Journal Issue
Share
Article

Chad: Selected Issues

Author(s):
International Monetary Fund
Published Date:
February 2009
Share
  • ShareShare
Show Summary Details

CHAPTER II An Assessment of Fiscal Sustainability for Chad

I. Introduction

43. The pursuit of sustainable fiscal policies is critical for oil exporters. The costs of failing to pursue sustainable fiscal policies in oil exporting countries has been dramatically illustrated in the past by the boom-bust cycles of the 1970s and 80s. In those episodes, countries that had excessively expanded their spending and borrowed on the back of high oil revenues had to abruptly adjust their fiscal policy when oil prices declined sharply and debt repayments came due. This sharp fiscal adjustment led to severe negative consequences for growth. In this regard, an assessment of fiscal sustainability is an important element of the surveillance agenda for an oil exporter such as Chad, particularly given limited oil wealth and highly volatile oil prices, as illustrated by recent experience.

44. Chad is a new oil producer with relatively small proven reserves and high extraction rates. Chad started to produce and export oil from its one producing region in 2003, with revenues commencing fully in 2004. Proven reserves are estimated at about 600 million barrels and are expected to be depleted by 2032. In this regard, Chad’s oil production is currently smaller than most other African oil exporters. Extraction rates in Chad, proxied by production as a share of reserves, are about 8 percent, which is high compared with other region producers.

45. Chad is confronted with key policy challenges that are well known to countries abundant in natural resources. Key challenges include (i) managing volatile and unpredictable oil revenue flows to prevent pronounced fluctuations in economic activity and the real exchange rate: (ii) addressing the exhaustible nature of oil resources with prudent intertemporal decisions on consumption and saving, and (iii) ensuring a high return to public spending, through consistency with absorptive capacity and a sustainable fiscal deficit. In this regard, oil producers have often succumbed to strong political economy pressures for large government spending increases during periods of positive terms of trade shocks, which prove difficult to reverse when oil prices are less favorable or production declines.

46. This paper examines how Chad has fared on these challenges by providing an assessment of fiscal sustainability and progress on the policies to ensure such sustainability. The paper has three sections. In the first section, fiscal policy in Chad in recent years is compared with that in 28 other oil producing countries, including in a historical perspective of the boom-bust cycles of the 1970s that culminated in the debt crisis of the 1980s. In the second, section the paper goes beyond relative comparisons with other oil countries, and provides and assessment of whether fiscal policy (measured by the non-oil primary deficit (NOPD)) is sustainable under various permanent income hypothesis type approaches. A final section concludes and discusses briefly reforms to ensure fiscal sustainability in Chad.

II. Fiscal Developments Over the Oil Era

47. Fiscal policy has been expansionary since the onset of oil production (Figure 1). The NOPD has deteriorated from about 3 percent of non-oil GDP in 2003 to about 22 percent of non-oil GDP on a commitment basis in 2007.13 Given an improvement in the level of non-oil revenues from about 9 percent of non-oil GDP in 2003 to 11 percent of non-oil GDP in 2007,14 the deterioration in the NOPD is mainly accounted for by an increase in domestically financed primary spending of about 20 percentage points of non-oil GDP. The large increase in oil revenues of over 27 percent of non-oil GDP in the same period financed this large increase in domestic primary spending in spite of a significant reduction in foreign financing of 7 percentage points of non-oil GDP.

Figure 1.Evolution of Selected Fiscal Aggregates-2003-2008 (Proj.)

Source: Chadian Authorities and Fund Staff Estimates

48. A large expansion in primary current spending and domestically financed capital spending were mainly responsible for the fiscal expansion. Primary current spending increased by about 14 percentage points of non-oil GDP, while domestically financed capital spending increased by about 6 percentage points in the period 2003–2007. Higher security spending in the form of increases in the number of military personnel, increases in goods and services to support a larger troop contingent, and in exceptional security spending,15 accounted for about 56 percent of the overall increase in primary current spending during 2003-07 (Figure 2). The remainder of the overall increase derives mainly from increased transfers to finance losses of the state cotton and energy companies (Cottonchad and STEE respectively) and a major increase in the civil service wage bill. The latter accounted for 55 percent of the overall increase in the wage bill and 29 percent of the total increase in the number of public sector workers during 2003-2007.

Figure 2.Evolution of main components of Primary Current Spending-2003-2008 (Proj.)

Source: Chadian Authorities and Fund Staff Estimates.

49. Primary current spending and the wage bill have significantly increased as a share of non-oil revenues. Primary current spending as a share of non-oil revenues has risen from the already worrisome level of 112 percent in 2003 to a level of 224 percent in 2007. In particular, the wage bill as a share of non-oil revenues has increased from about 60 percent in 2003 to 80 percent in 2007. These ratios highlight a clear vulnerability to sustaining current spending levels if oil revenues were to decline significantly.

50. Public debt has fallen as a share of non-oil GDP. The significant improvement in the overall balance generated by the oil revenue increase (Figure 3) led to a significant reduction in financing needs. Combined with the continuous servicing of debt and the decline in foreign financing as a share of non-oil GDP observed during 2003 and 2007, this led the nominal external debt to grow less than 10 percent in nominal terms, from CFAF 736 billion in 2003 (54 percent of non-oil GDP) to CFAF 794 billion in 2007 (43 percent of non-oil GDP). A gradual repayment of domestic debt including arrears also contributed to the decline in domestic debt, which fell from 142 billion in 2005 (9 percent of non-oil GDP) to 99 billion in 2007 (5 percent of non-oil GDP).16

Figure 3.Evolution of the Overall Balance17 2003-2008 (Proj.)

Source: Chadian Authorities and Fund Staff Estimates.

51. These trends in fiscal indicators have been broadly maintained in 2008. The NOPD is expected to have reached 28 percent of non-oil GDP mostly due to further increases in primary current spending driven by higher security spending. The February 2008 attacks on the capital and the desire to preempt further attacks led to additional troop hiring, wage increases for the military, and additional exceptional security spending (mainly military equipment purchases) of 6 percentage points of non-oil GDP in additional current spending. These developments resulted in a further increases in primary current spending and the wage bill as a share of non-oil revenues to 297 percent and 94 percent, respectively. The overall balance is expected to have further improved to about 8 percentage points of non-oil GDP, and external and domestic debt levels reduced further to 37 percent of non-oil GDP and 3 percent of non-oil GDP, respectively.

52. Chad has spent more of the increase in oil revenues it experienced than most oil producers in the recent oil price boom. Following Ossowski et. al. (2008), Figure 4 below graphs the ratio of the cumulative change in the non-oil primary balances against the cumulative increase in oil revenues as a share of GDP during the period 2000-2006.18 Up to 2006, Chad had spent about 70 percent of the total accumulated oil revenues earned between 2000 and 2006. This is more than most oil producers in the region with the exception of Sudan. With respect to the sample of other 28 oil producing countries, only 7 countries spent more than Chad during the 2000-2006 period.19

Figure 4.Utilization of Oil Revenues by Oil Producing Countries 2000-2006

53. Weaknesses in public financial management (PFM) have exacerbated problems in budget control and spending. Main weaknesses are: (i) abuses of exceptional and extra-budgetary spending procedures; (ii) lengthy and poorly-enforced public procurement processes combined with redundant and ineffective controls encouraging the use of noncompetitive bidding and the circumvention of the budget legal framework; (iii) deficiencies in the budget classification; (iv) deficiencies in planning, budgeting and monitoring of multiyear investment projects; (v) poor cash management, lack of effective commitment controls, and earmarking-based fragmentation of budget procedures; and (vi) insufficient fiscal transparency.

54. The behavior of Chad’s major fiscal aggregates in the light of international experience raises questions about fiscal sustainability. The evolution of Chad’s fiscal aggregates during the recent high oil price period exhibits important similarities with countries that were forced to undertake costly fiscal adjustment during the 1970s/80s boom-bust cycle. In addition to a large expansion of the non-oil primary deficit and spending, several of these countries experienced a growing wage bill (such as in the Republic of Congo, Trinidad and Tobago), a rapidly growing investment budget with questionable rates of return on many projects and over the availability of recurrent spending allocations to ensure infrastructure was maintained (such as in the Republic of Congo, Gabon, Nigeria, Trinidad Tobago), and increasing subsidies to inefficient sectors (see Gylfanson (2001) for examples and discussion).20 One exception is Chad’s relatively low public debt ratios, which in part reflect the lack of lending opportunities during decades of civil war, but also some degree of prudent debt management.

III. Assessing Sustainability and the Sustainable Non-Oil Primary Balance

55. A variety of approaches have been discussed to formulate sustainable medium-term fiscal policy in oil producing countries. Sustainable fiscal policy is defined as one where the intertemporal budget constraint of the government (i.e. the present value of government spending needs to be less or equal to the present value of non-oil revenues plus net government assets21) is satisfied in the long term while meeting financing constraints in the short term. As highlighted in Katz et al. (2004), most approaches entail guiding fiscal policy with explicit (i.e. when enforced by law) or implicit rules such as for example a “bird in hand” or balanced budget approach. A “bird in hand” policy is one that is sustainable by definition, since it consists of only spending the interest earned on existing financial assets resulting from oil savings. A balanced budget rule is also typically considered sustainable since it involves adjusting spending to the level of revenues.22 These rules are sometimes viewed as extremes since the first implies saving a large fraction of oil revenue, while the second implies spending all oil revenue as it materializes.

56. The permanent income hypothesis (PIH) approach is an important method to derive rules to guide medium-term fiscal policy. It’s importance derives from it providing rules which address key fiscal policy challenges facing oil producers. By linking government consumption to oil wealth rather than current oil revenues, it provides rules to guide fiscal policy that largely delink government consumption from short-term fluctuations in oil prices, thereby smoothing the government’s consumption path. A more stable consumption path helps address the concern of fiscal policy exacerbating the volatility of economic activity and the real exchange rate generated by oil price fluctuations, and reduces the risks of exceeding absorptive capacity constraints that accompany rapid increases in spending. Finally, it allows incorporation of different welfare criteria with relative ease through welfare functions. The PIH approach requires a chosen welfare function to be maximized subject to the intertemporal budget constraint to derive fiscal policy rules. Since PIH derived fiscal policy rules emerge from a welfare maximization, the PIH approach adds a sense of optimality to the sustainability concept allowing to discriminate among rules which imply sustainable paths.23

57. The PIH approach has been applied to several oil producing countries. This includes countries facing similar challenges to Chad, such as Congo, Cameroon and Gabon.24 This is because, in addition to the advantages discussed above, medium-term fiscal frameworks guided by PIH are flexible in accommodating country-specific circumstances. Convergence to a sustainable NOPD may be achieved by front-loaded fiscal adjustment (such as in one period adjustment models as in Barnett and Ossowski (2003)) or more gradually to minimize adjustment costs or allow for habit persistence (such as in Leigh and Olters (2006) or Carcillo et al. (2007)). Also, a low income country could opt for gradualism permitting higher levels of investment spending early on (as discussed in Takizawa et al. (2004)). However, unless that investment is productive not only in terms of a social rate of return but also in terms of future revenues, the delayed adjustment implies a lower sustainable NOPD. PIH approaches can also easily allow the assumption of oil wealth being consumed over a relatively long but finite time period to capture political economy considerations (as suggested in Clausen (2007)).

58. This section provides estimates of sustainable long-term NOPD that would result if some of these approaches were applied to Chad. Four different types of rules were simulated: a bird in hand rule, a balanced budget rule, and two types of rules derived from the PIH approach.25 The first PIH derived rule is one where a flat government consumption annuity as a share of non-oil GDP is computed on the basis of an estimate of oil wealth and a finite number of years in which oil wealth would be consumed (further discussion below).26 The long-run sustainable NOPD is then computed by subtracting from an assumed level of non-oil revenues the level of primary spending which is a function of the annuity and an assumed level of non-oil revenues. Adjustment to a long-term NOPD is front loaded with this rule, since government consumption would need to adjust in one period to the long term sustainable level. The second PIH derived rule is more realistic, and entails adjustment to a sustainable long-term NOPD gradually in the context of an infinite horizon model (Carcillo et al. (2007). In the latter, government primary consumption evolves according to a weighted average of (i) the level of government primary consumption that could be financed by wealth27 if adjustment were done in one period (i.e. no habit persistence) and (ii) the level of government primary consumption in the previous period (see Appendix for the precise formula). A steady state level of government primary consumption is achieved after a certain number of years depending on how fast the adjustment takes place. The non-oil primary balance path is then computed by subtracting from a non-oil revenue assumption the government primary consumption level. The weights in the government primary consumption formula depend on how important habit persistence or adjustment costs are assumed to be. The higher the habit persistence/adjustment costs coefficient assumed, the more gradual the adjustment is. This comes at the expense of a lower long term sustainable NOPD given that the model does not assume productive government investment.28

59. The annuity calculation is a key part underlying the estimates of government consumption for most PIH rules presented in the paper.29 The annuity refers to an implied stream of payments over a specified period of time which can be finite or infinite (proxying for very long time periods). The annuity is used to transform a projected stream of oil revenues into a hypothetical government consumption annuity with the same present value. Therefore, the calculation is done in three steps (i) calculate present value of the oil revenue stream under the assumption of a certain interest rate for each year; (ii) set the total present value of oil revenue equal to the total present value of the hypothetical government consumption annuity; and (iii) express the government consumption annuity relative to a certain denominator.30Figure 5 illustrates the case of a government consumption annuity which is constant as a which is constant as a share of non oil GDP.

Figure 5.Oil Revenue, Annuity, and Asset Accumulation

(Constant Government Consumption Annuity as a Share of Non Oil GDP)

60. The main oil sector and macroeconomic assumptions that entered the baseline calculation of NOPDs are broadly consistent with the baseline scenario of the staff report and DSA (Table 1).31

Table 1.Main Macroeconomic, Oil Sector, and Habit Persistence Parameter Assumptions
2009–2012

Average
2013–2016

Average
2017–2020

Average
2021–2045

Average
2046–2070

Average
Oil Sector Assumptions
Reference oil price (dollars per barrel) 1/66.379.686.6101.4--
Oil Production (millions of barrels) 1/45.035.420.312.0--
Tax take (net oil revenues as a share of oil sales valued at the Doba after well price) 1/35363636--
Macroeconomic Variables
Real Non Oil GDP growth (in percent)4.74.44.13.23.0
Real Interest Rate (in percent)4.04.04.04.04.0
Non oil revenues (in percent of non oil GDP)10.011.112.216.417.5
Other Parameters
Habit persistence coefficient assumption for the estimation using the Carcillo et al. (2007) model0.60.60.60.60.6
Source: Fund Staff Estimates

2021-2045 shows only the average until 2032, the expected last year of production.

Source: Fund Staff Estimates

2021-2045 shows only the average until 2032, the expected last year of production.

61. Results suggest that Chad’s current NOPD levels exceed sustainable levels by a large margin (Table 2). The results in the table below suggests that for all the approaches estimated the long run sustainable NOPD ranges from a small surplus to low single digit deficits.32 The results for the PIH approach derived rules also illustrate the cost of gradualism in fiscal adjustment, since following a more gradual approach results in a lower sustainable long run NOPD. The key result is that irrespective of the approach, these NOPD levels are far smaller than Chad’s expected NOPD in 2008, which is projected to be about 28 percent of non-oil GDP.

Table 2.Average Levels of Non-Oil Primary Deficits Under Different Approaches
2009–20122013–20162017–20202021–20452046–2070
Bird in Hand−1.4−4.4−5.1−2.80.0
Balanced Budget−10.1−8.5−3.7−0.10.9
PIH: Villafuerte, Leigh and Carcillo (2007) (infinite horizon)−13.1−8.9−6.5−1.00.4
PIH: Constant government consumption annuity as a share of non oil GDP (finite horizon)−1.1−2.3−2.6−2.6−2.6
Source: Fund Staff Estimates.
Source: Fund Staff Estimates.

62. The results are robust to sensitivity analysis on the parameters of the models including on oil revenues (Table 3). For all the sensitivity tests on the parameters, results imply that the long term NOPDs would remain in the low single digits.33 In particular, larger percentage term deviations from the baseline are considered for oil prices and production. This is to illustrate that even if oil prices or production were significantly more favorable than assumed in the baseline (for example due new production or significantly more favorable oil prices), given the current relatively low levels of oil production, long-term NOPDs would remain in the low single digits.

Table 3:Sensitivity Analysis of NOPD Estimates
2009-2012

Average
2013-2016

Average
2017-2020

Average
2021-204

Average
2046-2070

Average
2009-2012

Average
2013-2016

Average
2017-2020

Average
2021-2045

Average
2046-2070

Average
Bird in Hand (1)Balanced Budget (2)
Baseline Parameters−1.4−4.4−5.1−2.80.0−10.1−8.5−3.7−0.10.9
Sensitivity Tests
Oil Reference Price
Higher Oil prices (30 percent higher each year than baseline starting in 2009)−2.3−7.1−8.3−5.0−0.6−19.9−14.8−6.7−0.70.9
Lower Oil prices (30 percent lower each year than baseline starting in 2009)−0.4−1.6−1.7−0.50.6−0.4−1.6−0.20.60.9
Oil Production
Higher Oil Production (30 percent higher each year than baseline starting in 2009)−2.0−6.4−7.5−4.5−0.4−16.9−13.7−6.1−0.60.9
Lower Oil Produciton (30 percent lower each year than baseline starting in 2009)−0.8−2.4−2.6−1.20.4−3.4−3.3−1.20.40.9
Tax Take
Higher Tax Take (10 percent higher each year than baseline starting 2014)−1.4−4.5−5.3−3.0−0.1−10.1−9.2−4.1−0.20.9
Lower Tax Take (10 percent lower each year than baseline starting 2014)−1.4−4.4−4.9−2.60.0−10.1−7.8−3.20.00.9
Non oil Revenues
Higher Non Oil Revenues (10 percent higher each year than baseline starting 2009)−1.4−4.4−5.1−2.80.0−10.1−8.5−3.7−0.10.9
Higher Non Oil Revenues (10 percent lower each year than baseline starting 2009)−1.4−4.4−5.1−2.80.0−10.1−8.5−3.7−0.10.9
Real Interest Rate
Higher real interest rate (=4.5 percent)−1.6−4.8−5.5−3.1−0.1−10.1−8.5−3.7−0.10.9
Lower real interest rate (=3.5 percent)−1.1−4.0−4.6−2.60.1−10.1−8.5−3.6−0.10.9
Long run real Growth Rate
Higher long run real growth rate (=3.5 percent)−1.4−4.4−5.0−2.70.1−10.1−8.5−3.6−0.10.9
Lower long run real growth rate (=2.5 percent)−1.4−4.4−5.1−3.0−0.1−10.1−8.5−3.7−0.11.0
Source: Fund Staff Estimates
Source: Fund Staff Estimates
2009–2012

Average
2013–2016

Average
2017–2020

Average
2021–2045

Average
2046–2070

Average
2009–2012

Average
2013–2016

Average
2017–2020

Average
2021–2045

Average
2046–2070

Average
Villafuerte, Leigh and Carcillo (2007)

(Infinite Horizon) (3)
Constant government consumption annuity

as a share of non oil GDP (4)
Baseline Parameters−13.1−8.9−6.5−1.00.4−1.1−2.3−2.6−2.6−2.6
Sensitivity Tests
Oil Reference Price
Higher Oil prices (30 percent higher each year than baseline starting in 2009)−13.8−9.7−7.4−1.7−0.4−2.9−4.1−4.4−4.5−4.5
Lower Oil prices (30 percent lower each year than baseline starting in 2009)−12.4−8.0−5.7−0.21.20.9−0.2−0.5−0.6−0.6
Oil Production
Higher Oil Production (30 percent higher each year than baseline starting in 2009)−13.7−9.6−7.2−1.6−0.2−2.5−3.7−4.0−4.1−4.1
Lower Oil Produciton (30 percent lower each year than baseline starting in 2009)−12.6−8.3−5.9−0.40.90.4−0.8−1.1−1.2−1.2
Tax Take
Higher Tax Take (10 percent higher each year than baseline starting 2014)−13.2−9.0−6.6−1.00.3−1.2−2.4−2.7−2.8−2.7
Lower Tax Take (10 percent lower each year than baseline starting 2014)−13.1−8.9−6.5−0.90.4−0.9−2.1−2.4−2.5−2.5
Non oil Revenues
Higher Non Oil Revenues (10 percent higher each year than baseline starting 2009)−13.7−9.6−7.1−1.00.5−1.1−2.2−2.5−2.6−2.6
Higher Non Oil Revenues (10 percent lower each year than baseline starting 2009)−12.5−8.2−6.0−1.00.3−1.1−2.2−2.5−2.6−2.6
Real Interest Rate
Higher real interest rate (=4.5 percent)−13.3−9.1−6.7−1.20.2−1.5−2.6−3.0−3.0−3.0
Lower real interest rate (=3.5 percent)−13.0−8.7−6.4−0.80.5−0.7−1.8−2.2−2.2−2.2
Long run real Growth Rate
Higher long run real growth rate (=3.5 percent)−12.4−8.4−6.3−1.10.2−0.7−1.9−2.2−2.3−2.3
Lower long run real growth rate (=2.5 percent)−13.9−9.4−6.7−0.80.6−1.5−2.6−2.9−3.0−3.0
Habit persistence
Larger Habit parameter (=0.7)−14.5−9.7−6.8−0.90.5
Smaller Habit parameter (=0.5)−12.1−8.6−6.4−1.00.3
Source: Fund Staff Estimates
Source: Fund Staff Estimates

IV. Reforms to Restore Fiscal Sustainability in Chad

63. A significant reduction of the NOPD towards sustainable long term levels is urgently needed in Chad. The rapid growth in the NOPD in recent years to well beyond sustainable levels, high ratios of primary spending and the wage bill as a share of non-oil revenues, and expenditure of a high proportion of oil revenues, are all indicative of the need to adjust. While the difficult security situation contributed to these developments generating large security spending increases, the oil windfall has led to large spending increases across all other categories, including continued financing of loss making state enterprises and other transfer programs, the wage bill, and a very large expansion of domestically financed investment. Beyond fiscal policy, weaknesses in budget control have been an important factor too.

64. The fiscal policies pursued in the recent past and a declining outlook for oil prices confront Chad with a real expenditure rationalization problem. Given that an insufficient cushion of savings was built in the recent high oil price years to preserve expenditures during a period of decline in oil prices, and that spending remains high in relation to non-oil revenue, a significant portion of the adjustment will need to come from expenditure rationalization.

65. A permanent income framework that allows for gradual adjustment seems well suited to guide Chad’s path towards a sustainable long term NOPD level. Given Chad’s large investment and priority sector spending needs and large NOPD, fiscal frameworks guided by rules such as a constant government annuity as a share of GDP, a bird in hand approach or any other rule that essentially entails eliminating most of the NOPD in a short period of time (see Table 2 and Appendix Table 1), may be both infeasible and undesirable. Balanced budget rules are also problematic given that the large revenue fluctuations experienced by oil exporters would imply large expansions in spending when prices are high and large cuts when revenues decline, exacerbating fluctuations in aggregate demand. The recommended gradual PIH adjustment guidance should be complemented by ensuring that the adjustment path is consistent with short-term financing constraints.

66. Several important reforms are needed to achieve a considerable reduction in the NOPD. These include:

  • Revenue administration reform and a review tax policy to strengthen the level of non-oil revenues. With respect to the revenue administration, this includes strengthening the operations of the large taxpayers unit, improving the selection criteria and quality of the audits, simplifying and automatizing customs procedures and ensuring that both the tax administration and customs have the material and human resources to carry out its functions appropriately. On tax policy, the review should focus on broadening the tax base including by reassessing the need for tax incentives and exemptions, and simplifying the tax structure to eliminate taxes with small collections that increase the administrative burden.

  • Expenditure rationalization with a view to enhance the quality and poverty orientation of spending. Efforts in this area should be focused on implementing long standing reforms to rationalize the wage bill (biometric census of the civil service, computerization of the payroll, using the banking system to carry out payments, and demobilization of army personnel if the security situation permits), and reduce the cost of subsidies to loss making public enterprises through structural reforms (including introduction of an adequate billing system to improve collections and moving forward with ongoing initiatives to reduce fuel costs in STEE). A critical challenge is to improve the quality, cost and consistency with absorptive capacity of the investment budget. To this end, projects with sufficiently high rates of return should be selected through careful PRSP consistent prioritization, sufficient allocations should be provided for needed recurrent expenditures, non competitive procurement should be reduced, and ongoing projects should be completed before starting new ones.

  • Public Financial Management reforms to improve budget control and medium-term fiscal planning. This includes strict enforcement of budget procedures to avoid spending commitments without appropriations, issuance of clear regulations reducing the recourse to exceptional budget procedures, simplification and computerization of the spending chain to avoid undue delays in budget execution and procurement, introduction of an international standard budget classification to better track poverty reducing spending, strengthening the planning, budgeting and monitoring of execution of investment projects so that sufficient allocations are provided to ensure the finalization of ongoing multi-year investment projects, preparation of monthly treasury plans to improve cash management, and regular approval of budget review laws and publication of budget execution reports to enhance transparency and accountability.

Appendix II
Appendix Table 1.Average Level of NOPD Under Additional PIH Approach Rules
2009–20122013–20162017–20202021–20452046–2070
PIH: Constant real government consumption (infinite horizon)−3.7−3.9−3.4−1.9−0.4
PIH: Constant real per capita government consumption (infinite horizon)0.0−1.0−1.1−1.0−0.7
PIH: Constant real government PIH: Constant real government−4.3−4.3−3.8−2.1−0.5
PIH: Constant real per capita government consumption (finite horizon)−1.6−2.5−2.6−2.2−1.8
Source: Fund Staff Estimates
Source: Fund Staff Estimates
Appendix Table 2.Sensitivity Analysis of NOPD Under Additional PIH Approach Rules
2009-2012

Average
2013-2016

Average
2017-2020

Average
2021-204

Average
2046-2070

Average
2009-2012

Average
2013-2016

Average
2017-2020

Average
2021-2045

Average
2046-2070

Average
Constant real government consumptionConstant real per capita government consumption
(infinite horizon)(infinite horizon)
Baseline Parameters−3.7−3.9−3.4−1.9−0.40.0−1.0−1.1−1.0−0.7
Sensitivity Tests
Oil Reference Price
Higher Oil prices (30 percent higher each year than baseline starting in 2009)−6.9−6.6−5.7−3.3−1.1−1.3−2.2−2.3−2.0−1.6
Lower Oil prices (30 percent lower each year than baseline starting in 2009)−0.2−1.0−0.9−0.30.41.40.30.00.10.2
Oil Production
Higher Oil Production (30 percent higher each year than baseline starting in 2009)−6.2−5.9−5.2−3.0−0.9−1.0−1.9−2.0−1.7−1.4
Lower Oil Produciton (30 percent lower each year than baseline starting in 2009)−1.2−1.8−1.6−0.70.21.0−0.1−0.3−0.20.0
Tax Take
Higher Tax Take (10 percent higher each year than baseline starting 2014)−3.9−4.1−3.6−2.0−0.4−0.1−1.1−1.2−1.0−0.8
Lower Tax Take (10 percent lower each year than baseline starting 2014)−3.5−3.7−3.2−1.7−0.30.1−0.9−1.1−0.9−0.6
Non oil Revenues
Higher Non Oil Revenues (10 percent higher each year than baseline starting 2009)−3.7−3.9−3.4−1.9−0.40.0−1.0−1.1−1.0−0.7
Higher Non Oil Revenues (10 percent lower each year than baseline starting 2009)−3.7−3.9−3.4−1.9−0.40.0−1.0−1.1−1.0−0.7
Real Interest Rate
Higher real interest rate (=4.5 percent)−4.3−4.3−3.8−2.1−0.5−0.7−1.6−1.8−1.5−1.2
Lower real interest rate (=3.5 percent)−3.1−3.4−3.0−1.6−0.20.8−0.3−0.5−0.4−0.2
Long run real Growth Rate
Higher long run real growth rate (=3.5 percent)−3.7−3.9−3.4−1.80.1−1.0−1.1−0.9−0.5
Lower long run real growth rate (=2.5 percent)−3.7−3.9−3.4−2.0−0.60.0−1.0−1.2−1.0−1.0
Baseline Parameters−4.3−4.3−3.8−2.1−0.5−1.6−2.5−2.6−2.2−1.8
Sensitivity Tests
Oil Reference Price
Higher Oil prices (30 percent higher each year than baseline starting in 2009)−7.8−7.3−6.3−3.7−1.2−3.8−4.5−4.4−3.9−3.2
Lower Oil prices (30 percent lower each year than baseline starting in 2009)−0.5−1.2−1.1−0.40.30.7−0.4−0.6−0.5−0.3
Oil Production
Higher Oil Production (30 percent higher each year than baseline starting in 2009)−7.0−6.6−5.7−3.3−1.1−3.3−4.0−4.0−3.5−2.9
Lower Oil Produciton (30 percent lower each year than baseline starting in 2009)−1.5−2.1−1.9−0.90.10.0−1.0−1.1−1.0−0.7
Tax Take
Higher Tax Take (10 percent higher each year than baseline starting 2014)−4.5−4.6−4.0−2.2−0.5−1.8−2.6−2.7−2.3𢄡1.9
Lower Tax Take (10 percent lower each year than baseline starting 2014)−4.0−4.1−3.6−2.0−0.4−1.5−2.3−2.4−2.1−1.7
Non oil Revenues
Higher Non Oil Revenues (10 percent higher each year than baseline starting 2009)−4.3−4.3−3.8−2.1−0.5−1.6−2.5−2.6−2.2−1.8
Higher Non Oil Revenues (10 percent lower each year than baseline starting 2009)−4.3−4.3−3.8−2.1−0.5−1.6−2.5−2.6−2.2−1.8
Real Interest Rate
Higher real interest rate (=4.5 percent)−4.7−4.7−4.1−2.3−0.6−2.1−2.9−2.9−2.5−2.1
Lower real interest rate (=3.5 percent)−3.8−4.0−3.5−1.9−0.4−1.2−2.1−2.2−1.9−1.5
Long run real Growth Rate
Higher long run real growth rate (=3.5 percent)−4.3−4.3−3.8−2.0−0.3−1.6−2.5−2.5−2.1−1.4
Lower long run real growth rate (=2.5 percent)−4.3−4.3−3.8−2.2−0.7−1.6−2.5−2.6−2.4−2.3
Source: Fund Staff Estimates
Source: Fund Staff Estimates

The Carcillo et al (2007) government primary spending formula as a share of non-oil GDP (gt)

where φt=α(1+γt)Rt, α is the habit persistence parameter, γt+i is the real growth rate of non oil GDP in year t+i, Rt+i is one plus the real interest rate in year t+i, γ is the long term real growth rate of non oil GDP, zt+j is the level of oil revenue as a share of non oil GDP for year t+j, τt+j is the level of non oil revenues for year t+j and bt − 1 is the level of debt as a share of non oil GDP in the previous period for i=1,...j and j=1,...., infinity.

This equation shows that, with habit formation, the government primary spending is a linear combination of the government primary spending level in the previous period and the permanent income level as indicated in the text. The implication is that if the previous period’s NOPD is higher than the permanently sustainable level, then the NOPD is expected to adjust to the permanently sustainable level over a number of periods. In contrast, without habits (α = 0), the policy would be to adjust the NOPD abruptly to the permanently sustainable level in a single period.

References

    BallasoneFabrizioTakizawaHajime and ZebregsHarm2006Managing Russia’s Oil Wealth: An Assessment of Sustainable Expenditure PathsIMF Country Report No. 06/325 pp 343.

    BaunsgaardThomas2003Fiscal Policy in Nigeria: Any Role for Rules?IMF Working Paper No. 03/155 (Washington: International Monetary Fund).

    BarnettSteven. and OssowskiRolando.2003Operational Aspects of Fiscal Policy in Oil Producing Countries” inJ.DavisR. Ossowski and A. Fedelino2003Fiscal Policy Formulation and Implementation in Oil-Producing Countries (Washington: International Monetary Fund).

    BasdevantOlivier2006Are Diamonds Forever? Using a Lifecycle Approach to Analyze Botswana’s Reliance on Diamond RevenuesIMF Country Report No. 06/228 pp 419.

    ClausenJens2007Calculating Fiscal Sustainability Benchmarks for Botswana”IMF Country Report No. 07/375 pp 211.

    CarcilloStephaneLeighDaniel and VillafuerteMauricio2007Catch-Up Growth, Habits, Oil Depletion, and Fiscal Policy: Lessons from the Republic of CongoIMF Working Paper No. 07/80 (Washington: International Monetary Fund).

    GylfansonThorvaldur2001Lessons from the Dutch Disease:Causes, Treatment, and CuresInstitute of Economic Studies Working Paper Series.

    de Carvalho FilhoIrineu2006Managing Oil Wealth”IMF Country Report No. 06/340 pp. 214

    KatzM. U. BartschH. Malothra and CueM.2004Lifting the Oil Curse—Improving Petroleum Revenue Management in Sub-Saharan Africa” (Washington: International Monetary Fund).

    LeighDaniel and Jan-PeterOlters2006Natural-Resource Depletion, Habit Formation, and Sustainable Fiscal Policy: Lessons from GabonIMF Working Paper No. 06/193 (Washington: International Monetary Fund).

    Sala-i-MartinXavier and ArvindSubramanian2003Addressing the Natural Resource Curse: An Illustration from Nigeria”IMF Working Paper No. 03/139 (Washington: International Monetary Fund).

    SeguraAlonso2006Management of Oil Wealth Under the Permanent Income Hypothesis: The Case of São Tomé and PríncipeIMF Working Paper No. 06/183 (Washington: International Monetary Fund).

    TakizawaHajime and others2004Are Developing Countries Better Off Spending Their Oil Wealth Upfront?IMF Working Paper No. 04/14 (Washington: International Monetary Fund).

    OssowskiRolando. and others2008Managing the Oil Revenue Boom: The Role of Fiscal InstitutionsIMF Occasional Paper No. 260 (Washington: International Monetary Fund).

The NOPD is defined consistently with the staff report as total revenue excluding grants, oil revenue and interest on government assets, minus total expenditure excluding interest payments and foreign financed investment. Following Barnett and Ossowski (2003), this paper focuses on the NOPD rather than the overall balance because it eliminates fluctuations due to reasons unrelated to the fiscal stance and fiscal sustainability (such as volatile oil prices) and makes clear that fiscal proceeds from oil should be viewed as financing. Recent movements in the overall balance which switched from a deficit of about 16 percent non oil GDP in 2003 to a surplus of about 2 percent in 2007 mainly due to higher oil prices and production illustrates this point clearly.

Non-oil revenues remained roughly at their historical levels between 8 and 9 percent of non-oil GDP until 2006. The increase occurred mainly in 2007 reflecting buoyancy in income taxes related to significant wage increases, the collection of income tax arrears from previous years, and improved collection from property taxes.

Exceptional security spending captures military spending generally not expected at the time of the budget spent in response to unexpected security threats and includes purchases of military equipment.

Reliable and comprehensive domestic debt information prior to the 2005 Mazars report is not available. Debt ratios as a share of GDP are as follows: 46 percent of GDP in 2003 and 24 percent in 2007 for external debt, 5 percent of GDP in 2005 and 3 percent of GDP in 2007.

The definition of the overall balance data presented above excludes grants and is on a cash basis consistent with the staff report presentation.

Since execution data was not available for part of the sample for 2007, data until year 2006 is used to maximize the size of the sample. Year 2000 is the year when oil prices started their escalation and is therefore used as the beginning year of the oil boom.

The sample includes the following 28 oil producing countries: Algeria, Angola, Azerbaijan, Bahrain, Brunei, Cameroon, Congo, Ecuador, Gabon, Indonesia, Iran, Kazakhstan, Kuwait, Mexico, Nigeria, Norway, Oman, Qatar, Russia, Saudi Arabia, Sudan, Syria, Timor Leste, Trinidad and Tobago, United Arab Emirates, Venezuela, Vietnam, and Yemen.

See the following papers for a more detailed discussion of the developments in the mentioned countries: Carcillo et. al. (2007) for the Republic of Congo, Leigh and Olters (2006) for Gabon, Irineu de Carvalho Filho (2005) for Trinidad and Tobago and Sala-i-Martin and Subramanian (2003) for Nigeria.

Net assets of the government includes oil wealth, or the present discounted value of future net oil revenues.

Note however that this is strictly true if a country is not a net debtor and the horizon is not finite. If the horizon is finite and a country starts applying the rule as a net debtor, a balanced budget rule would not satisfy the intertemporal budget constraint since at some point surpluses would need to be run to pay down the debt principal.

For example, it is clear that a balanced budget rule would not be supported by the permanent income approach since it would imply large procyclical responses of government spending to changes in oil prices which would not be optimal under standard welfare functions.

There are many other oil and natural resource rich countries more generally for which the permanent income hypothesis approach has been applied. These include for example: Sao Tome and Principe (Segura (2006)), Botswana (Basdevant (2006), Clausen (2007)), Russia (Balassone, Takizawa and Zegrebs (2006)), Nigeria (Baunsgaard (2003)), Trinidad and Tobago (Irineu de Carvalho Filho (2006))

Only these four rules are discussed in the main text. However, estimates were also prepared for four other rules which are derived from traditional one-period adjustment PIH approaches a la Barnett and Ossowski (2003). They differ from the PIH rules in the text mainly in terms of the underlying welfare function, which leads to different measures of government consumption being kept constant over time, i.e. real aggregate consumption and real per capita consumption. For each of these two other measures, the NOPDs are computed for versions where wealth is consumed in an infinite or in a finite but long period of time (63 years). The results for these rules are presented in the Appendix Tables 1 and 2.

The finite number of years assumed in the calculation is 63 years. See Clausen (2007) for additional discussion.

The definition of wealth in the model includes oil wealth, net financial assets and the present value of non-oil revenues.

See Carcillo et al. (2007) for a detailed derivation of the weighted average formula for the government primary consumption which is also shown in the appendix to illustrate the nature of the calculations. The assumption that there is no productive investment by the government is a limitation of the model but it still provides a useful benchmark since it is consistent with the limited evidence showing any significant relationship between public investment and growth.

The annuity calculation was carried out for all PIH rules estimates included in this paper, with the exception of Carcillo et al. (2007) where the calculations for government consumption are more complicated and follow the formula included in the appendix.

The annuity on oil wealth can be calculated in several ways depending on the rule. One is to calculate a government consumption annuity which is constant as a share of non-oil GDP as in the first PIH rule discussed in the text. Other alternatives are government consumption annuities which are constant in real terms or constant in real per capita terms for a finite period of time or for an infinite time period. These alternatives are the ones used to compute the results for the other PIH rules which are presented in the appendix.

All macroeconomic variables and oil sector variables determining the level of oil revenues are the same as in the baseline scenario, with the exception of the non-oil real GDP growth rate and the real interest rate. For the purpose of the estimations of the long-term NOPDs, the assumptions for the latter variables had to be adjusted to avoid a situation where the real interest rate is permanently below the real non-oil GDP growth rate, as this would trivially solve any sustainability problem. To address this problem and avoid forcing an assumption of an unrealistically high real interest rate, the real non-oil growth rate after 2012 is gradually adjusted for it to converge to a long term level of 3 percent by 2032 assuming that there is a long catching-up period in which growth rates will remain above this lower long term rate. The real interest rate is assumed to be flat at 4 percent during the whole period, a level consistent with the assumption in Carcillo et al. (2007) for the Republic of Congo. While in Chad the large majority of public debt and assets earn relatively low real interest rates given concessionality and investment constraints imposed by the BEAC, it would be unrealistic to assume these rate levels for the purpose of this exercise since both features are likely to change in the next 20 to 30 years.

This includes the four additional rules discussed in footnote 13. The results are presented in Appendix Table 1.

This result also holds for the four additional rules mentioned in footnote 13 whose results are discussed in the Appendix. See Appendix Table 2 for details.

Other Resources Citing This Publication