Selected Issues and Statistical Appendix

This Selected Issues paper and Statistical Appendix presents a debt sustainability analysis for Afghanistan. The analysis is based on a simple spreadsheet model that calculates indicators of sustainability in terms of ratios of the net present value (NPV) of debt and debt service to relevant macroeconomic aggregates over 2005–30. Five scenarios are examined in the paper. Key vulnerability ratios are assessed under each scenario and compared with sustainability thresholds under the enhanced Heavily Indebted Poor Countries (HIPC) Initiative. The results obtained are also discussed in detail in the paper.


This Selected Issues paper and Statistical Appendix presents a debt sustainability analysis for Afghanistan. The analysis is based on a simple spreadsheet model that calculates indicators of sustainability in terms of ratios of the net present value (NPV) of debt and debt service to relevant macroeconomic aggregates over 2005–30. Five scenarios are examined in the paper. Key vulnerability ratios are assessed under each scenario and compared with sustainability thresholds under the enhanced Heavily Indebted Poor Countries (HIPC) Initiative. The results obtained are also discussed in detail in the paper.

I. Debt Sustainability Analysis1

A. Introduction and Summary

1. Undertaking a debt sustainability analysis (DSA) for Afghanistan presents unique challenges. The lack of complete debt information, the presence of disputed or unverified claims, and the almost total absence of historical data on key macroeconomic variables precludes use of the Fund’s standard template for assessing external sustainability.2 Even the more nuanced approach to debt sustainability for low income countries presents difficulties, given that standard assessments of institutional capacity (for example, through the World Bank’s Country Policy and Institutional Assessments—CPIA) have yet to be completed for Afghanistan.

2. Data and other constraints notwithstanding, the need for an assessment of Afghanistan’s debt sustainability over the medium‐ and long-term remains. Particularly in the post-election environment, as ministries and other government institutions come to grips with their respective portfolios, a quantitative assessment of how the debt burden might evolve under different scenarios could be an important input in developing a debt management strategy. In this context, the current DSA focuses less on issues pertaining to pre-existing debts and claims (a central concern of a previous DSA), and more on the impact of future borrowing and the behavior of key vulnerability indicators under different scenarios. It takes as a baseline assumption a complete write-off of all pre-existing bilateral claims.3 In an effort to address some concerns specific to Afghanistan, the current exercise also departs slightly from the standard template by examining the impact of different borrowing strategies.

3. Five scenarios are examined in the current exercise: 1) a baseline scenario, drawing on the current macroeconomic framework and a nominal level of borrowing over the long term; 2) a more pessimistic scenario, with slower projected growth in GDP, exports, and government revenue; 3) the more pessimistic scenario with the addition of a sudden exchange rate depreciation; 4) the pessimistic scenario with a lower degree of concessionality in external loans; and 5) the preceding scenario under a more aggressive external borrowing strategy. Key vulnerability ratios are then assessed under each scenario and compared with sustainability thresholds under the HIPC Initiative (Table I.1 and Figures I.1I.5).

Table I.1.

Islamic State of Afghanistan: Debt Sustainability Analysis, 2005–30 1/

(In percent; unless otherwise indicated)

article image

All debt indicators refer to public and publicly guaranteed debt. The NPV of debt is calculated using DSA discount rates based on January-June 2004 SDR Commercial Interest Reference Rates (CIRRs).

The NPV of debt ratio is based on a 3—year average of exports of goods and services; the debt-service ratio is based on current year figures.

Revenues are defined as general government revenues, excluding grants.

Figure I.1.
Figure I.1.

Islamic State of Afghanistan: NPV of Debt to Exports, 2004–30

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Figure I.2.
Figure I.2.

Islamic State of Afghanistan: NPV of Debt to Revenue, 2004–30

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Figure I.3.
Figure I.3.

Islamic State of Afghanistan: NPV of Debt to GDP, 2004–30

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Figure I.4.
Figure I.4.

Islamic State of Afghanistan: Debt Service to Exports, 2004–30

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Figure I.5.
Figure I.5.

Islamic State of Afghanistan: Debt Service to Revenue, 2004–30

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

4. The results of the various scenarios indicate that, assuming a “clean slate” with regard to old bilateral claims, and reasonably strong economic and revenue performance, there is a comfortable margin with regard to key sustainability thresholds. Through the successive shocks outlined in Scenarios 1–5, however, this margin quickly diminishes, leading to situations where Afghanistan’s debt and debt service burden would be deemed unsustainable. Throughout all the projections, however, it is critical to keep in mind that the projected sustainability margin may be substantially less than shown in the baseline, depending on how bilateral claims are resolved, and given the strong potential for domestic and external shocks.

B. Methodology

5. Data constraints present challenges to undertaking a meaningful debt sustainability analysis.4 The destruction of debt records and documentation, for example, means that a substantial portion of Afghanistan’s debt is yet to be verified and reconciled with creditors. While the Afghan authorities have made substantial progress in this area over the past year (aiming at a full reconciliation by end-March 2005), information is still lacking for a number of bilateral creditors. In addition, the status of certain claims—most notably those of Russia, which are disputed by Afghanistan—have yet to be settled. The weak statistical base underlying key macroeconomic variables, such as GDP and selected balance of payments indicators, is also a concern, given that current assessments of “sustainability” rely principally on ratios based on these figures. Finally, the large structural shifts that are emerging as Afghanistan moves ahead with macroeconomic stabilization and restructuring cast into doubt the reliability of virtually any set of projections, and certainly those stretching over 25 years.

6. While the constraints and limitations noted above call for some caution in interpreting results, a revised DSA for Afghanistan—building on the previous exercise completed in April 2003—is both timely and warranted. The current exercise is designed to be forward looking. The Afghan authorities, while relying primarily on grants to finance reconstruction and development, have nevertheless already begun to contract new program and project loans from multilateral and bilateral creditors. As development capacity increases, and in the face of daunting social and economic challenges, further recourse to external loans is likely. In this context, it is vital to highlight the potential vulnerabilities that could arise from undertaking such obligations, as an input to developing an explicit and prudent external debt management strategy that could guide the government in future years. Of particular concern is to avoid the same pitfall seen in many other low-income countries, where the rapid accumulation of external loans—even on concessional terms—led to an unsustainable debt burden, and subsequent recourse to rescheduling or other forms of debt relief. 5

7. The current debt sustainability analysis is based on a simple spreadsheet model that calculates indicators of sustainability in terms of ratios of the NPV of debt and debt service to relevant macroeconomic aggregates over the period 2005–30. The analysis is based on assumptions for (a) treatment of existing claims; (b) GDP growth, exports, and fiscal revenues; and (c) the level and terms of future loans from multilateral and bilateral creditors. The baseline scenario for the model is taken, to the extent possible, from the macroeconomic framework developed for the current staff-monitored program (SMP). A set of assumptions is then used for the behavior of key macroeconomic aggregates, to extend the exercise beyond the medium-term currently covered by the SMP framework.

Existing claims

8. A critical assumption in the exercise relates to the treatment of all bilateral claims. As noted earlier, the authorities are currently engaged in a comprehensive effort to identify, verify, and reconcile official claims vis-a-vis bilateral creditors. Staff estimate these claims (excluding Russia) to be in the range of $294 million, of which $124 million has been verified. The true value of the remainder has a considerable margin for error, given the lack of physical records and surrounding uncertainty regarding outstanding obligations and accumulated arrears. The current DSA exercise assumes that all bilateral debt will be forgiven, including the Soviet-era Russian claims on the order of $10.8 billion—equivalent to about 190 percent of 2004/05 estimated GDP.

9. Total forgiveness of these bilateral claims is a working assumption that simplifies the DSA exercise, allowing a focus on the implications of debt and debt service connected with recent and future borrowing. This assumption may be viewed as heroic, given recent information that some bilateral creditors may insist on some form of rescheduling rather than debt forgiveness, and while the position of a range of other bilateral creditors has not yet been determined. Consequently, when assessing the figures presented in this DSA, and evaluating Afghanistan’s external vulnerability under different scenarios, it is critical to remember the existence of these potential liabilities, and the downside risk that they represent.

Macroeconomic assumptions

10. A number of alternative scenarios are used to illustrate the implications of different macroeconomic outcomes for external sustainability. As noted above, the DSA draws on the current macroeconomic framework under the SMP to generate a baseline scenario. This baseline (Scenario 1) assumes: (a) average real GDP growth of 9.5 percent during the medium term (2005–2009), and annual growth of 5 percent thereafter; (b) growth in exports of 17 percent over the medium term, and 7 percent thereafter (roughly in line with nominal GDP); and (c) an annual increase in government revenue equivalent to 1 percent of GDP, to a maximum level of 15.5 percent (achieved by 2016).6, 7 Alternative scenarios include the following changes to key variables:

  • Scenario 2 examines the impact of a substantial reduction in GDP and export growth, along with a slower growth rate for government revenue. Under this scenario, real GDP is projected to grow by an average of 4.7 percent over the medium term, and 2.5 percent per year thereafter. Exports are subjected to a shock of similar magnitude, growing by an average 9 percent during 2005-09, and 4.8 percent thereafter. Government revenue, meanwhile, increases by only one-half percent of GDP per year, until reaching 12 percent in 2019.8

  • Scenario 3 combines the more pessimistic outlook for macroeconomic and revenue figures with a one-time nominal depreciation of the Afghani. Under this scenario, slower GDP, export, and revenue growth are accompanied by a 30 percent depreciation of the Afghani against major currencies. While this represents a substantial shift, it is not out of line with recent experience. Between March and August 2004, for example, the Afghani appreciated by 16 percent, driven in part to sizeable inflows of donor assistance. A temperance of these or other capital flows, or some sort of reversal could engender a substantial depreciation.

Level and terms of new debt

11. Scenarios 1–3 hold as a critical assumption that the government continues its prudent approach to external borrowing. Annual disbursements of external loans are moderate at about $170 million per year, and the terms of these loans are assumed to be highly concessional, with a grant element of at least 60 percent (in line with the performance criteria for contracting of external debt set out in the staff monitored program). The level of new loans, and the terms under which they are disbursed are a key element in assessing sustainability.9 These assumptions center around continued strong donor support in the form of grants, and a willingness to offer loans on terms that will help to ensure medium term external and fiscal sustainability. While such an outcome is not beyond possibility, it is far from guaranteed.

  • Scenario 4 assesses the impact of a lower degree of concessionality in external borrowing. The performance criteria contained in the SMP (a minimum grant element of 60 percent for all newly contracted external loans) represents a strict standard—considerably higher than the 35 percent level more often found in Fund programs—but one deemed appropriate for Afghanistan. Scenario 4 illustrates the consequences of new external borrowing at rates consistent with a 35 percent grant element, and under less optimistic macroeconomic outcomes (consistent with the conditions in Scenario 3).

  • Scenario 5 builds on Scenario 4, by considering an alternate borrowing strategy. Specifically, it is assumed that the authorities pursue a more aggressive borrowing strategy, based on a given path for GDP and other key variables consistent with Scenario 1, and with a view to allowing debt to accumulate at a rate consistent with a stock equivalent to 35 percent of GDP by 2015. The level of borrowing would then decrease, so as to allow the debt/GDP ratio to fall below 20 percent after 2030. Such a strategy would be consistent with a shortfall in grants, or a decision by the government to finance a range of additional investment projects through loans, with the notion that a 35 percent debt/GDP ratio is an acceptable and sustainable burden. Scenario 5 tests the impact of implementing such a borrowing strategy under the less favorable macroeconomic, revenue, and concessionality assumptions of Scenario 4.

C. Debt Sustainability

12. Five indicators are used to assess the sustainability of Afghanistan’s external debt: (a) NPV of debt as a percentage of exports (NPV/X); (b) NPV of debt as a percentage of fiscal revenues (NPV/R); (c) NPV of debt as a percentage of GDP (NPV/GDP); (d) debt service as a percentage of exports (DS/X); and (e) debt service as a percentage of revenues (DS/R). Under the enhanced HIPC Initiative, a country’s external debt is considered unsustainable if the NPV/X ratio exceeds 150 percent; in addition, the DS/X ratio should normally be within the range of 15–20 percent or less. For very open economies (with an exports-to-GDP ratio of at least 30 percent), which have a heavy fiscal debt burden despite strong efforts to generate revenue (with a fiscal revenue-to-GDP ratio of at least 15 percent), the debt sustainability threshold is a NPV of debt-to-revenue ratio of 250 percent.

13. These ratios provide useful benchmarks for assessing sustainability. However, in line with recent thinking on debt sustainability for low income countries,10 Afghanistan’s particular context should also be kept in mind. GDP, for example, depends to a large degree on agricultural output, which has already shown to be highly volatile. The exchange rate, while relatively stable since the introduction of the new Afghani, may also be subject to bouts of instability in the face of underlying shifts in demand, uneven donor flows, and the eradication of the opium economy. Further, institutional capacity is still weak, and the quality of policy-making (while strong thus far) is uncertain over the medium term.11 These factors suggest that the sustainability thresholds, and consequently the margin of safety, might be considerably lower than shown here. Special attention should also be given to ratios of debt and debt service relative to government revenue. Given the very low level of government revenues—especially compared to expenditure needs—the proportion spent on debt service becomes a more critical concern.

14. Under the first scenario (forgiveness of all existing bilateral debt, strong GDP and export growth, and limited borrowing on very concessional terms), Afghanistan’s debt burden falls within a sustainable range throughout the projection period. The NPV/X debt ratio rises close to 70 percent during 2015–17, but declines thereafter to 48 percent by 2030. Debt service would be limited to below 3 percent of exports for all of the projection period, while the ratio of debt service to government revenues remains in the 2.5 percent range in the short term before declining to below 2 percent for the remainder of the period.

15. Scenario 2 highlights the importance of assumptions regarding Afghanistan’s macroeconomic and fiscal outlook. Under this second scenario (substantially weaker GDP and export growth, along with a more gradual path for revenue growth), Afghanistan’s debt burden continues to fall within the sustainable range during the projection period, but comes substantially closer to sustainability thresholds. Debt relative to exports rises above 100 percent by 2015, peaking at around 110 percent by 2021 before falling to 102 percent by 2030. Debt service as a share of government revenue continues to remain modest, and well below the sustainability threshold throughout the period. However, the less robust growth of the economy equates to lower revenue earnings versus the baseline scenario. As a result, both DS/R and DS/X, after falling in the short run, rise steadily over the medium and long-term.

16. In Scenario 3, the additional impact of an exchange rate shock illustrates another potential vulnerability. For purposes of the simulation, exports are assumed to be priced in Afghani—allowing for a full pass-through of the appreciation on export values.12 The relative fall in the value of exports is enough to push the NPV/X ratio to 143 percent by 2021, falling to 132 percent by 2030. Similar increases are apparent in debt service/exports and debt service/revenue ratios. However, reflecting the high level of concessionality in external borrowing, these ratios remain in the sustainable range throughout.

17. Scenario 4 evaluates the impact of borrowing on less concessional terms. Specifically, external loans are assumed to have a grant element of only 35 percent—the minimum level for loans on “concessional” terms.13 Applied to the less optimistic macroeconomic conditions outlined in Scenario 2, the NPV/X ratio breaches the 150 percent sustainability level by 2012, and rises as high as 187 percent by 2021. Debt and debt service relative to government revenue also raise some concerns. NPV of debt to revenue rises above 200 percent by 2009, and stays above this level until 2017, while debt service as a share of revenue also climbs considerably—with total debt service consuming 10–11 percent of government revenue after 2014. The latter represents a sizeable burden for a government faced with the daunting challenges of reconstruction and pressing social needs.

18. Scenario 5 illustrates the potential risks associated with a less cautious approach to external borrowing. While such an approach might be sustainable under the favorable conditions outlined in Scenario 1 (the NPV/X ratio quickly breaches the sustainability threshold, but debt service as a share of exports remains below 12 percent throughout the period), a less optimistic macroeconomic picture raises serious concerns. Under the assumptions presented in Scenario 4, the NPV/X ratio reaches a level more than double that considered sustainable by 2012, and climbs as high as 423 percent by 2021, while debt service relative to exports hits 15 percent by 2014 and rises above 20 percent by 2018—staying above this level for the remainder of the period. Further, debt service as a share of government revenue quickly rises to substantial levels, moving beyond 20 percent by 2013 and peaking at 28 percent by 2026.

D. Policy Implications

19. The scenarios described above highlight a number of important considerations for Afghanistan with respect to external debt. The following are among the most critical points for consideration by the government and bilateral creditors/donors.

20. Relief for existing, pre-war claims is essential. As highlighted in the DSA conducted in 2003, debt sustainability would be difficult to attain without the forgiveness of most existing bilateral debt. Indeed, the current analysis takes this scenario as a given that these claims are not restructured, but written off in their entirety. Failure to achieve this “clean slate” would imply a higher base for debt and debt service for all scenarios, which could make external and fiscal sustainability difficult even under the most optimistic scenario, depending on the terms.14

21. The terms and conditions surrounding new borrowing must remain highly concessional. Scenario 4 highlights the potential fragility of external and fiscal sustainability should Afghanistan undertake external borrowing on less generous terms. This will be a key concern for the government moving forward, given the potential for donor support (in the form of grants) to fall off over time, and the pressures to find alternative financing for key development projects. The temptation to accept loans offered by bilateral or even multilateral institutions with less than the most generous level of concessionality should be avoided, where possible.

22. The underlying premises behind a national debt strategy should take into account the low level of foreign exchange receipts derived from the current account, as well as the extremely low starting point of government revenue.15 Borrowing strategies that focus on different aggregates, such as debt to GDP, while ignoring key solvency ratios could potentially put Afghanistan’s fiscal and external vulnerability at risk-and quite rapidly depending on macroeconomic outcomes. This is a risk highlighted by the experience of a number of transition economies, where borrowing strategies based on projected macroeconomic outcomes led to sustainability problems when growth and export opportunities did not measure up to expectations.

23. Layered upon these fundamental policy principles must be the knowledge that Afghanistan faces a daunting array of challenges. An unstable security situation, the presence of a large and illicit opium economy, the reemergence of drought, and government institutions which are largely still in the making imply the potential for considerable volatility in coming years. Indeed, shocks of one sort or another seem almost inevitable given the many downside risks. Continued mobilization of grant financing, caution with regard to taking on new external liabilities, and ensuring that they are tied to only the most pressing social or development needs will be crucial.

II. Medium-Term Structural Fiscal Reforms16

A. Introduction

24. This note describes the medium-term agenda for fiscal structural reforms in Afghanistan. The task of rebuilding this fragmented nation and responding to the extremely high social needs is complicated by the ongoing high level of insecurity, the significant opium production, the dependency on donor financing and weak institutions. In order to promote fiscal sustainability, the reform agenda should be anchored around three key themes, namely:

  • Continuing to consolidate fiscal management by further incorporating the fiscal operations in the provinces and by integrating the considerable donor assistance into the budget. The 34 ‘provinces’ referred to in this note are not a separate sub-national tier of government but constitute a deconcentrated level of the central government, although in the recent past many have exercised significant independence;

  • Rebuilding and broadening the domestic tax base so as to be able within nine years to fund operating expenditures entirely by domestic revenue; and

  • Developing effective public administration and public expenditure management (PEM) systems with the capacity to plan and implement a more comprehensive national budget.

25. This agenda encompasses an extremely broad range of measures, with many, on the revenue side, in public administration and in PEM, already underway. These measures partly build on the work of several FAD technical assistance missions, whose strategies and work plans were largely endorsed by the authorities and donor community. They are consistent with the three-step process commonly applied in post-conflict countries, which aims to rebuild fiscal institutions by: (a) creating a legal and/or regulatory framework for fiscal management; (b) establishing and/or strengthening the fiscal authority; and (c) designing appropriate revenue and expenditure policies while simultaneously strengthening revenue administration and PEM.17 Fundamental expenditure policy reforms are likely to take longer to develop, and implementation will rely on the efficacy of the underlying expenditure management systems and the capacity of the public administration.

26. Section B of this note outlines the steps being taken to consolidate the fiscal framework. Section C briefly discusses the medium-term macro-fiscal outlook. Sections C and D describe the steps being taken to rebuild and broaden the domestic tax base and to improve public expenditure policy and management.

B. Consolidating the Fiscal Framework

27. More than 20 years of war and internal conflicts massively eroded the central government’s control over fiscal operations. Although many administrative institutions remained formally in place, along with some well-established laws (e.g., budget, tax, and audit) and procedures (e.g. a public accounting manual), most had ceased to function due to a lack of resources and qualified staff. While the PEM system is de jure highly centralized, central government control over provincial fiscal operations was very weak, resulting in the only partial transfer of revenues to the center and the lack of coordination of expenditures. The task of consolidating fiscal operations was daunting as the basic PEM and tax systems were not operational in practice and needed to be updated and improved. The revenue situation was exacerbated by the proliferation of levies and taxes (many with no legal basis) imposed by different government agencies and local authorities that resulted in a complex tax system, and the absence of a functioning payments system to facilitate tax payments.

28. The process of developing a comprehensive budget that incorporates the provinces and donors, can be divided into the following three stages: the first stage, which has largely been completed, is the reestablishment of basic control over domestic revenues and those expenditures managed by the central government, in addition to providing a framework for coordinating the relatively massive amount of donor assistance. The second stage, which reflects the government’s current status, is the gradual deepening of the reforms to consolidate the management of domestic and donor funded fiscal operations. The final stage entails a single, comprehensive, transparent and sustainable budget, where spending patterns reflect government priorities. The broad elements of the three stages are outlined below.

Stage one: controlling domestic operations and coordinating donor spending

  • The government progressively reestablished control over the Operating Budget, which comprises domestic revenues, salaries and wages of civil servants at the center and in the provinces and a small amount of maintenance and capital expenditure. This involved an extensive effort to build the basic fiscal institutions, including establishing the physical infrastructures and developing administrative capacity (see section below).

  • The coordination of sizeable donor-funded projects was important to ensure that spending patterns reflected government priorities. In April 2002, the government adopted the National Development Framework (NDF), which sets out a domestically owned development strategy, under which donors could implement projects. The NDF has three core pillars: (a) sustainable human capital—humanitarian assistance and social policy; (b) physical infrastructure and natural resources–laying the basis for a private sector-led growth strategy; and (c) the creation of sustainable growth through the private sector. The NDF outlined 16 national programs (NPs) across these pillars, providing a framework for the establishment of a national development budget (NDB).18 The government also developed a Consultative Group mechanism, whereby lead ministries and donors manage the NPs.

  • An explicit mechanism to record and monitor donor assistance was also established at an early stage. The Ministry of Finance (MoF) developed a Donor Assistance Database (DAD) to record information about donor pledges, commitments and disbursements.19

  • The operating budget is funded by domestic revenues and multi-donor grants. The latter are channeled though the multi-donor-administered Afghanistan Reconstruction Trust Fund (ARTF) and the smaller UN-administered Law and Order Trust Fund of Afghanistan (LOTFA).20 While contributing substantially to the financing of the budget, the ARTF also provides an important mechanism for improving compliance with basic fiduciary standards and promoting the transparency and accountability of reconstruction assistance. The ARTF is a unique external funding mechanism that supports the development and use of the government’s own PEM systems, and avoids creating parallel systems within government that could divert limited capacity. The government initially spends money through its own PEM systems and an independent monitoring agent reimburses funds after checking that the expenditures are eligible (i.e. not for security) and have been made in compliance with basic financial management procedures. The monitoring agent provides an important fiduciary safeguard, both for donors and for the government as the domestic internal and external audit functions are still in the process of being established.21

Stage two (current): consolidating the management of domestic and donor spending

  • The Core Budget concept was introduced in June 2004 (for fiscal year 2004/05) to combine the operating budget with development expenditures that are channeled through the treasury’s accounts. The Core Budget is considered to be the portion of the National Budget that is controlled by the Government and for which it is accountable. However, many donors still earmark their funds for particular development projects and require separate government bank accounts (controlled by the Treasury, but separate from the treasury single account, TSA, in 2004), limiting thereby the extent to which resources may be transferred within the core budget (Box 1).

  • The government also collects and disseminates information on the development expenditures that donor agencies plan to implement directly. This External Budget is monitored by the MoF and reported in the DAD (commitments, rather than pledges, were reported in the 2004/05 budget). Table II.1 reports the relative shares of operating expenditures, core budget development expenditures and committed external budget expenditures for 2001/02 to 2004/05.

Table II.1.

Budget, 2001/02–2004/05

(In millions of U.S. dollars)

article image
Source: Ministry of Finance. October 2004 Financial Report.

Core Development Budget expenditure is disbursed through the government’s treasury Accounts.

Reflects commitments by donors for specific projects undertaken directly by donors but reported to MOF. (the budget documents showed a higher figure due to some partially unfunded programs)

  • The NDF continues to be refined to provide the framework for coordinating donor spending. For the 2004/05 budget, all spending under the NDF was subdivided into NPs and, a new category, National Priority Programs (NPPs). These programs are designed to coordinate donors around government policies and larger, sometimes sector-based, programs. In theses initial stages of development, these programs largely remain a collection of donor-funded projects, which can present a complex and confusing management task. By having the overall coordination framework of the NDF, it is hoped that implementation will be more effective as competition and duplication are reduced. These programs also introduce the concept of a budget constraint (and expenditure choices), which is initially driven by the availability of donor financing and an assessment of absorptive capacity.

  • The core budget aims to gradually link operating and development expenditures. Starting with the formulation of the FY 2003/04 core budget, the operating budget was expected to include operating and maintenance expenditures for completed development projects and sufficient resources to support the implementation of ongoing and new development projects.

  • As Figure II.1 and Table II.1 indicate, a substantial amount of donor spending remains outside the core budget. While this spending is included in budget documents, it is essentially ‘off-budget’ for accounting purposes. This may undermine the coherence of the government’s policy framework, where the proliferation of projects may be driven by bilateral interests, and lead to duplication and other forms of waste. The high share of direct donor spending in the overall fiscal operations also creates a significant ‘parallel’ public service, whereby some basic services are predominantly delivered by donor projects and NGOs.

Figure II.1.
Figure II.1.

Budget Spending, 2004/05

(in percent of GDP)

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Stage three: Expanding the Core Budget and simplifying budget management

  • The existence of multiple donor budgets and earmarked sources of funds considerably complicates fiscal management (see also Box 1). The earmarking of funds for specific expenditures reduces the flexibility of the budget to respond to changes in the macroeconomic environment or policy shifts, and may inadvertently skew allocations towards lower priority areas. Over the long term the development budget should be transformed from an aggregation of donor-funded projects into a comprehensive government-led demand-driven expenditure program. The distinction between operating and development expenditures should also disappear.22

  • As the government’s financial management capabilities and planning capacity improve, donor development assistance is expected to be increasingly channeled through the Core Budget. However, in order to achieve this, the authorities recognize the need to demonstrate their ability to manage resources in an efficient, accountable and transparent manner to give donors the requisite fiduciary assurances. Commensurately, donors must also be willing to progressively support the development of government PEM systems by channeling their funds through the budget.23

  • More fundamental sector-based financial planning within a medium-term budget framework must be pursued to develop a politically viable, stable, efficient, and responsive center-provincial fiscal system. The authorities must also develop strategies to better integrate some of the expenditures currently being implemented directly by donors. Donor programs have created a significant ‘parallel’ public service that currently delivers a substantial amount of critical services to the population. For example, in the health sector the government has largely adopted a purchaser-provider model, where implementation is largely carried out by NGOs or other donor assisted programs. The central government is focusing on its stewardship role, which involves developing and enforcing a regulator framework, setting standards etc. as well as some limited direct provision. Eventually, the majority of donor support might be provided to the general budget, without being earmarked for specific programs, and capacity would have to be established for the authorities to manage the purchaser role.

C. Medium-Term Fiscal Framework

29. The authorities will remain highly dependant of foreign assistance over the medium-term. To secure ongoing donor support for their development strategy, the government, in collaboration with some of the main donors, has conducted a medium-term needs assessment and defined a broad strategy aimed at attaining the Millennium Development Goals (MDGs) for economic growth and poverty reduction. This strategy, presented in the document Securing Afghanistan’s Future (SAF), was endorsed by donors at a conference held in Berlin in March 2004. Operating expenditure needs are estimated at $7.2 billion over the next seven years, of which $4.3 billion would be financed by domestic revenue and $2.9 billion from external resources (Table II.2). The development program, which is wholly donor-financed, is estimated to cost $24.6 billion over seven years. Table II.2 sets out the aggregate medium-term fiscal projections set out in the SAF report.24

Table II.2.

Islamic State of Afghanistan: Projected Medium-Term Fiscal Framework, 2004/05–2010/11

(In millions of U.S. dollars)

article image
Source: Government of Afghanistan, Securing Afghanistan’s Future, March 2004.Notes: the figures for FY2004/05 differ slightly to those presented in the approved Core Budget.

30. Donor pledges are substantial but focus on the next three years. During the Berlin donor conference, donors pledged to provide $4.5 billion to Afghanistan for 2004/05, slightly exceeding the amount requested by the government.25 Donors also pledged to provide an additional $8.2 billion for the three-year period covering 2005/06–2007/08, compared with the government’s estimated requirement of $11.9 billion. Donors indicated that additional pledges would be made at a later date, as countries budgetary procedures often prevented them from making firm multi-year commitments.

31. The composition of expenditures over the medium-term will be largely determined by the availability and nature of external resources. As noted above, a considerable amount of the funding within the core and external budgets is not fungible. In determining the fiscal envelope the MoF must consider a number of constraints (see also Box 1), including the following:

  • The annual budget appropriation law and Central Bank law include a “no-overdraft” financing rule, forbidding the government to borrow by running an overdraft with the Da Afghanistan Bank (DAB). In addition, the government has made a policy commitment to use concessional loans only for development spending;

  • The absence of domestic debt instruments, such as Treasury bills or bonds, which would enable the government to raise additional domestic resources;

  • Many of the donor resources are earmarked: (a) LOTFA funds are earmarked to security-related expenditures, with the priority given to wage costs within the operating part of the core budget; (b) ARTF, International Development Association (IDA) and Asian Development Bank (AsDB) funds may not be used to finance security-related expenses; and (c) the majority of core and external budget funds are provided for specific projects.26

32. The priority for the government is to fund the operating budget. The fiscal policy commitments for the operating budget, outlined in Box 1, are designed to provide aggregate fiscal discipline, as expenditures are constrained by availability of domestic revenues and external (grant) resources. Given the large weight of salaries and wages in total operating expenditures, the funding requirements are largely driven by the ambitious public administration reform program (see section below). Domestic revenue mobilization, which is developing from a very a narrow base, is not expected to keep pace with the costs of the proposed public administration reforms over the near term, let alone to absorb the maintenance costs associated with the externally funded development programs. Figure II.2 outlines the expected profile, presented in SAF, for wages, overall operating expenditures and domestic revenue for the next ten years.

Figure II.2
Figure II.2

Islamic State of Afghanistan: Operating Budget Expenditure and Domestic Revenue Profile, 2003/04–2015/16

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Sources: Securing Afghanistan’s Future; and Fund staff estimates.

33. Under the government’s scenario, domestic revenue should be able to finance the wage bill within five years and total recurrent spending within nine years. In the interim, the government would be highly dependant on donor funding, but anticipates that the operating budget financing needs will be fully met though donor grants. The highly concessional loans provided by the international community have been earmarked in the 2004/05 budget to development programs and projects. Over the next seven years, development spending, which is not shown on the figure, would be entirely funded by external assistance.

34. On the revenue side, the fiscal revenue projections might be overly optimistic. Projecting revenue is particularly difficult in Afghanistan, where issues such as security, control over provincial transfers, the impact of opium production and eradication on the economy, as well as legal imports and significant tax policy reforms, obscure the outlook. Furthermore, the lack of reliable historical information also affects the quality of the estimates.

35. Reaching the ambitious revenue targets included in SAF will require robust economic growth and the continued strong implementation of a number of policy and administrative reforms. The IMF projections presented in Figure II.2 are based on the latest information available for 2004/05 and the five year reform programs that have been approved by the customs and tax administrations (see section below). The projections were based on updated macro economic information and follow broadly the same methodology as the authorities’ estimates (differing mainly on the expected effectiveness of implementation). The projections are driven by four elements:

  • Economic growth and inflation: The tax base is expected to grow in line with nominal GDP growth, or for import duties, in line with the growth in dutiable imports. These variables are particularly susceptible to the effects of the prolonged drought and also could be negatively effected in the short term by an opium eradication campaign;

  • Customs and tax policy reforms: The government enacted a sweeping reform of trade taxes towards the end of 2003/04 and implemented a wide range of tax reforms aimed at mobilizing revenue (outlined below);

  • Customs administration reforms: Customs duties are currently the predominant form of revenue in Afghanistan (Table II.3). Measures to improve customs administration will focus on improving enforcement and compliance. Progress encompasses implementation of international standards for classification of goods, progressive recovery of customs functions from other agencies (e.g., valuation of goods), and replacement of multiple customs declaration forms with a single form for all types of imported goods; and

Table II.3.

Islamic State of Afghanistan: Domestic Revenue Collection, 2003/04 and 2004/05

article image
Source: MOF, Fund staff estimates.

See section 2 for revised projections

  • Tax administration reforms: Measures to improve tax administration will also focus on enforcement and compliance. Also important, however, are such developments as the establishment of a Large Taxpayer Unit (LTU) within the MoF and improved provincial transfers.

36. The dependence of external funding and uncertain domestic revenue presents considerable risks to fiscal sustainability. Revisions to the macroeconomic assumptions and the rates of compliance, combined with delays to the implementation of some key tax reforms, may reduce the anticipated tax revenues as shown in Figure II.2. Unless remedial measures are taken to increase domestic revenue mobilization, by speeding up the reforms, additional external funding would be required to cover the operating budget over a more prolonged period or else the authorities would have to curtail some expenditures including those related to the current public administration reforms.27

37. In order to ensure that the budget remains fully funded over the medium term the authorities need to maintain donor commitments to fund the operating budget. In order to demonstrate the credibility of the budget and the reform program to donors, the authorities need to execute the budget as planned in addition to giving donors sufficient fiduciary assurances that their money would be spent in line with their objectives. The authorities will also have to negotiate with donors with regard to any shortfalls in domestic revenue within each year—to determine whether expenditures would be curtailed or additional external assistance provided. It should be noted that reducing aggregate spending may require ad hoc cuts in activities that may be inconsistent with broader policy objectives. Such negotiations with donors will determine whether the critical budget constraint should be the expenditure ceilings, as outlined in the SAF, or the operating budget deficit.

D. Rebuilding and Broadening the Domestic Tax Base

38. Revenue mobilization is critical to achieving fiscal consolidation. Domestic revenues have significantly increased over the last two years. In 2002/03, they reached $131 million, equivalent to 3.2 percent of GDP. In 2003/04, revenue collection reached $208 million, equivalent to 4 percent of GDP and outperformed Fund staff expectations. The budget target for 2004/05 has been set at $309 million, equivalent to around 5.5 percent of expected GDP.28 The customs and tax authorities have both adopted five-year reform plans, through 2007 and 2008 respectively, with the short-term focus on extending coverage in the provinces and enhancing tax compliance.

39. The domestic revenue base is concentrated in a relatively small number of taxes and is derived from a few locations (Table II.3). Import duties are the main source of domestic revenue, comprising over 50 percent of the total. More than 93 percent of total revenue is collected by central ministries and just five of the 34 provinces (i.e. Herat, Nangarhar, Kandahar, Balkh, and Kabul). The majority of revenues attributed to the central ministries are in the form of sales of goods and services and fees and charges (e.g. sale of telephone services, registration of motor vehicles, the sale of passports and over flight charges). Afghanistan’s fiscal system is very centralized and the provincial level of government refers to a deconcentrated level of central government, rather than to a subnational level of government. With the creation of the TSA, provincial administrations can no longer spend money directly from the revenues they collect as these are automatically swept into the TSA and then transfers are made to fund expenditures in accordance with the government’s budget. This has considerably improved the reporting and transfer of revenues collected in the provinces to the central government, a critical component of the move towards fiscal consolidation (see Statistical Appendix Table 31)

Cash Management in Afghanistan

Government cash management may be broadly defined as the cost-effective management of the government’s short-term cash flows and cash balances to ensure the smooth government payments.

The options for cash management in Afghanistan are severely limited. Key constraints include:

  • The government’s commitment to a strict “no-overdraft” financing rule, which forbids the government to finance its operations through running an overdraft with DAB is embodied in the annual Budget Decree. The new central bank law further codifies this policy;

  • The absence of any domestic government debt market that could be used to borrow or place liquidity;

  • The government policy to fund, to the extent possible, the operating budget exclusively through domestic revenue and external grants, and to use concessional borrowing only for development spending.1/

The overriding requirement of cash management is to ensure that the government has, at all times, a sufficient cash buffer to meet its obligations. With monthly expenditures averaging Af 2.3 billion during the second quarter of 2004/05 domestic government cash balances peaked at Af 2.3 billion at the start of the current fiscal year but rapidly declined thereafter (see Figure). This decline was mainly attributed to the slow rate of disbursement from ARTF and LOTFA. As a result of the cash shortfall, general program loan accounts have served as a temporary buffer for bridging cash shortages in the operating budget to avoid disruptive cash rationing. To avoid a repeat of this in the future, the authorities must maintain a prudential cash reserve, which should be available to compensate for any unexpected shortfall in other sources of funding.

Continuing to improve cash planning and forecasting is a high priority for the implementation of both fiscal and monetary policies. There are commonly three stages in the development of a basic cash management system (with Afghanistan currently at stage 1), these are:


Local currency cash balances 1382–1383

Citation: IMF Staff Country Reports 2005, 034; 10.5089/9781451800203.002.A001

Stage 1: Improving the control over cash flows. The recently developed Treasury Single Account (TSA) helps to integrate government accounts, with balances regularly swept into a single account held by the MoF. Other measures being developed to control cash flows include: modern payment procedures (e.g. Electronic Funds Transfer) to suppliers and direct employee salary payments (currently the plans are only for Kabul). Commitment controls should also be introduced to facilitate expenditure management and help coordinate the timing of payments;

Stage 2: Developing a system of cash plans and targeting of the amount of cash reserve to be available for the government payments. The should allow timely action to anticipate possible cash shortfalls in the future and assist with the coordination of fiscal and monetary policy. The main constraint is the capacity in the MoF to develop this capability; and

Stage 3: Development of short term instruments to manage liquidity (including Treasury bills) designed to offset temporary cash imbalances because of timing of cash flows in and out of government. The MoF has, so far, been reluctant to consider the creation of domestic debt.

1/ It should be noted that an initial advance from the ARTF, of $50 million, provided the liquidity for cash advances to provinces and expenditures in the center.

40. The remaining provinces and ministries either have much lower revenue capacity or suffer from poorer collection rates. Many ministries and provinces collect a minimal amount of revenue and, with around 90 different active taxes, it is highly likely that the administrative costs of some taxes outweigh their revenue gains. This suggests some scope for rationalizing the number of taxes collected, which should be helped by the ongoing work to strengthen the classification of revenue into more internationally recognized categories.

41. Customs administration reform will continue to dominate the revenue mobilization effort over the medium-term. In 2004 the customs administration implemented measures to: (a) improve compliance to ensure that all provincial customs revenues are transferred to the TSA; and (b) simplify the tariff system, including the use of market exchange rates for import valuation, a streamlined tariff structure (moving from 25 tariff rates to six), a reduction in tariff dispersion (with rates ranging from 2.5–16 percent, compared with 0–150 percent previously), and a more effective broker process was established. Ongoing customs reforms include: (a) improving customs infrastructure through major building projects supported by USAID and the World Bank; (b) introducing of a computer system, ASYCUDA; (c) re-organizing and re-building the capacity of central and regional customs offices through an extensive DFID training program; and (d) establishing effective enforcement mechanisms and reducing illegal charges.

42. Tax administration reform: the tax base is likely to remain relatively narrow over the medium-term and reforms being implemented will focus on large taxpayers. The current set of reforms include: (a) the adoption of tax reform package for 2004/05; (b) the creation of a Large Taxpayer Office; (c) the nationwide roll out of the new Taxpayer Identification Number (TIN) to the major provincial centers; and (d) the consolidation of recent tax legislation amendments to address the excessive use of tax holidays, exemptions and ‘special agreements that were in danger of eroding the tax base.29 Provisions for income tax, capital gains, and company tax will be introduced under a new tax law and the government must also ensure that new laws being drafted for the mining, oil and gas sectors are consistent with the tax code.

43. Moving to a broad-based consumption tax: The current indirect tax system contains serious deficiencies, including: (a) a narrow tax base leading to low revenue yield; (b) serious potential cascading (albeit generally at a low rate); and (c) disincentives to export. The phased extension of the recently introduced taxes on selected services should pave the way to a broader-based consumption tax but it is unlikely that the government will have the administrative capacity to implement this for at least five years.

E. Public Expenditure Policy and Management

44. Further improvements to the legislative framework and PEM systems are needed to remove the main impediments to budget consolidation. Most of these improvements have already taken place at the central MoF, with improvements in line ministries and in the provinces being much slower: a TSA has been established, consolidating all government accounts; a computerized management information system has been set up in the ministry of finance and will be gradually rolled out to other parts of the government; donor-funded capital spending is increasingly being channeled through the treasury; and a public financial management law has been prepared and is expected to be approved shortly.

45. The government urgently needs to publish in the official gazette recent amendments to the tax laws, including the tax reform package and the amendments to ensure that the revenue legislation is the only source of concessions or exemptions, and adopt a new organic budget law with strengthened rules for budget coverage, preparation, execution and reporting. High priority also needs to be given to a new customs code, internal audit and procurement legislation that should help to improve compliance rates.

46. The lack of capacity to execute spending plans, combined with problems caused by ongoing insecurity, are severely hampering the ability of the government and donors to implement development programs. This is particularly the case for reconstruction projects for which implementation capacity is very low. Considerable amount of technical assistance is being provided to line ministries, designed to improve planning, financial management, procurement and audit capacity, but the experience of other post-conflict countries suggests that this problem will persist over the medium-term.30 The rate of spending is also likely to be negatively effected if the security situation deteriorates. It is therefore important that the government manages public expectations and produces realistic budgets that reflect the limited absorptive capacity, as opposed to ‘wish lists’.

47. Building an effective and affordable public administration will be a major challenge. The authorities’ policy, as set out in the 2004 policy document Securing Afghanistan’s Future, is to broadly maintain the size of the civil service but to invest heavily to build its capacity.31 These proposals entail significant additional recurrent and investment costs over the medium-term. The wage bill will therefore constitute a growing proportion of the operating budget in the short term, growing faster than the projected increases in GDP, and remain significant over the medium term (Figure II.2).

48. The authorities consider the move to a small, but in the medium term a relatively high-cost, public service as unavoidable if capacity is to be restored. According to World Bank estimates, the average pay for civil servants is likely to reach a multiple of almost five times per capita income within five years, significantly above the average for Asia (Table II.4), leveling off thereafter. However, in weak economies with very low capacity, relatively higher salaries are commonly paid in order to retain well qualified staff (Table II.4).

Table II.4.

Public Sector Wages

article image
Sources: Government Financial Statistics database (IMF); International Financial Statistics database (IMF); World Economic Outlook database (IMF); IMF Staff; the online World Bank Cross-National Data on Government Employment & Wages; and Schiavo-Campo, Salvatore, Giulio de Tommaso and Amitabha Mukherjee (1997), An International Statistical Survey of Government Employment and Wages, Policy Research Working Paper 1806 (Washington: The World Bank).

Data refer to 1996–2000 average.

Data refers to the 1996–2000 average.

49. Over the longer term, the ratio between average civil service pay and per capita GDP can return to levels more in line with regional averages. The government has already started a comprehensive pay and grading review across the civil service that will provide an alternative to the necessarily fragmented schemes that comprise the current capacity building framework and will consider questions of long-term sustainability.

50. The current approach entails significant fiscal risks. Despite the anticipated growth in domestic revenues, donors will be required to continue to finance a large share of salaries and wages over the medium-term. The authorities must ensure that the proposed salary costs (as set out in SAF) can be funded by available domestic and external financing over the medium-term. This will likely be a complex task as the authorities may need to respond to shortfalls in domestic revenue or fluctuations in donor funding, which is traditionally volatile. To retain external support the government must also demonstrate that this approach is leading to commensurate improvements in government capacity and continue to resist the significant pressures for general or sector pay increases—these pressures may be exacerbated by the number of donor-financed projects used to fund staff at greatly varying pay rates. For example, the average civil servant is paid around $50 per month, while the main capacity building program (the Priority Reform and Restructuring, PRR) pays up to $300 in additional allowances per month, and programs to attract well-qualified managers may pay substantially more. The pay rates for the new Afghan National Army are also significantly above those of the rest of the civil service.

51. Significant technical assistance (TA), including by FAD, has established the framework for consolidating budget management systems. The framework includes: establishing the TSA; introducing a basic computerized management information system (AFMIS, currently only operational in the MoF in Kabul, but there are plans gradually extended computerization to the provinces); reforming payment systems; developing accounting and auditing standards and capacity. While significant progress has been made, these reforms will be slow to implement and most public financial management practices outside the MoF remain rudimentary and paper-based.

52. Afghanistan is receiving a considerable amount of TA in the areas of PEM, tax policy and revenue administration from a wide range of donors. Several FAD technical assistance missions have assisted in the reform process with the focus of advice in the following areas: (a) recommending reform strategies and specific short–and medium-term priority measures; (b) helping to design programs of technical assistance to ensure successful; and (c) monitoring and evaluating the implementation of reforms and to suggest improvements. Some lessons from this TA include: (a) the pace of reforms has been slowed by the poor security situation. For example, it was a daunting task to reestablish the central government’s authority over the provinces; (b) the inclusion of TA recommendations as conditionality in Fund-supported programs helped with implementation; (c) flexibility in the overall strategy is important to respond to changing circumstances on the ground; and (d) good donor coordination is critical for success, especially when a large number of donors is involved.32


Table 1.

Islamic State of Afghanistan: Key Economic Indicators, 2001/02–2004/05

(In units as indicated)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates.
Table 2.

Islamic State of Afghanistan: Gross Domestic Product at Current Prices, 2001/02–2004/05

(In millions of new Afghanis)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates and projections.
Table 3.

Islamic State of Afghanistan: Gross Domestic Product at Constant Prices, 2001/02–2004/05

(In millions of new Afghanis at 2002/03 prices)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates and projections.
Table 4.

Islamic State of Afghanistan: Gross Domestic Product at Constant Prices, 2001/02–2004/05

(Annual percentage change)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates and projections.
Table 5.

Islamic State of Afghanistan: Sectoral Shares of Gross Domestic Product, 2001/02–2004/05

(In percent of GDP at market prices)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates and projections.
Table 6.

Islamic State of Afghanistan: Savings-Investment Balances at Current Prices, 2001/02–2004/05

article image
Sources: Afghan authorities; and Fund staff estimates and projections.

Data originating from the fiscal accounts.

Determined as a residual.

Equivalent to foreign direct investment reported in the balance of payments, and some new private investment.

Equivalent to the current account deficit, excluding grants.

Table 7.

Islamic State of Afghanistan: Opium Production and Prices, 1994–2004

article image
Sources: United Nations Office on Drugs and Crime, Afghanistan Opium Survey for 2004, November 2004; and FAO/WFP, Crop and Food Supply Assessment Mission to Afghanistan.
Table 8.

Islamic State of Afghanistan: Agricultural Crop Production, 2000/01–2004/05

(In tons)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates.
Table 9.

Islamic State of Afghanistan: Land Use, 2001/02–2004/05

(In thousands of hectares)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates.
Table 10.

Islamic State of Afghanistan: Cultivated Land Area for Fruits and Vegetables, 2001/02–2004/05

(In hectares)

article image
Sources: Data provided by the Afghan authorities; and Fund staff estimates.