Scaling up Public Investment, Growth and Debt Sustainability1
1. Burkina Faso, like many developing countries, has significant infrastructure gaps. In 2011, the World Bank estimated that the country would need to invest 11 percent of GDP in infrastructure over 10 years to close the gap with the continent’s leading performer (Briceño and Dominguez, 2011). Nevertheless, according to the 2016 African Infrastructure Development Index, the country still lags in the bottom 20 performers for the continent. While natural resource wealth from gold and other commodity exports has the potential to accelerate the country’s development, resource-rich developing countries that plan to increase public investment financed in part by external borrowing may carry substantial debt risk. For example, soaring oil prices in the 1970s allowed many oil-exporting countries to undertake ambitious investment projects, financed by oil revenues and external borrowing but the increase in interest rates and the collapse of oil prices in the 1980s contributed to a sequence of debt crises.
2. The central policy question is how best to make use of resource revenues to support the economy while maintaining debt sustainability. Many factors come into play in addressing this issue including: commodity prices; investment budget absorptive capacity, rate of return and efficiency of investment projects, private sector response, Dutch disease, availability of financing, government debt composition, efficiency of revenue mobilization and expenditure rules. While it is difficult to estimate quantitatively how these different elements knit together, the analysis here will make use of a model to examine the empirical evidence, provide a framework for policy advice and, above all, try to better understand the linkages between the underlying determinants. In doing so, it will focus on the macroeconomic effects of scaling up public investment in terms of the trade-offs associated with different financing options and will briefly address the impact of shocks to the resource sector.
B. The Country Context
3. Prior to the sociopolitical events of 2014-15, Burkina Faso had experienced sustained growth. A recent IMF African Department publication2 identifies Burkina Faso as one of the few non resource-rich low-income countries in Sub-Saharan Africa (SSA) that have been able to achieve consistently high growth rates over a 15-year period. The report identifies several determining factors underpinning this outturn: improved macroeconomic management, stronger institutions, increased aid, and higher investment in human and physical capital. Between 1990 and 2015, real GDP growth became steadily higher and less volatile (Figure 1) as the rolling five-year average real GDP growth rate rose steadily from 3.0 to 5.5 percent with a simultaneous decline in the rolling five-year standard deviation from 3.8 to 1.4 percent (Figure 1). Medium-term-oriented policies and important structural reforms have aimed at stabilizing growth, notably in agriculture, which is particularly vulnerable to weather shocks. Following the recent downturn, the trend growth rate is expected to resume and improve at 6-7 percent over the medium term as the sociopolitical climate fully stabilizes, investment increases and energy supplies improve. Although the share of gold mining in national output is low (mostly due to outdated national accounts that are in the process of being rebased), it nonetheless accounted for 58 percent of exports and 14 percent of fiscal revenues in 2014, down from a peak of 75 percent and 19 percent, respectively, in 2011-12, when international prices were nearly 30 percent higher.
Figure 1.Real GDP Growth, 1990-2015
Source: Countries Authorities; and IMF staff estimates.
4. Burkina Faso has a good track record of maintaining macroeconomic stability even in the face of severe shocks in recent years, but higher, more diversified, and more sustainable growth is essential in order to substantially reduce poverty and inequality. The authorities’ National Economic and Social Development Plan (PNDES) aims to accelerate growth through a major increase in public investment, equivalent to about 17 percent of GDP annually until 2019 compared to 12.5-13 percent of GDP in the baseline and resulting in a substantial increase in the projected debt-GDP ratio (Figure 2). Investment will focus on energy and transportation infrastructure, which are considered the main bottlenecks to growth. The analysis aims at assessing the potential growth impact of this ambitious scaling-up of investment and how the authorities can generate the necessary fiscal space to increase investment without jeopardizing macroeconomic stability. Estimates of the growth impact of public investment in the literature suggest a range of 0.2-0.5 percent onto the growth rate for each additional percentage point of public investment.
Figure 2.Total Public Debt, Public Investment and Fiscal Deficit, 2015-19
Sources IMF Staff Estimates
5. Investment budget execution has been a key concern for program management for many years but the issue has become critical in the past 2 years with execution rates falling to 67.3 and 75.5 percent in 2014 and 2015 respectively. The problem has been particularly acute as regards externally-financed investment. Over 2013-15, investment spending overall declined from 14 percent of GDP to just over half that level at 7.7 percent of GDP with both domestically financed and externally-financed investment declining sharply. Absorption capacity issues were certainly a factor in this underperformance as was the uncertainty regarding the political process in terms of its impact on externally-financed investment. The main issues identified as constraints to investment budget execution at the outset of the current ECF-supported program included redundant payment approval systems, a lack of automated processes and procurement delays. Many of the structural benchmarks defined for program implementation specifically targeted improvements in these areas with a limited degree of success. At any rate, these low rates of investment budget execution were significantly sub-optimal for a country with Burkina Faso’s development needs.
6. Remaining capacity constraints and bottlenecks are deep-seated and broad-ranging. A recent FAD mission highlighted weaknesses at all stages of the investment process: there is no rigorous process to select projects, resulting in poor quality or delays due to insufficient preparation; ongoing issues with budget classification result in only about 44 percent of total expenditure in the investment budget actually resulting in fixed capital formation; procurement processes are unduly long and complex; and execution monitoring and reporting remains inefficient. As a result, there has been increasing recourse to exceptional procedures such as “deblocage de fonds” and single sourcing for procurement which are difficult to monitor and not cost-efficient. The state’s audit authority (ASCE) has also identified domestic arrears amounting to about 1 percent of GDP, linked at least in part with lack of control over project execution or with invalid contracting or procurement procedures. The mission recommended a battery of specific measures to improve project selection and execution, such as better integration of the capital budget into the overall budgetary architecture and clearer assignment of responsibilities for projects critical to structural transformation. The authorities have also expressed interest in a Public Investment Management Assessment (PIMA) to be held as soon as possible to develop more comprehensive recommendations to further improve project selection and implementation in the long term with a view to bringing about a permanent improvement in performance in this critical area.
C. Impact of Public Investment Scaling-Up: Principles and Underlying Determinants
7. A key issue in determining the outcome of any scaling-up initiative is the efficiency of public investment. Each additional unit of investment spending does not necessarily result in an equivalent unit of productive public capital. In developing countries high productivity of infrastructure can coexist with low returns to public investment because of inefficiency in investing due to corruption and scarce technical capacity but investment efficiency can drop also because abrupt investment scaling-up that encounters absorptive capacity constraints (supply bottlenecks and lack of complementary infrastructure are typical examples). Insufficient maintenance can also shorten the life of existing capital. Accounting for the operations and maintenance expenditures of installed capital is crucial for assessing the growth effects and debt sustainability of public investment scaling-up. It can therefore be assumed that the depreciation rate increases proportionally to the extent to which effective investment fails to maintain existing capital.
8. Crowding in (or crowding out) is another important issue to consider as private investment responds both to interest rates and tax rates, as well as to the higher rates of return generated by more public capital. Higher public capital raises the marginal product of private capital but when the government uses domestic resources to invest, these resources are not available for private investment and consumption (crowding out). Moreover, by creating distortions, tax increases to finance the scaling-up efforts or reductions in transfers lower consumption and welfare. A further consideration is that recurrent infrastructure costs should be covered at least partially by user fees: failure to do so exacerbates debt sustainability risks. The final outcome depends on the interaction of these often conflicting influences. In the long run, crowding in will occur if the projects concerned are of good quality, and address bottlenecks to private sector development, such as essential road or energy infrastructure. However, during the transition, crowding out may dominate, especially early on and especially if there is not enough foreign financing available.
9. A conservative approach to scaling up would set moderate targets for public investment bearing in mind that public capital and non-resource output growth do not increase until the medium-term. A gradual approach would develop spending plans while anticipating some of the increase in non-resource output while an aggressive approach would anticipate future output growth with a big immediate increase in public investment spending. Due to absorptive capacity constraints, the aggressive approach may only deliver an accumulation of public capital and output comparable with the gradual approach. Under a conservative approach public debt will increase less as a share of GDP and even decline over the longer term, thus implying no future fiscal adjustment. Under a gradual approach public debt rises gradually to a peak in the medium term, then declines over the longer term to more sustainable levels as public capital builds up and non-resource output expands, with the result that any required fiscal adjustment is limited. An aggressive approach implies a much bigger build-up of public debt in the medium term, and would require a painful fiscal adjustment in order to service the accumulated debt.
10. If, over time, institutions, governance, and management practices are improved public investment becomes, on average, more efficient and absorptive capacity constraints less binding. If investment projects are better designed, selected, and implemented, the return of investment could increase: it will increase productivity with a more positive impact on capital stock accumulation, growth and incomes. Good governance and careful project selection and execution are key determinants of the impact public investment has on growth and thus optimizing the effectiveness of investment is crucial to maximize its positive effects and limit the build-up of debt needed to fund such investment.
11. Natural resource revenue has the potential to substantially increase government revenue, but this is inherently uncertain and volatile. A gradual public investment scaling-up anticipating some but not all future potential revenue would be appropriate given Burkina Faso’s infrastructure investment needs and the uncertainty regarding mining production/revenue. This would help translate mining revenues into non-resource growth without compromising fiscal stability. Due to absorptive capacity constraints, an aggressive approach is not likely to yield tangibly better growth outcomes and poses threats to debt sustainability.
D. Model Simulations (DIGNAR)
12. The analysis includes a model estimation of the macroeconomic implications of the PNDES scenario compared to staff’s baseline projections using the Debt, Investment, Growth and Natural Resources (DIGNAR) model. The model, developed by the IMF’s Research Department, combines the debt sustainability framework developed in Buffie et al. (2012) with the natural resource model in Berg et al. (2012). It is particularly well-suited for assessing the investment-growth nexus together with debt sustainability in resource-rich developing countries that, like Burkina Faso, intend to scale up public investment, financed by a combination of borrowing and resource revenues. The analytical framework includes a natural resource sector in addition to traded and non-traded goods sectors, and is calibrated to reflect important features of developing countries, including varying degrees of investment efficiency, limited absorptive capacity, Dutch disease, and a detailed fiscal specification reflecting the operation of fiscal buffers. Taking resource revenues and public investment policy as given, the framework can simulate the macroeconomic outcomes of scaling up public investment, accounting for the investment-growth linkages and the feedback effect on non-resource revenue. DIGNAR models were previously used to assess public expenditure scaling-up options in several countries, notably Chad, the Republic of Congo, Mozambique and Kazakhstan, as well as for a group of four fragile countries in West Africa (Deléchat et al., 2015, IMF, 2015, IMF, 2014, Melina and Xiong, 2013, and Minasyan and Yang, 2013).
13. The model analysis sets out to compare the impact of PNDES scaling up compared to a baseline projection which is essentially defined by the debt sustainability analysis for the sixth ECF program review (corresponding to the gradual approach outlined above). Under the scaling-up scenario (which represents the aggressive approach to scaling up described above), the profile of public investment is defined by the authorities’ projections under the 2017 budget proposal which would raise capital spending to 17 percent of GDP. The ratio of investment spending to GDP is assumed to decline gradually to 15 percent of GDP by 2026 and to remain at that level until the end of the simulation period in 2035. The trajectory for the investment budget in the baseline is considerably less ambitious with investment expected to reach 12 percent of GDP in 2017 and rise progressively to 14 percent of GDP in 2026 before declining gradually towards 10 percent of GDP by 2036 (Figure 3).
Figure 3.Public Investment (% of GDP)
14. A key feature of the assessment is the role of the investment efficiency parameter, which for analytical purposes is assumed in the PNDES scenario to drop to 35 percent, from the 50 percent efficiency rate assumed in the baseline. The rationale for this assumption is that an aggressive increase in public investment spending will immediately encounter the absorption capacity constraints holding back investment spending historically and will result in a decline in the quality of spending relative to the baseline. These constraints will take the form (inter alia) of the lack of complementary infrastructure, insufficient provision for maintenance and supply bottlenecks (Figure 4). GDP growth is consequently lower in the long run than in the baseline due to lower productivity in both the traded and nontraded sectors resulting from the relatively lower increments to the public capital stock (Figure 5).
Figure 4.Public Investment Efficiency (%) Figure 5.Non-resource GDP Growth (%)
15. The results of the simulation indicate that long-term debt sustainability would be compromised by the increase in spending and borrowing implied by the PNDES scenario. Figure 6 indicates that the debt/GDP ratio would increase beyond the limit defined by the WAEMU criteria by the end of the simulation period (the 70 percent limit would be breached sometime around 2030). By contrast, the baseline projects a debt/GDP ratio stabilizing at around 50 percent of GDP by the end of the simulation period. Welfare would also be lower under the scaling up scenario as private consumption levels would ultimately decline by 7 percent relative to baseline levels in response to decline in disposable incomes resulting from the higher levels of taxation needed to maintain fiscal sustainability.
Figure 6.Total Public Debt (% of GDP)
E. Sensitivity Analysis: Risks to the Baseline from Resource Revenue Volatility or Fiscal Policy
16. The analysis also includes 2 variants on the baseline scenario. In the first, the price of gold is assumed to drop by about 30 percent relative to the baseline while the second variant assumes that there will be no increase in taxation to finance the higher investment spending in the baseline. This is motivated by the fact that, in many countries, there may be political economy constraints that limit the feasibility of tax increases. In the gold price scenario, the loss in resource revenues (from 15 to 10 percent of total revenues) would result in an increase in the debt/GDP ratio by around 12 percentage points (Figures 7 and 8). There would also be a slight decline in GDP and in welfare under this variant due to increased indirect taxation to make up for revenue shortfalls.
Figure 7.Resource Revenue (% of tot. rev.) Figure 8.Total Public Debt (% of GDP)
17. As regards the variant where the tax rate is fixed, the principal effect is the increase in the debt-to-GDP ratio (Figure 9) by even more than in the scaling-up case. Public debt would reach almost 90 percent of GDP by 2030, well beyond the level defined by the WAEMU criterion. This underscores the importance of ensuring adequate measures to maintain fiscal sustainability in the context of scaling-up initiatives. In the standard model approach, consumption taxes are set to increase to close off the fiscal financing gap arising from scaled-up investment spending not covered by additional borrowing. When the tax rate cannot adjust and is fixed (Figure 10) at its initial level, then additional (non-concessional) borrowing is required. The increase in the debt-to-GDP ratio is a full 35 percentage points above the baseline value and 3 points above that of the scaling-up scenario assuming partial fiscal adjustment. Moreover, the increase in debt is accounted for by increases in both external non-concessional and domestic debt which emphasizes the importance of continued vigilance in the area of debt management.
Figure 9.Total Public Debt (% of GDP) Figure 10.Consumption Tax Rate (%)
18. The results point to some clear conclusions for best practice in scaling up public investment. “Big-push” investment efforts, while designed to accelerate growth, are likely to run up against significant absorption capacity constraints. These constraints will diminish the efficiency of investment spending and result in a lower rate of public capital accumulation and productivity increase than when a more measured approach is adopted. The empirical evidence from the experience of many countries (including a recent paper on Burkina Faso (Kabedi-Mbuyi et al., 2016) suggests that the results of aggressive scaling-up initiatives are mixed. Model simulations for the aggressive scaling-up scenario implied by the PNDES indicate that it is likely to be inconsistent with long-term fiscal and debt sustainability. In addition, there are non-negligible risks to even the moderate gradual scaling up scenario outlined in the baseline stemming, for example, from commodity price shocks. The central policy lesson in this regard is that good investment budget management is one of the keys to prospects of success in the implementation of the PNDES. Recent initiatives undertaken by the Burkinabè authorities will help in this respect, particularly the revised procurement code which will contribute to significantly improving investment budget execution. Building on the results of the Public Investment Management Assessment (PIMA) exercise to be launched in the coming months to strengthen the key institutions which shape the planning, allocation, and implementation of public investments will also be critical: IMF research suggests that strengthening these institutions could close up to two-thirds of the public investment efficiency gap (IMF, 2015).
AraujoJuliannaGraceLiMarcosPoplawski-RibeiroLuis-FelipeZanna2016 “Current Account Norms for Resource Rich and Capital Scarce Countries” Journal of Development Economics Vol. 120 pp. 144–156.
BergAndrewRafaelPortilloShu-Chun S.YangLuis-FelipeZanna2012 “Public Investment in Resource-Abundant Developing Countries” IMF Working Paper 12/274 (Washington D.C.: International Monetary Fund).
Briceño-GarmendiaCeciliaCarolinaDomínguez-Torres2011 “Burkina Faso’s Infrastructure: A Continental Perspective” Africa Infrastructure Country Diagnostic (AICD) country report (Washington D.C.: The World Bank Group).
BuffieEdward E.AndrewBergCatherinePattilloRafaelPortilloLuis-FelipeZanna2012 “Public Investment, Growth, and Debt Sustainability: Putting Together the Pieces” IMF Working Paper 12/144 (Washington, D.C.: International Monetary Fund).
DelechatCorinneWillClarkPranavGuptaMalanguKabedi-MbuyiMesminKoulet-VickotCarlaMacarioToomasOravManuelRosalesReneTapsobaDmitryZhdankinSusanYang2015 “Harnessing Resource Wealth for Inclusive Growth in Fragile States” IMF Working Paper 15/25 (Washington D.C.: International Monetary Fund).
GuptaPranavGraceLiJiangyanYu2015 “From Natural Resource Boom to Sustainable Economic Growth: Lessons for Mongolia” IMF Working Paper 15/90 (Washington D.C.: International Monetary Fund).
HakuraDalia S.AdrianAlterMatteoGhilardiRodolfoMainoCameronMcLoughlinMaximilienQueyranne2015 “Sustaining more Inclusive Growth in the Republic of Congo” African Departmental Paper 15/02 (Washington, D.C.: International Monetary Fund).
International Monetary Fund (IMF)2014 “Chad—2013 Article IV Consultation and Assessment of Performance under the Staff-Monitored Program—Staff Report” IMF Country Report 14/100. (Washington, D.C.: International Monetary Fund).
International Monetary Fund (IMF)2015 “Making Public Investment More Efficient” IMF Staff Report (Washington, D.C.: International Monetary Fund).
Kabedi-MbuyiMalanguMame AstouDioufConstant LonkengNgouana2016 “Walking a Fine Line: Public Investment Scaling-Up and Debt Sustainability in Burkina Faso” African Departmental Paper 16/03 (Washington, D.C.: International Monetary Fund).
MelinaGiovanniYiXiong2013 “Natural Gas, Public Investment and Debt Sustainability in Mozambique” IMF Working Paper 13/261 (Washington, D.C.: International Monetary Fund).
MelinaGiovanniShu-Chun S.YangLuis-FelipeZanna2014 “Debt Sustainability, Public Investment, and Natural Resources in Developing Countries: the DIGNAR Model” IMF Working Paper 14/50 (Washington D.C.: International Monetary Fund).
This paper was jointly prepared by Liam O’Sullivan (African Department) and Jun Ge (Research Department).
Regional Economic Outlook, Sub-Saharan Africa, October 2013, Keeping the Pace, IMF.