- Alexei Kireyev, and Gaston Mpatswe
- Published Date:
- October 2013
2012, “Building Inclusive Growth in the Philippines.” IMF Country Report No. 12/50 (Washington: International Monetary Fund).
2011, “Inequality and Unsustainable Growth,” IMF Staff Discussion Note, April 8, SDN/11/08 (Washington: International Monetary Fund)..
2005, “Access to Financial Services: A Review of the Issues and Public Policy Objectives,” Policy Research Working Paper Series 3589 (Washington: World Bank).
Enquête Suivi de la Pauvreté au Sénégal(ESPS-II 2010–2011). Dakar, Agence Nationale de la Statistique et de la Demographic.
International Monetary Fund, 2011, “How Inclusive Has Africa’s Recent High-Growth Episode Been?” Regional Economic Outlook: Sub-Saharan Africa (Washington, October).
2003, “Measuring Pro-Poor Growth,” Economics Letters, Vol. 78, pp. 93–99.
World Bank, 2005, Pro-Poor Growth in the 1990s (Washington: World Bank).
Quasi-fiscal costs related to the power sector in Senegal are significant. Electricity tariff subsidies are well targeted and generally do not benefit the poor. With high international oil prices and under current policies, the subsidies are expected to remain high until less costly technologies are introduced.
Power Sector: Current Challenges
The Senegalese power sector faces tremendous challenges. The sector relies heavily on oil imports for power generation; only about 10 percent of available electricity supply is purchased from regional hydropower. Primarily as a result of increasing oil prices, tariffs set at levels well below full cost recovery, lack of proper maintenance, and delays in making investments in capacity to match an increasing power demand, the supply of electricity has become insufficient and unreliable, and the financial situation of the national utility company (SENELEC) has weakened significantly over time. SENELEC has had a large structural operating deficit over recent years; despite significant budgetary transfers, its net income has been negative since 2005 (except in 2009). This situation culminated in a major power crisis in 2010–11, with regular power outages affecting the well-being of the population and economic activity. An emergency plan (“Takkal”) was put in place in 2011 to address the situation.
Tariff Gap and Quasi-Fiscal Costs
The setting of the tariff follows a well-defined regulatory process. SENELEC’s customers are billed based on the amount of kWh of electricity they consume. Prices are differentiated by voltage, with users of higher voltages billed at higher prices. A price differential also applies within the same voltage band, with higher consumption priced at a higher tariff. Under the current regulatory framework, changes in tariff levels and/or schedule must be authorized by the Power Sector Regulatory Commission (CRSE in French) before taking effect. The regulatory framework includes mainly four elements: (i) a formula to determine, under baseline tariff and economic conditions, the maximum revenue the utility is allowed to collect to cover its operating and capital costs and make reasonable profits; (ii) conditions that could trigger a tariff adjustment; (iii) a compensation mechanism from the budget when tariff adjustments are not allowed by the government; and (iv) periods for revising the formula and the tariff conditions and for paying tariff compensation if any.
The tariff, in practice, has not been set to allow cost recovery; this has required substantial compensation from the budget. Some tariff adjustments were authorized between 2007 and 2009, though not high enough to reduce the large spread between the average price and the average cost per billed kWh.6 Because of a flawed compensation mechanism that relied on overly optimistic assumptions about the future (e.g., when cost-effective technologies would become available), the tariff compensation received by SENELEC was insufficient and led to losses. In July 2011, the tariff formula was revised and became more realistic. This led to a substantial upward revision of the 2011 tariff compensation (from an initial estimate of CFAF 45 billion to about CFAF 96 billion). The tariff in 2011 was on average about 40 percent below the level suggested by the CRSE formula. Despite this large tariff gap, electricity remains very costly in international comparison.
|Total power supply (GWh)||2,171||2,192||2,306||2,400||2,489||2,618||2,559|
|Power supply (percent)|
|Of which: Regional hydropower||12||11||8||10||10||10||10|
|Power sales and losses (percent of supply)|
|Historic sale prices and costs (CFAF/kWh)|
|Average tariff (Comp. incl.)2||94||109||122||144||133||132||169|
|Average operating cost3||97||129||112||126||110||130||147|
|Fuel costs (SENELEC only)||49||77||67||105||71||111||163|
|SENELEC results (CFAF billion)|
|Operating subsidies received||26||33||37||60||40||28||103|
|Operating results (comp & subsidies excl.)||−33||−56||−40||−61||−27||−80||−100|
|Operating results (with subsidies)||−7||−23||−3||−1||13||−52||4|
The new tariff compensation system, although a major improvement, may not provide a full picture of the losses in the electricity sector. This reflects a number of issues, such as assumptions concerning the cost structure, which are fixed for three years, or the existence of technical and nontechnical losses not entirely compensated for by the formula. Although the collection of billed electricity units has reportedly been close to 100 percent, the losses between production and billing have been about 20 percent. This is way above usual technical losses (e.g., related to transportation and transformation), which in efficient systems are contained below 10 percent. Total losses, measured as the difference between the cost of units actually supplied to the power grid and the amounts recovered, amounted in 2011 to an estimated CFAF 141 billion (2.1 percent of GDP). The tariff gap, as measured by the formula, made up about two-thirds of this amount. The rest corresponds to “unaccounted” losses, basically reflecting SENELEC’s inefficiencies. What matters from a public finance perspective is obviously total losses, given that SENELEC is government-owned. The losses not compensated through the formula eventually have to be borne by the taxpayer one way or another.
|(In CFAF billion)|
|Total quasi-fiscal costs||85||133||83||95||67||106||141|
|Underpricing (estimated tariff gaps)||59||100||46||53||28||61||97|
|SENELEC liabilities (stock)||149||248||252||312||336||402||385|
|Liabilities to suppliers||62||138||131||171||187||239||257|
|(In percent of GDP)|
|Total quasi-fiscal costs||1.8||2.7||1.5||1.6||1.1||1.7||2.1|
|Underpricing (estimated tariff gaps)||1.3||2.0||0.9||0.9||0.5||1.0||1.4|
|SENELEC liabilities (stock)||3.5||5.4||5.1||5.8||5.6||6.7||6.0|
|Liabilities to suppliers||1.3||2.8||2.4||2.9||3.1||3.7||3.8|
Figure A.1.Quasi-fiscal Costs as Percent of GDP, 2005–11
Figure A.2.Quasi-fiscal Costs as Percent of SENELEC Sale Receipts, 2005–11
If oil prices remain high, electricity subsidies will remain large in 2012 and decline only gradually. Absent any changes in the tariff and efficiency gains at SENELEC, the other main factors affecting electricity subsidies are the level of international oil prices, electricity consumption, and technology. With oil prices expected to remain high in the medium term and power consumption likely to increase, only the introduction of more efficient production units (such as natural gas, hydropower, or coal-fired plants, as envisaged by the authorities) could lead to a significant reduction of electricity subsidies. It therefore makes it critical to finalize the energy sector strategy and to implement expeditiously these investments in new production units, as it will take a few years before they come onstream and contribute to much lower electricity costs.
Distributional Effects of Electricity Subsidies
Beyond being a drain on the budget, electricity subsidies are poorly targeted. Subsidies are typically justified by the need to make power services affordable to low-income and vulnerable households. However, empirical evidence on Senegal (and elsewhere) from the mid-2000s suggests the following about tariff subsidies7:
They are not necessarily benefiting the poor, primarily because most of them are not connected to the power grid. This access factor reflects not only the unavailability of electricity in the areas where poor households tend to live but also poor households’ choice not to connect to the power grid even when they can. Based on data from 2005–06, a World Bank report shows that all indicators on electricity consumption increase with household income. Households with lower income are less connected and have less access, and even when they have access they are less likely to connect to the power grid because of connection fees. Connection and access rates in rural areas were estimated at 15 percent and 39 percent, respectively.
Even if they were benefiting the poor in absolute terms, their distributional effects have been regressive because electricity consumption is unevenly distributed across regions and income groups. Using a simple measure of targeting, the World Bank’s analysis suggests that only 12.5 percent of the subsidies implicit to the tariff structure benefited the poor. Similarly, Fund analysis concluded that the increase in tariff (22 percent) during 2005–07 had a very limited effect on the welfare of poor households in rural areas.8
Tariff subsidies distort prices, and as such constitute an inefficient way of allocating resources within the economy. A lower tariff gives an incentive to electricity customers (households and businesses) to consume more electricity than they would otherwise do and thereby exacerbates the burden on public finances.
Tariff subsidies divert important resources needed to finance pro-poor and priority spending. Although most of the benefits go to individuals or businesses that do not need subsidization, their costs are very large and crowd out government spending in other areas, including the electricity sector itself. For instance, the cumulative transfers received by SENELEC over the last five years amounted to about CFAF 400 billion, which is nearly equivalent to resources needed for the construction of coal-fired plants that would increase production capacity by more than 50 percent. Similarly, annual transfers to SENELEC were comparable to, or higher than, the resources allocated for capital spending in the health or education sector (Figure A.3).
Figure A.3.Transfers to SENELEC, Capital Spending in Health and Education Sectors, 2005–11
Higher debt-financed public investment may have substantial impact on growth and debt sustainability. Using an open-economy dynamic general equilibrium model, several scenarios are constructed by altering assumptions about the composition of financing. The results of the model suggest that increased public investment would have a positive impact on growth, but could lead to debt sustainability problems if financed mainly by nonconcessional borrowing10
An ambitious investment policy has been central to Senegal’s development strategy. Public investment as a share of GDP has doubled since 2000, rising from 5.5 percent to nearly 11 percent in 2011. With the support of donors, the authorities have focused investment spending on expanding and modernizing transport infrastructure to facilitate access to markets. Major infrastructure projects include the Blaise Diagne International Airport, the Dakar-Diamniadio toll road, other road projects, and the modernization of the port of Dakar. These projects are intended to spur private sector development, diversify economic activity away from Dakar, and support poverty reduction.
This section explores the impact of a higher public investment path in the next few years. The public investment ratio is expected to remain relatively high in the program’s macroeconomic framework for 2012–17. This note assumes a further scaling up of public investment to 15 percent of GDP in 2013 and a subsequent gradual decrease back to levels assumed in the program. The spike in the next few years, although not the program’s central scenario, is based on the pipeline of current and future investment projects maintained by the authorities; it is therefore plausible, and implicitly corresponds to an early implementation of major investments in the energy and transport sectors.
The model’s parameters were calibrated to match data for the Senegalese economy. Table A.3 summarizes the calibration of the main parameters. In those instances where Senegal-specific estimates were not available, the parameters were set to fit a “generic” low-income country, as described by Buffie et al. (2011). Key parameters affecting debt dynamics are absorptive capacity, the return on infrastructure investment, investment efficiency, and user fees.
|Absorptive capacity parameter||11||IMF staff|
|Initial return on infrastructure investment||0.313||Dessus and Herrera (1996)|
|Efficiency of public investment||0.70||Briceño-Garmendia et al. (2011)|
|User fees for infrastructure services (% of recurrent costs)||0.83||IMF staff|
|Value added in nontraded sector||0.450||Country authorities|
|Capital’s share in value added in the traded sector||0.400||Country authorities|
|Capital’s share in value added in the nontraded sector||0.550||Country authorities|
|Cost share of nontraded inputs in the production of capital||0.500||Country authorities|
|Depreciation rate||0.10||Country authorities|
|Trend per capita growth rate||0.017||Country authorities|
|Initial real interest rate on domestic debt||0.035||IMF staff|
|Initial real interest rate on private external debt||0.035||IMF staff|
|Real interest rate on concessional loans||0.000||IMF staff|
|Initial real interest rate on public commercial loans||0.065||IMF staff|
|Initial public domestic debt-to-GDP ratio||0.10||IMF staff|
|Initial public external concessional debt-to-GDP ratio||0.192||IMF staff|
|Initial public external commercial debt-to-GDP ratio||0.103||IMF staff|
|Initial private external debt-to-GDP ratio||0.200||IMF staff|
|Grants-to-GDP ratio||0.026||Country authorities|
|Remittances-to-GDP ratio||0.094||Country authorities|
|Initial ratio of infrastructure investment to GDP||0.105||IMF staff|
|Initial consumption VAT||0.18||Country authorities and IMF staff|
|Labor ratio of nonsavers to savers||4.3||Country authorities and IMF staff|
Absorptive capacity constraints are significant in Senegal. When absorptive capacity is limited, high investment rates lead to large cost overruns, usually due to coordination problems or supply bottlenecks during the implementation phase of public investment projects. The assumption for Senegal is that, on average, public investment projects incur costs overruns equal to 74 percent of the size of the initial investment. This estimate is based on data provided by the authorities on initial estimates and final costs of large infrastructure projects in Senegal. To arrive at costs overruns equal to 74 percent of initial investment, taking into account the projected path of investment, the parameter that controls absorptive capacity constraints was set to 11.
The gross return on infrastructure investment is assumed to be 31.3 percent. Estimates of the return on infrastructure investment vary significantly Foster and Briceño-Garmendia (2010) found returns ranging from 17 to 24 percent on electricity, water and sanitation, irrigation, and road projects. Estimates by Dalgaard and Hansen (2005) cluster between 15 percent and 30 percent. Buffie et al. (2011) assume a 30 percent return on infrastructure investment for a generic low-income country. The assumption of a 31.3 percent return in Senegal is taken from work by Dessus and Herrera (1996) on the impact of public investment on growth in developing countries.
Not all investment spending is assumed to contribute to capital accumulation. The investment efficiency parameter captures the idea that some investment spending may be wasted or spent on poor projects. The assumption here is that each dollar of public investment in Senegal increases the stock of productive infrastructure by 70 cents. This assumption is based on Briceño-Garmendia et al. (2011) and compares with a benchmark value of 0.60 used by Buffie et al. (2011) for sub-Saharan Africa.11
User fees allow governments to recoup a significant share of recurrent costs associated with the use of public capital. Even if investment has a high rate of return, it may not pay for itself from a fiscal perspective if user fees are low and the benefits of investment accrue mainly to the private sector. The estimate of 0.83 for Senegal is based on Briceño-Garmendia et al. (2011) and IMF staff estimates. According to Briceño-Garmendia et al. (2011), tariffs in the power and water sectors cover 72 percent and 91 percent of recurrent costs, respectively. Estimates of tolls collected on the Dakar-Diamniadio toll road point to 85 percent of recurrent costs covered by the collection of user fees. The estimate of 83 percent for Senegal compares to a sub-Saharan African benchmark value of 0.50 used by Buffie et al. (2011).
Steady-state ratios and other parameters were set to either their average over the past 10 years or the level at end-2011. The model generates results by comparing the projected values of certain variables with their steady-state values. For example, the steady-state value for investment spending was set to 9.5 percent of GDP, equal to the average level of public investment to GDP over the last 10 years. The model then assesses the impact of a projected increase in investment spending from the steady-state level.
Higher public investment, financed through a combination of concessional and nonconcessional external borrowing and domestic borrowing, would increase per capita income permanently but also significantly raise the debt ratio over the medium term. This scenario most closely resembles the current financing mix in Senegal, where the government finances public investment through a combination of external and domestic sources.12 The supply of concessional financing is assumed here to be finite and decreasing over time, whereas nonconcessional financing gradually increases (both variables are exogenous). Domestic financing is generated endogenously by the model to cover the remaining gap. Public debt rises to 60 percent by the 10th year of the projection period before starting to decline (Figure A.4). The impact on growth is positive but limited by some crowding out of private investment from domestic borrowing.
Figure A.4.External and Domestic Borrowing
Relying exclusively on fiscal adjustment to finance higher public investment looks unrealistic. In this first alternative scenario, the government is assumed to finance public investment by increasing consumption taxes as necessary while avoiding any recourse to borrowing. The results suggest that economic growth would accelerate and the ratio of public debt to GDP would decrease marginally, but the consumption tax rate would need to jump from 18 percent to 33 percent and remain at elevated levels over the medium term, which does not seem feasible (Figure A.5).
Figure A.5.Fiscal Adjustment Only
Source: Senegalese authorities and IMF staff estimates.
Financing higher public investment with concessional borrowing, without any fiscal adjustment, appears to be the best option from a growth and debt sustainability perspective. If the government is assumed to have unlimited access to external concessional financing to cover all planned investment spending, the results would be generally favorable. Public debt to GDP would initially rise, reaching 57 percent by the eighth year of the projection period, but trend downward thereafter (Figure A.6). The boost to growth would be even greater than in the fiscal adjustment and baseline scenarios, as consumption and private investment would not be impacted by higher taxes or domestic borrowing. The question, though, is whether unlimited access to external concessional financing is a realistic assumption. Recent history suggests that concessional resources are limited and will need to be supplemented by other types of financing.
Figure A.6.Concessional Borrowing Only
Source: Senegalese authorities and IMF staff estimates.
Relying exclusively on nonconcessional external borrowing would lead to a permanently higher debt ratio, raising debt sustainability concerns. Similar to the previous scenario, external borrowing without any fiscal adjustment would have a positive impact on growth, due to higher levels of public investment, private investment, and consumption (Figure A.7). However, the higher cost of borrowing would result in a rapid and continuous increase in the debt, with the ratio of public debt to GDP reaching 68 percent by the end of the projection period.13
Figure A.7.Nonconcessional Borrowing Only
Source: Senegalese authorities and IMF staff estimates.
The results of the model suggest that a modest scaling up of public investment would benefit growth without jeopardizing debt sustainability, provided it was not financed exclusively through nonconcessional borrowing. Financing public investment through concessional borrowing would yield even more favorable results, assuming such financing was available. It should be noted that the outcome depends critically on key structural conditions, such as absorptive capacity, the return on infrastructure investment, investment efficiency, and user fees. By improving the structural environment for public investment, Senegal could arguably pursue a more ambitious investment program without threatening debt sustainability. Conversely, a worsening of the structural environment would call for a more cautious approach.
2011, “Senegal’s Infrastructure: A Continental Perspective,” World Bank Policy Research Working Paper 5817 (Washington: World Bank).
2011, “Public Investment, Growth and Debt Sustainability: Putting Together the Pieces,” IMF Working Paper WP/12/xx (Washington: International Monetary Fund).
2005, “The Return to Foreign Aid,” Discussion Paper no. 05–04 (Copenhagen: Institute of Economics, University of Copenhagen).
1996, “Le Rôle du Capital Public dans la Croissance des Pays en Développement au cours des Annees 80,” Documents Techniques de l’OCDE 115 (Paris: Organization for Economic Cooperation and Development).
The author of this chapter is Gaston Mpatswe.
The growth accounting exercise uses a standard Cobb-Douglas production function, with an elasticity of output with respect to capital of 0.3 and an annual depreciation rate of the capital stock of 6 percent. The economically active population and labor force (from World Bank database) are used as proxies for labor input.
This assumption is critical to medium-term growth prospects; without it, per capita income would continue to increase only very slowly.
The author of this chapter is Alexei Kireyev.
The author of this chapter is Gaston Mpatswe.
Recent tariff adjustments: (i) a 6 percent increase in October 2007; (ii) a 17 percent increase and a new tariff structure in August 2008; (iii) a 12 percent reduction in January 2009; followed by (iv) an increase of 8 percent in July 2009.
World Bank, “Senegal—Poverty Diagnosis,” 2008, pp. 83–88; IMF, “Senegal: Selected Issues,” 2008.
Senegal. “Selected Issues.” IMF Country Report 08/221.
The authors of this chapter are Salifou Issoufou and Andrew Jewell.
In collaboration with Senegal’s Ministry of Economy and Finance, Direction de la Prévision et des Etudes Economiques.
According to Briceño-Garmendia et al. (2011), 70 percent of public investment spending goes to capital expenditures and the rest goes to operations and maintenance. Based on the concept of efficiency of public investment in the model, this translates into a value of 0.70 for the efficiency parameter.
Domestic borrowing in this case also includes borrowing from investors in the West African Economic and Monetary Union area.
Nonconcessional borrowing is assumed to carry a real interest rate of 6.5 percent, equal to a nominal rate of 9 percent (close to the initial yield on the 2011 Eurobond) minus a 2.5 percent world inflation rate.