A. Introduction

1. The mining industry is the main driver of Mauritanian exports. Extractive industries comprised on average about 83 percent of the country’s total exports in 2009–13. Iron ore is the largest component of Mauritania’s commodity export basket, alone accounting for over half of overall exports. The iron ore mining industry has substantial expansion plans, with the public mining company SNIM² aiming to triple its mining output in the next 10 years, with additional expansion of mining production expected from the private sector. The projected increase in iron exports will be a major driver of the projected strong output growth over the next decade.

2. The authorities have taken advantage of the recent period of high iron prices and access to concessional borrowing to ramp up investment plans. Mauritania needs to address urgent investment needs to become a middle-income country. Both job creation through economic diversification and improvement of the living standards of the population require better infrastructure, including transport and electricity generation and distribution. Further irrigation in the south of the country can increase agricultural output and increase employment. The authorities have recently increased public investments to address these needs more rapidly, taking advantage of higher fiscal revenues from extractive industries due to higher commodity prices during 2012–14 and available, mostly concessional, borrowing.

3. Investment in Mauritania, however, is subject to the usual capacity constraints prevalent in lower-income countries. An IMF study (Dabla-Norris et al. (2011)) calculated a Public Investment Management Index (PIMI) for a number of emerging markets and lower-income countries. PIMI is a composite index of the efficiency of the public investment management process that draws upon country diagnostics on public investment management systems conducted by the World Bank, existing budget survey databases and assessments carried out by donors, and expert surveys (Figure 1). The index evaluates four stages of public investment: appraisal, selection, managing and evaluation. Mauritania is positioned roughly in the middle of the PIMI sample of countries (and behind, for instance, Tunisia but ahead of Egypt). While the country can achieve positive results through public investments, capacity to invest efficiently is an issue to consider when making investment decisions, and its improvement can lead to significant benefits.

Public Investment Management Index (PIMI)

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Public Investment Management Index (PIMI)

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF

• Download Figure
• Download figure as PowerPoint slide

Public Investment Management Index (PIMI)

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF

• Download Figure
• Download figure as PowerPoint slide

B. Model Description

4. The DIGNAR model provides a framework in which to analyze the benefits of higher investments versus their costs in terms of the impact on a country’s debt level.³ The model is designed to analyze the nexus between natural resources revenues management, public investment, and public debt in the context of investment capacity constraints. The analytical framework is based on Araujo et al. (2013), Buffie et al. (2012), Berg et al. (2013) and Melina et al. (2014) and includes a natural resource sector, limited investment efficiency and absorptive capacity as well as a detailed fiscal specification reflecting the choices by the authorities in financing investments through external debt, taxes and drawings from the sovereign wealth fund (currently Mauritania’s oil fund, which was set up at the start of the short-lived oil boom in 2006 and amounted to almost US$100 million, equivalent to 2 percent of GDP in 2014). Taking resource revenues and public investment policy as given, the framework can simulate the macroeconomic outcomes of investing resource revenue, accounting for the investment-growth linkage and the feedback effect on non-resource revenue. In allowing the evaluation of the level of future private consumption after public investment has borne its fruit, against the cost of foregone current private consumption resulting from financing that investment through taxes, the model can also provide insights regarding the intergenerational impact of the public investment program.

5. The DIGNAR model provides a stylized representation of a small open economy with iron production and public investment needs like the Mauritanian economy. The particular features of the model are:

• The government finances its consumption and investment expenditure with taxes, debt and savings accumulated in the sovereign wealth fund (SWF). The path of external concessional debt is given by the projection of repayment of the existing stock of concessional debt and new concessional debt disbursements consistent with those in the debt sustainability analysis (DSA) based on the needs of the public investment program. The government thus only decides the level of non-concessional (commercial) external debt. Commercial foreign debt is subject to a risk premium that depends on the stock of external debt. Fiscal revenues are collected from the extractive and non-extractive sectors. The government chooses the consumption tax (VAT) rate, starting from the initial level of 14 percent in place at the end of 2014. In every period, the government can choose to close the fiscal gap with an increase in taxes and/or debt, as well as by drawing down savings in the SWF. While Mauritania’s SWF is currently the country’s oil fund, the model application assumes that, should the government collect substantial revenues from the extractive sector of the economy, it would wish to save some of the receipts and could do so in a more general extractive SWF.

• The physical capital formation process⁴ is subject to absorptive capacity and government efficiency constraints.⁵ In particular, effective government investment is a fraction of government expenditure on investment. In the Mauritanian model calibration, investment efficiency (the ratio of effective investment to investment expenditure) is currently assumed at 0.5 in the non-natural resource sector, a value suggested for other lower-income countries, and rising over the next decade to 0.6, a level applied in the cases of some transition economies (e.g. Kazakhstan).⁶ This increase assumes that Mauritania will implement substantial improvements to its public investment management process over the coming ten years across the four areas used to calculate the PIMI index and, especially, in project management and appraisal, where it is lagging behind its peers. Furthermore, to capture the idea of rising investment costs due to absorptive capacity constraints, investment efficiency is assumed to fall when the expenditure level exceeds a certain threshold.⁷

• Two types of households. Rule-of-thumb consumers are liquidity constrained and consume all of their disposable income in each period. Optimizing households are subject to borrowing constraints and have only limited access to financial markets to acquire international bonds with portfolio adjustment costs, which restrict the degree of capital account openness. The private sector pays a premium on foreign debt over the interest rate that the government pays on its own external debt.

• Three production sectors. In the non-extractive economy traded and non-traded goods are produced according to a Cobb-Douglas production function with three inputs: labor, private capital and public capital. The difference between these two sectors resides in the modeling of technological progress—assumed to be exogenous in the non-traded sector whereas in the traded sector it is subject to learning-by-doing externalities and depends positively on previous-period output. The intuition is that once traded-sector production starts falling, knowledge and skills can be lost. Growth in both sectors of the non-extractive economy is endogenously determined as a function of behavior of firms and households, which in turn respond to government policies and outcomes in the third sector. The latter, the extractive sector, produces iron ore and output is assumed to be exogenous, with both price and quantities taken as given. Given iron ore’s dominant role in Mauritania’s extractive sector and the projected rise in iron ore production, iron ore mining is taken to represent Mauritania’s entire extractive sector.

C. Investment and Revenue Scenarios

6. The DIGNAR analysis of Mauritanian public investments is based on a 2x2 matrix of scenarios, considering a baseline and an aggressive investment path under baseline and downside iron ore price projections.

• Baseline investment path is consistent with the projections in the staff report, which assumes a high level of public investments over the next three years. Public investment as a percent of GDP remains anchored at the 2018 level (around 10 percent of GDP) afterward. This path is driven by the authorities’ investment plans in conjunction with a projection for the extractive sector output and extractive fiscal revenues as well as new external debt disbursements embedded in the DSA.

• Aggressive investment path assumes that the authorities continue investing at the baseline peak level currently projected for 2015, at almost 15 percent of GDP, through 2018 before starting the decline projected in the baseline to reach a level of about 10 percent of GDP by 2021.

• Baseline iron ore price projection is based on an iron ore price of USD72 per ton, which is close to the prevailing price in the world commodity markets in late 2014. This projection also assumes that all currently planned mining projects come to execution and start production within one year of the scheduled date. This will more than triple the iron ore production of Mauritania over the next decade and production will be anchored at over 65 million tons per year by 2030.

• Downside iron ore price projection assumes a further 10 percent decline in prices to about USD 65 per ton. This is close to futures market prices in 2015-16 as of end-2014. It is assumed that the lower price would lead to canceling of at least one substantial mining project and delays in the implementing others. The overall iron ore mining capacity will still increase substantially over the coming years to 44 million tons by 2026.

Mauritania: Iron Ore Projections, 2014–29

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mauritania: Iron Ore Projections, 2014–29

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

Mauritania: Iron Ore Projections, 2014–29

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

D. Results

7. The simulations show that the ambitious public investment plans would increase output but also lead to a decrease in private consumption under current commodity prices.

• The results for the baseline investment path under the baseline iron ore price projections suggest that the authorities’ current investment plans are sustainable but hurt private consumption. Given the high poverty levels in Mauritania, foregone consumption will be costly. The model would finance the beginning of the investment plan with a combination of additional debt, and drawings on the resources currently in the oil fund, later replacing the latter source with an increase in consumption taxes once the oil fund runs out of resources. This policy change will decrease private consumption by about one percent below its steady state value in the medium run. The lower consumption will in turn reduce non-extractive output to its steady-state value after an initial rise. Only once the investments have begun fully bearing fruit in 2019 will consumption start increasing, returning to the steady state value by 2028, with non-extractive output rising almost one percent above its steady state. In the meantime, total public external debt rises substantially in the coming two years, peaking at slightly over 70 percent of GDP, but then stabilizes and begins steadily declining by 2019.

• If iron prices fall further and mining expansion projects are delayed, the current investment plan will become very costly in terms of foregone private consumption and will reduce nonextractive output. As extractive revenues of the government fall—due to the lower price and output in iron mining relative to the baseline scenario—the authorities will have to increase taxes and public external debt further. This will result in a sharp fall in private consumption and a substantial decrease in non-extractive output in the medium run. At the same time, public external debt will increase further and then decrease more slowly than in the baseline price scenario. Overall, these results suggest that the authorities’ baseline investment plans would be very costly to maintain in terms of foregone private consumption should commodity prices decrease further⁸

• A more aggressive investment path would result in higher output at a substantially higher foregone consumption cost in the medium run. A longer period of scaled-up investments would succeed in increasing non-extractive output more than 2.5 percent above the steady state, almost three times the difference achieved in the baseline investment scenario. Yet, this would come at the cost of substantially higher consumption taxes required to finance the investment path, which would, in turn, depress private consumption by up to 3 percent below its steady state value. Consumption would eventually recover, but the overall consumption loss in the medium and long run would be very large and likely exceed the benefits of the successful efforts to increase non-extractive output.

• A further iron ore price decline would be particularly damaging to private consumption levels if the authorities pursue a more aggressive investment path. The negative impact on consumption would effectively combine the declines seen in the previous two scenarios, resulting in a steep drop in the next five years followed by further stagnation and even decrease in the long run. Non-extractive output would still initially increase above its steady-state level as a result of the higher investments but then fall back as a result of the fall in private consumption. The level of external public debt would rise above 80 percent of GDP by 2020 and only slowly decrease after that. Overall, the costs in terms of consumption as well as the debt dynamics of maintaining an aggressive public investment plan in the face of lower iron prices would be very large.

Mauritania: Model Simulations, 2014-29

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF Staff estimates

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mauritania: Model Simulations, 2014-29

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF Staff estimates

• Download Figure
• Download figure as PowerPoint slide

Mauritania: Model Simulations, 2014-29

Citation: IMF Staff Country Reports 2015, 036; 10.5089/9781498354226.002.A004

Source: IMF Staff estimates

• Download Figure
• Download figure as PowerPoint slide

E. Conclusions and Policy Implications

8. The DIGNAR application points to the potential costs in terms of foregone private consumption of a scaled-up public investment plan in Mauritania, calling for flexibility in implementation should extractive revenues fall short of expectations. While higher levels of public investments will benefit the country in terms of non-extractive output and eventually private consumption, in the medium run their cost will be substantial. In particular, their financing will require higher debt and tax levels, the latter of which will depress private consumption even under the baseline investment plans and price projections. The model suggests that a higher-than-currently-envisaged investment path would have even larger costs. Even in the baseline investment path, the authorities will need to remain flexible in implementation in the face of volatile global commodity prices. Should prices decrease further and mining expansion projects be delayed or scaled back, too much of the planned investments would have to be financed through higher debt or taxes, potentially hurting private consumption even in the long run.

9. Further improvements in investment capacity would increase the benefits of public investments in Mauritania and ultimately reduce their costs. The model application already assumes an improvement over the next decade in the country’s public investment efficiency. A faster or more substantial improvement would increase the positive impact of the planned investment path on the economy, or alternatively achieve the same results as in the model baseline at lower foregone consumption costs. The components of the PIMI index with particularly low values relative to peers—those capturing project management and appraisal—suggest starting points for efforts to further improve public investment efficiency in Mauritania.