilities of project managers and supervising officers in line ministries;

    • ⚬ Templates for monthly reporting of project costs and physical progress for submission to MFED, NSO, GICO, and Cabinet to avoid duplicative reporting obligations;

    • ⚬ A risk-based approach for central oversight of major projects;

    • ⚬ Criteria that define when a fundamental review and re-appraisal of ongoing projects are required and how it should be implemented. For example, if costs have increased by more than 15 percent than expected at a specific implementation milestone, the review process is triggered; and

    • ⚬ A certificate of project completion which provides an explanation of delays, cost overruns and any adjustments to the project scope.

  • Restoration of the Development Project Management System (DPMS), which is scheduled to be operational once again in mid-2017.

  • Involve project managers from project inception, not only at the implementation stage. Project managers should guide projects through appraisal, selection, procurement, and implementation phases.

  • Consultants should be appointed on a medium-term basis to supplement implementing agency staff.

C. Public-Private Partnerships

Issue 6: Strengthen the MFED oversight role

Challenges in fully implementing the 2009 PPP policy might expose the government to excessive risks arising from PPPs as the current legal framework is unclear, the MFED staff are not involved in assessing procurement of all PPPs, and the fiscal obligations to the government of Botswana are not comprehensively quantified.

Recommendations:

  • Ensure that all PPP projects are assessed and monitored by the PPP unit at the MFED, regardless of the public entity procuring them and whether or not they require any type of government support.

  • Require that the MFED’s PPP unit compiles and reports in an annex accompanying the budget documents data on existing PPP contracts, including:

    • ⚬ Investment by PPP companies (i.e., construction costs of PPP projects regardless of whether they are undertaken by the central government, sub-nationals, or SOEs and parastatals);

    • ⚬ Long-term cash-flow forecasts of gross costs and revenues of PPP companies and their associated present value;

    • ⚬ Estimates of government’s future payments to PPP companies and their associated present value; and

    • ⚬ Risk assessment of PPP contracts (e.g., explicit guarantees, traffic risks, force majeure, etc.), and risk mitigation measures undertaken by government.

  • Reinforce the PPP unit to ensure that is adequately staffed and is fully operational to assess, monitor, and report fiscal implications of PPP contracts. In particular, the use of a dedicated lawyer is warranted, given the need to clarify and/or develop the legal and regulatory framework underpinning PPPs. If new positions cannot be created, commissioning from other public entities (such as the Office of the Attorney General) could be considered.

Issue 7: Clarify the legal and regulatory framework for PPPs

In the absence of a dedicated PPP legal framework, existing challenges in the implementation of the PPP policy give rise to regulatory uncertainty, increasing fiscal costs for the government.

Recommendations:

  • Ensure that the proposed amendments to the PPAD Act are reviewed by the MFED’s PPP unit taking into account all relevant aspects of procuring PPPs in line with the PPP policy.

  • Ensure that the MFED’s PPP unit is in charge of and has adequate staff to develop the PPP enabling regulations and process guidelines once the amendments to the PPAD Act are approved.

Appendix I. Project Management Tool for Monitoring Project Execution

Planning and Costing:

Line Ministry:

Authority:

Date:

Compiled by:

A01ufig3
Note: in the database/template—the three stages would run sequentially—the division is for presentation only.

Procurement and Implementation:

Line Ministry:

Authority:

Date:

Compiled by:

Cash Flow and Completion Summary

Line Ministry:

Authority:

Date:

Compiled by:

Appendix II. Botswana: Gatekeeping Role of the Ministry of Finance and Economic Development

Appendix III. Botswana: Public Investment Management Capacity Building Plan

article image
article image
article image
KeyDP = Development ProgramsBA = Budget AdministrationLM = Line MinistriesGABS = IFMS technical support unitNSO = National Strategy OfficePPP = PPP Unit in MFEDPPADB = Public Procurement and Asset Disposal Board
1

See staff estimate: Chapter II - Efficiency and Impact of Public Investment Management.

2

Mineral revenues account for 35 percent of government revenues.

3

Since 2006, the government introduced formal debt limits on domestic and external debt equivalent to 20 percent of GDP each.

4

Peer countries refers to both emerging economies (EMEs) and comparator countries, the latter including Lesotho, Mauritius, Namibia, Swaziland, Seychelles, Cape Verde, and South Africa. The methodology for estimating public capital stock is detailed in the IMF Board Paper “Making Public Investment More Efficient,” June 2015.

5

It should be noted, that estimated capital stock for South Africa is not strictly comparable with regional peers, due to partial data at the municipality level and the exclusion of the SOE sector.

6

This is similar to the case of South Africa and Mauritius where a large share of total public infrastructure procured by SOEs (42 percent in south Africa and about 30 percent in the case of Mauritius)

7

2009 Law, section 2.1.

8

There is no centralized information of the value of existing projects. The estimates are based on data available at the World Bank’s PIIAF database. According to the later, only one power purchase agreement is currently in operation. Anecdotal evidence suggest that other PPP exist in other sector, even though relatively small in magnitude (e.g., health clinics, buildings).

9

Botswana Systematic Country Diagnostic, World Bank, March 2015; Botswana’s OECD Investment Policy Reviews, OECD 2014.

10

The methodology used in the calculation of the efficiency frontier and efficiency gap is detailed in the 2015 IMF’s Board Paper “Making Public Investment More Efficient.” Botswana’s public capital spending data corresponds to the development budget figures, which include some recurrent spending components, potentially overestimating public investment. On the other hand, Botswana efficiency gap would be even higher to the estimate presented above, if the capital spending by SOEs were to be considered as an additional input in the public investment process.

11

Part IV of the 2005 stocks, bonds and treasury bills act states only the limits.

12

NDP 9 operated from 2003–09 and NDP 10 from 2009–16.

13

IMF Article IV Report, 2016 (page 46) and Budget in Brief 2016/17 (Table 3).

14

NDP 11, Chapter 5.

15

Estimation made by analysis done by the department for development programs.

16

Botswana has 16 districts and 36 sub districts.

17

The Auditor General’s report for 2014/15 shows that Lobatse Council had not repaid the balances on all of its nine loans by March 31, 2015.

18

The budget book is published in February–two months before the start of the fiscal year

19

Literal 6.3 of the “Public-Private Partnership Policy and Implementation Framework,” MFED, 2009.

20

Regulatory uncertainty typically increases financial costs for the government and/or deters private sector participation in PPPs.

21

In Botswana, parastatals are mostly monopolies covering the power, water, transport, and communication sectors, which are typically the key drivers of a PPP portfolio.

22

Both SADC headquarters and the PPA signed by BPC in 2016 required explicit government guarantees.

23

At present, the PPP unit has only 2 ½ staff, without no dedicated lawyer assisting them. International experience suggest that legal support is critical particularly at early stages of the PPP project cycle.

24

According to the WB’s database (PPIAF) APR Energy operates a 70 MW diesel emergency power plant at Matshelagabeli, with a capital stock of 0.5 percent of GDP at the end of 2015.

25

BPC signed in September 2016 a power purchase agreement (PPA) of USD800 million (P8.7 billion) for expansion of the Morupule B Power Station to a consortium from Asia.

26

These include: (i) starting a business (down 10 ranks); (ii) getting credit (down 5 ranks); (iii) protecting minority investors (down 3 ranks); (iv) paying taxes (down 3 ranks), and (iv) enforcing contracts (down 2 ranks).

27

Publishing multi-year forecasts is relatively new in Botswana, with the first MTFF produced for the FY14-15 budget. Underlying ministerial projections were published in the Budget Call Circular for the first time to support the FY16-17 budget process. This meant that an assessment of multi-year forecast errors over time at the global and ministry level was not possible.

28

See institution 2.

29

See institution 10.

30

See institution 9.

31

See institution 11.

32

Communicated verbally by senior staff of the Development Programs section.

33

Part II, Financial Paper No. 2 of 2016/2017 is a good example.

34

See I.11, Protection of Investments.

35

Warrants serve as annual authorization to spend registered in the GABS.

36

Because of the large number of households without bank accounts, some payments must be made by cash. To support cash payments, each of the 32 Revenue Offices (responsible for both revenue and expenditures) has at least two commercial bank accounts, which are supplied from the Remittance Account. Each Revenue Office has access to GABS in which to record cash transactions.

37

The Bank of Botswana is the central bank.

38

For example, see Statement 3, Report of the Auditor General on the Accounts of the Botswana Government for the Financial Year ended March 31, 2015.

39

Dated June 2015, with support from the Regional Technical Assistance Center (RTAC) in Southern Africa.

Botswana: Technical Assistance Report-Public Investment Management Assessment
Author: International Monetary Fund. Fiscal Affairs Dept.
  • View in gallery

    Botswana: Public Investment and GDP Growth

    (In percent of GDP and percent change)

  • View in gallery

    Botswana: Public Debt and Deficit

    (In percent of GDP)

  • View in gallery

    Current vs. Capital Public Spending, Comparison to Peers, 2015

    (In percent of GDP)

  • View in gallery

    Botswana: Composition of Public Spending, 1990–2015

    (In percent of total)

  • View in gallery

    Public Investment Trends: Comparison with Peers

    (2005 PPP$-adjusted, percent of GDP)

  • View in gallery

    Public Capital Stock, Comparison with Peers, 2015

    (2005 PPP$-adjusted, percent of GDP)

  • View in gallery

    Botswana: Public Capital Stock per Capita

    (2005 PPP$-adjusted, percent of GDP)

  • View in gallery

    Public Capital Stock per Capita: Comparison with Peers, 2015

    (2005 PPP$-adjusted, percent of GDP)

  • View in gallery

    Comparison of Volatility of Public Investment1

    (Average 2010-13)

  • View in gallery

    Botswana: Execution of Capital Expenditure

    (Percentage executed, relative to revised budget)

  • View in gallery

    Botswana: General Government Investment by Function, Average 2010–15

    (In percent of total general government investment)

  • View in gallery

    Emerging Market Economies: General Government Investment by Function, Average 2010–15

    (In percent of total general government investment)

  • View in gallery

    Botswana: Composition of Public Investment by Level of Government

    (In percent of total public investment)

  • View in gallery

    Botswana: General Government Investment by Sources of Funding, 1990–2014

    (In percent of total general government investment)

  • View in gallery

    Public-Private Partnerships Capital Stock, Comparison with Peers, 2015

    (2005 PPP$-adjusted, percent of GDP)

  • View in gallery

    Perception of Infrastructure Quality, 2006–15

  • View in gallery

    Botswana: Infrastructure Quality, Decomposition, 2006–15

  • View in gallery

    Measures of Infrastructure Access and Service Delivery, 2015

  • View in gallery

    Power Generation Capacity, Regional Comparison, 2006

  • View in gallery

    Efficiency Frontier, Quality Indicator, 2008–15

  • View in gallery

    Efficiency Gap, Quality Indicator, 2008–14

  • View in gallery

    Strength of Public Investment Management Institutions

  • View in gallery

    Botswana: Total Spending to GDP

    (percent)

  • View in gallery

    Botswana: Capital Budget Shares

    (percent of total budget and outturn)

  • View in gallery

    Variances in Major Infrastructure Projects, 2009–10, 2011–12

    (percent)

  • View in gallery
  • View in gallery
  • View in gallery
  • View in gallery