Pakistan: Selected Issues and Statistical Appendix

This Selected Issues paper analyzes Pakistan's tax reform and revenue performance. The paper assesses the nature and magnitude of Pakistan's actual twin debt problem in a historical context, and reviews the policy options. The study highlights the factors explaining the recent stagnation in merchandise exports, and analyzes the country's export performance with regard to the linkages between performance and the structure of exports. The paper reviews a survey of poverty in Pakistan, and also provides a statistical appendix report of the country.

Abstract

This Selected Issues paper analyzes Pakistan's tax reform and revenue performance. The paper assesses the nature and magnitude of Pakistan's actual twin debt problem in a historical context, and reviews the policy options. The study highlights the factors explaining the recent stagnation in merchandise exports, and analyzes the country's export performance with regard to the linkages between performance and the structure of exports. The paper reviews a survey of poverty in Pakistan, and also provides a statistical appendix report of the country.

II. Debt and Debt Sustainability Issues in Pakistan

A. Introduction21

48. It is widely believed that Pakistan suffers from acute debt problems. High levels of public and external debt are typically mentioned among the two or three most immediate economic policy issues in Pakistan.22 Many observers also argue that the usual ramifications associated with debt problems—low investment ratios, crowding out of essential social spending and development expenditure, balance of payments financing problems, and increasing constraints on macroeconomic policy management—have also begun to surface.

49. These concerns about debt problems are not surprising given that Pakistan’s public and external debt stocks at about 92 percent and 58 percent of GDP, respectively, are high by international standards. Interest payments on public debt soaked up 48 percent of revenue in the consolidated government budget, while scheduled external debt service payments reached a stunning 64 percent of current foreign exchange receipts in the last fiscal year (1999/2000). Another indication of the problem is the external debt emergency of 1998/99, when a comprehensive restructuring of Pakistan’s external debt was needed in the wake of the balance of payments crisis that began to unfold after May 1998.

50. Recent debt developments and current debt data suggest that debt problems will not disappear quickly. Under the new program to be supported by a Stand-By Arrangement with the Fund, short-term balance of payments viability requires yet another round of flow restructuring of the external debt. Progress towards achieving a sustainable medium-term debt service profile will also be an important policy challenge. Moreover, while perhaps less immediate, domestic debt has become a problem for policymakers, as the high real interest rates on domestic currency debt may push the debt dynamics on to an unstable path.

51. This Section attempts to assess the nature and magnitude of Pakistan’s actual “twin” debt problem in a historical context and to review policy options. The analysis focuses on two related sets of issues. The first set revolves around the burden associated with external and public debt. The second set concerns debt sustainability, that is, issues related to the question of how policies would need to be adjusted to ensure that external and public debt could be serviced in an orderly fashion, while other goals of economic policy, such as high economic growth and macroeconomic stability, could be met.

52. In Pakistan, as in so many other countries, public debt and external debt issues are often treated synonymously. Although they are closely interrelated, especially in view of the large share of public and publicly guaranteed external debt in both total external and total public debt, the two dimensions are nevertheless distinct. Analytically, they need to be explored separately. Public debt issues mainly concern the capacity of the government to service its debt in domestic currency while external debt issues have to do with the capacity of the economy to raise the foreign exchange needed to meet the debt service obligation toward nonresidents. Instruments and issues differ in the two domains although policy actions in either domain typically affect each other. Moreover, good policy strategies in one domain need not be consistent with good strategies in the other, and analysis along the two dimensions is required to arrive at a strategy that is consistent with overall macroeconomic objectives.

53. At the outset, it should be noted that the lack of a comprehensive debt monitoring system complicates the analysis of debt issues. In particular, the total amounts and the composition of both public and external debt remain subject to uncertainties for two reasons. First, different agencies are responsible for monitoring and reporting, which has led to the regular publication of only a subset of debt data by each agency. Second, the various components are not yet fully consolidated into total domestic public debt and total external debt. Despite these caveats, the Section attempts to provide an integral perspective on the quantitative dimensions of Pakistan’s debt problems.

54. The Section is organized as follows: subsection B addresses issues related to the burden and sustainability of Pakistan’s external debt; the subsequent subsection focuses on the same issues for the public debt; and subsection D attempts to provide a policy perspective on Pakistan’s debt problems.

B. External Debt Burden and Sustainability

55. This section analyzes the burden and sustainability of Pakistan’s external debt levels. The external debt concept used is that of total external debt, that is, the total debt that resident public and private entities in Pakistan contracted from nonresidents. The focus on total external debt is important because the severity of the 1998/99 balance of payments crisis was, in part, related to the relatively large outstanding stock of short-term private external debt at the outset.

56. External debt sustainability is concerned with the capacity of the government and other parties that have contracted external debt to service the debt in an orderly manner. Orderly debt service is, however, only a necessary condition for debt sustainability.23 Sufficient conditions for a sustainable external debt level are that the debt can be serviced as scheduled under conditions of high medium-term economic growth and macroeconomic stability. In addition, sustainability is often understood to encompass the condition that the debt service capacity is robust to possibly persistent perturbations to the anticipated path of key variables. If these conditions are satisfied, then the burden of the external debt on the economy can be considered bearable.

57. As external debt, denoted with F hereafter, covers all debt owed to nonresidents, its sustainability is closely related to external current account sustainability, especially for a country at Pakistan’s level of financial market development. Under such conditions, external debt transactions should account for a large fraction of all transactions recorded in the financial account of the balance of payments. Accordingly, it is assumed in this Section that the change in external debt is about equal to the change in net foreign liabilities.

58. Refinancing aspects are important in the case of external debt as the mobilization of foreign exchange resources for amortization payments is an issue, not the least because the willingness of nonresidents to refinance the principal remains outside government control. Even if macroeconomic policies are, in principle, consistent with best practices, expectations of nonresidents may be such that foreign exchange flows remain scarce. Hence, while for domestic currency debt it is primarily interest payments on public debt that matter, total debt service matter a great deal more in the case of external debt, especially in a country with recent balance of payments difficulties.

Pakistan’s external debt during the 1990s—stylized facts

The level of external debt

59. Pakistan’s external debt amounted to 58 percent of GDP at the end of 2000 (Table II-1).24 Over the last few years, the ratio of external debt to GDP fluctuated around 57 percent. Compared to the beginning of the 1990s, however, the external debt as a percent of GDP increased by about 9 percentage points. In terms of exports or, more precisely, current foreign exchange receipts25—another frequently used yardstick to measure the external debt burden—the external debt fluctuated around 280 percent in recent years. Compared to the early 1990s, an increase in the external debt was also registered on the basis of this benchmark.

Table II-1.

Pakistan: External Debt, 1990–2000 1/

article image
Source: State Bank of Pakistan; Ministry of Finance; and Federal Bureau of Statistics.

External debt after rescheduling/restructuring as of June 30 in each year.

Including central bank.

Foreign Currency Bearer Certificates.

Public sector short-term debt and military debt.

Including state-owned commercial banks.

Nonresident foreign currency deposits.

Includes nonguaranteed private debt, including bank borrowing other than nonresident foreign currency deposits.

Short-term debt at original maturity plus amortization payments of medium-term debt of the following year.

The debtor and creditor composition of external debt

60. External debt contracted by the public sector or guaranteed by the government has been the dominant type of external debt by debtor for many decades. At end-1999, public and publicly guaranteed debt, henceforth public external debt, accounted for about 87 percent of total external debt at end-2000 (Table II-2). During the entire last decade, this debt category fluctuated around 48 percent of GDP and 235 percent of current foreign exchange receipts.

Table II-2.

Pakistan: Structure of External Debt, 1990–2000 1/

article image
Source: State Bank of Pakistan; Ministry of Finance.

External debt after rescheduling/restructuring as of June 30 in each year.

Including central bank.

Foreign Currency Bearer Certificates.

Public sector short-term debt and military debt.

Including state-owned commercial banks.

Nonresident foreign currency deposits.

Includes nonguaranteed private debt, including bank borrowing other than nonresident foreign currency deposits.

Short-term debt at original maturity plus amortization payments of medium-term debt of the following year.

61. While the dominance of public external debt remained unaffected, the 1990s nevertheless witnessed a profound change in the debtor composition of external debt, as the share of private external debt rose from about 2 percent at end-1990 to a maximum of 20 percent at end-1997.26 This increase also accounted for most of the increase in total external debt as a percent of GDP or current foreign exchange receipts registered between 1990 and 1997. With the debt crisis that began to unfold in 1999, the share of private debt decreased.

62. The debtor structure of private external debt varied during the last decade, partly because of the rise and fall of nonresident foreign currency deposits and partly because of the steady rise of other private debt. The latter debt category includes external liabilities of independent power producers (IPPs), which had begun to invest and operate in Pakistan after the market for electricity generation was opened for private sector participation in the mid-1990s.

63. In recent years, the composition of external public debt was stable in terms of creditors. At end-2000, about 80 percent was held by official creditors, about half of which is debt owed to bilateral creditors. Debt owed to the general, nonofficial public, which includes debt to commercial banks or to the general private sector in the form of bonds, only amounted to about 6 percent at end-2000. The remainder of external public debt includes military debt and short-term public debt other than debt owed to commercial banks.

The structure of external debt by instrument and maturity

64. With the large share of public and publicly guaranteed external debt owed to official creditors, Pakistan’s external debt is mostly long-term, if classified by the criterion of original maturity (maturity at the time of contraction) and is in the form of debt contracts that are generally not tradable. The share of tradable debt in the form of securities, which consists of Eurobonds and bearer securities in foreign currency, remained very small. Short-term public and publicly guaranteed external debt by initial maturity, which includes debt owed to commercial banks, foreign currency bearer certificates, and some central bank liabilities, was typically less than 10 percent of total public and publicly guaranteed external debt.

65. Private external debt was always more short-term in nature, especially foreign currency deposits. The share of demand and time deposits with a maturity of 12 months or less in total foreign currency deposits was typically above 80 percent.27 Details on the maturity structure of other private external debt is not available. According to the SBP’s latest annual report this debt category includes only medium and long-term debt (at original maturity), including supplier credits and cash loans for import financing.

66. Overall, short-term external debt at original maturity (excluding possible short-term elements in other private debt) was generally somewhat less than one fifth of total external debt during 1992–98. The amounts of short-term external debt were large enough to contribute to considerable external financial vulnerability, as the comparison with the usual yardstick of official foreign exchange reserves shows.

67. The picture on the short-term debt during the 1990s would not be complete if the rapid increase of effective short-term debt were not mentioned. The concept of effective short-term debt is more general because it includes all debt coming due over the next 12 months, including amortization payments on long-term debt, in addition to the short-term debt at original maturity. With the increasing amounts of scheduled debt service payments on medium-term public and private debt, which are discussed in more detail later, the amounts of effective short-term debt began to account for more than 20 percent of total external debt from the mid-1990s. Effective short-term debt exceeded official foreign exchange reserves by a large margin, which was yet another illustration of the increasing external financial vulnerability of the economy during the 1990s.

68. At this stage, it should also be noted that with the restructuring and rescheduling of the external debt during 1999–2000 (discussed in Box I-1), the distinction between short-term and other external debt has become blurred because some debt that was short-term, if classified by the initial maturity criterion, effectively became medium-term debt after rescheduling. Similarly, after end-1998, the effective debt also decreased if only amortization payments coming due after rescheduling are included.28 At end-2000, the outstanding amount of effective short-term debt after rescheduling was significantly lower than at end-1998.

The debt service burden of external debt

69. Arguably, the most important external debt related development during the 1990s was the dramatic increase in the debt service burden of external debt. When measured on the basis of scheduled payments, the burden almost doubled to about 13 percent of GDP by 2000 (Table II-3). In terms of current foreign exchange receipts, the scheduled debt service burden increased more than twofold to about 64 percent by 2000. Toward the end of the decade, the debt service burden, based on scheduled payments, obviously became increasingly unmanageable given export receipts and capital inflows.29 Debt service payments had to be restructured in the context of a comprehensive debt restructuring exercise, as described in Box II-1. Even after the debt restructuring and rescheduling, the actual debt service burden remained above 30 percent of current foreign exchange receipts during 1999–2000.

Pakistan: Debt Crisis and External Debt Restructuring After May 1998

In the aftermath of the events of May 1998, Pakistan had to embark on a comprehensive restructuring of its external debt service obligations during 1999–2000, as the debt service burden had become unmanageable. The main elements of the restructuring include:

  • In January 1999, the Paris Club provided debt relief on debt service from public and publicly guaranteed debt contracted prior to September 30, 1997 falling due between January 1, 1999 and December 31, 2000 (including arrears accumulated during the first half of 1999). In January 1999, the projected debt service relief granted over the 18-month period amounts to US$3.3 billion.

  • Pakistan froze withdrawals in foreign currency from all nonresident foreign currency deposits (FCDs) in May 1998, which amounted to US$4 billion at the time.1 Subsequently, it reached agreement with nonresident institutional investors on a more favorable repayment schedule for US$1.4 billion of FCDs. Other nonresident investors have been allowed to withdraw and cash their deposits in local currency (at the official rate until May 1999 and at the interbank rate after) or to swap them with the so-called Special U.S. dollar bonds issued by the Government of Pakistan.

  • In December 1999, Pakistan succeeded in exchanging three existing Eurobonds worth about US$610 million for a new six-year amortizing bond with a three-year grace period and a 10 percent coupon. Otherwise, US$450 million of repayments would have come due during 2000 (in addition, a put option on the remaining bond coming due in 2002 could have been exercised from February 2000).

  • In December 1999, Pakistan also reached agreement with eight commercial banks to restructure US$512 million of short-term trade credits. In addition, Pakistan rescheduled US$415 million of medium-term commercial bank credits in 1999.

  • The central bank succeeded in rolling over short-term and medium-term liabilities held by other central banks.

Overall, the amount of relief achieved through the restructuring of debt service obligations is estimated at about US$7.3 billion during 1999–2000. About 45 percent of the relief was provided by official, bilateral creditors.

1

Foreign currency withdrawals from bearer securities denominated in foreign currency (including the so-called Foreign Exchange Bearer Certificates, Foreign Exchange Bearer Certificates, and Dollar Bearer Certificates) and foreign currency deposits held by resident investors were also frozen. The corresponding outstanding liabilities amounted to about US$7.3 billion at end-May 1998.

Table II-3.

Pakistan: External Debt Service, 1992–2000 1/2/

article image
Source: State Bank of Pakistan; Ministry of Finance; and Federal Bureau of Statistics.

Fiscal year basis. Fiscal year runs from July 1 to June 30. For example, 1992 refers to the fiscal year running from July 1, 1991 to June 30, 1992.

Scheduled debt service before restructuring and rescheduling.

Including debt of state-owned commercial banks with government guarantee.

Scheduled debt service minus rescheduled debt service.

70. Another dimension of the debt service burden is the extent to which debt is effectively refinanced or rolled over through new debt inflows. The smaller are the amounts of refinancing or rollover funds, the more burdensome the debt service becomes because other means of financing are needed unless the external current account is in surplus. From this angle, a noticeable rise in the debt burden was also registered during the 1990s (Table II-4). The gross debt-related capital inflows decreased quite steadily after 1993/94, both in absolute value and in terms of GDP or debt service obligations. In fact, net debt-related capital flows (debt disbursements minus amortization) decreased steadily from a peak reached in 1993/94 and turned negative in 1997/98.

Table II-4.

Pakistan: External Debt Service and Capital Flows, 1992–2000 1/

article image
Source: State Bank of Pakistan; Ministry of Finance.

Fiscal year basis. Fiscal year runs from July 1 to June 30. For example, 1992 refers to the fiscal year running from July 1, 1991 to June 30, 1992.

Gross inflows minus scheduled amortization payments before restructuring and rescheduling.

The dynamics of external debt and debt service during the 1990s

71. The identification and quantification of the factors that contributed to the evolution of the external debt in the past typically provides useful information on prospects and policy issues. However, as argued below, the dynamics of the external debt cannot explain the dynamics of the debt service, which was determined by the changes in the debt structure.

Explaining the External Debt Dynamics

72. As shown in Appendix A, the change in the external debt as a percent of current foreign exchange receipts between t and t+1 can be decomposed into four factors that capture the most important balance of payments aggregates:

  • Net exports (the noninterest current account balance)30, which determine the need for the external financing of imports given overall receipts from exports of goods and service and private transfers.

  • The intrinsic debt-interest dynamics, which emanates from the difference between the interest rate on external debt and the growth of current foreign exchange receipts. If this difference is positive, the dynamics of interest compounding applies, which can lead to continued increases in the external debt unless net exports are, on average, positive and large enough to offset the interest bill. In general, a positive difference is expected, although a country like Pakistan is likely to be an exception given the significant share of concessional external debt.31

  • The accumulation of gross official foreign exchange reserves, which requires the accumulation of external debt unless the accumulation is offset by other capital flows or the current account balance.

  • Other factors, including other capital inflows such as foreign direct investment, which reduce the need for the accumulation of external debt.

73. The actual decomposition of Pakistan’s external debt dynamics during the 1990s according to this scheme can be found in Table II-5.32 In the following discussion of the results, it is convenient to distinguish between push and pull factors. Push factors are factors that would have contributed to increases in the external debt ratio if all other factors had remained unchanged. Pull factors are factors that would have pulled down the external debt ratio.

Table II-5.

Pakistan: External Debt Dynamics, 1992–2000 1/

article image
Source: Fund staff calculations based on data provided by the State Bank of Pakistan and Ministry of Finance.

Fiscal year basis. Fiscal year runs from July 1 to June 30. For example, 1992 refers to the fiscal year running from July 1, 1991 to June 30, 1992.

Averages for growth rate of forex receipts and the growth-adjusted interest rate on external debt.

Positive sign means contribution to an increase in the debt ratio.

Net exports

74. The generally large imbalance between exports and imports, which implied negative net exports (or noninterest external current account deficits), unambiguously was the most important “push” factor behind the increase in the external debt. These imbalances partly reflected Pakistan’s fundamental external vulnerability related to productivity developments in the agricultural sector. As agricultural outputs such as cotton and wheat are important determinants of the export supply potential as well as of the import needs, a series of productivity shocks in this sector during the 1990s led to highly pro-cyclical net exports. However, as illustrated in Chart II-1, cyclical variations in net exports due to a series of negative productivity shocks to the agricultural sector were only one part of the story. In addition, the overall performance of exports, including workers’ remittances during the 1990s, was disappointing, especially after significant growth in the 1980s, reflecting, inter alia, insufficient stabilization efforts and a lack of determination in carrying through the deep structural reforms needed to reduce the external vulnerabilities.

Chart II-1.
Pakistan: Exports, Imports, and Net exports
A02ct01
Source: Staff calculations based on data provided by the authorities.1/ Exports: exports of goods and services and workers’ remittances; imports: imports of goods and services and unrequited transfers (debit items).2/ Net exports equals exports minus imports as defined in footnote 1.
The intrinsic debt-interest dynamics

75. Interest payments on external debt were generally another push factor. As the growth rate of total exports fluctuated with a large amplitude in the last decade, the magnitude of the contribution depends critically on the period selected for the analysis. For the entire decade, the contribution to the debt dynamics was substantial at about 42 percentage points of current foreign exchange receipts. However, this magnitude is largely the result of a sharp drop in total exports in 1999 against the background of weak external demand related to the Asian crisis and the unfolding balance of payments crisis (workers’ remittances). Between 1992 and 1998, the contribution was merely 4 percentage points of current foreign exchange receipts compared to an increase in the external debt by 33 percentage points. A relatively small contribution on average was to be expected given the relatively constant and low nominal interest rate on external debt (Chart II-2). Table II-5 also shows how in years with favorable export growth, interest payments were a pull factor because interest rates on external debt adjusted for export growth were favorably low.

Chart II-2.
Pakistan: External Debt and Dynamics of External Debt Burden
A02ct02
Source: Staff calculations based on data provided by the Pakistan authorities.
Foreign exchange reserves and other factors

76. Changes in reserves were sometimes a push factor and sometimes a pull factor. Overall, the accumulation of reserves contributed to the increases in external debt, although the magnitude remained small as the reserves were quite often needed for balance of payments financing during the 1990s. Other determinants were a significant pull factor. These determinants include foreign direct investment, which began to contribute substantially to the balance of payments once the EPP program became operational in the mid-1990s. Other inflows were portfolio inflows, involving the sales of Pakistani assets to nonresidents.

Explaining the dynamics of the external debt burden

77. Analyzing the external debt dynamics also points to a another important aspect of Pakistan’s external debt problems. The change in the stock of external debt between 1991/92 and 1999/2000 cannot explain the sharp increase in the external debt service burden when measured against current foreign exchange receipts. As noted in Table II-5, the ratio of external debt to current foreign exchange receipts only increased by about 45 percent. Applying the beginning-of-period interest rate on total external debt, a seemingly innocuous assumption given the small variation over time, and the average debt maturity during the period to this increase suggests that the debt service burden should have increased by roughly 8 percentage points to about 39 percent of current foreign exchange receipts. The actual debt service ratio, however, more than doubled. It is important to note that this observation is not just the result of balance of payments arithmetic during a debt crisis, when rescheduling or exceptional financing is “below the line” and the calculation of the scheduled debt service “above the line” assume immediate repayment in full. If the same back-of-the-envelope calculation were done for the period 1992–98, a significant discrepancy between actual and “fitted” increase in the debt service ratio would also emerge.

78. The heavier debt service burden must be the result of either higher interest rates on the public debt, changes in the structure of external debt, or of a reduction in the growth of current foreign exchange receipts relative to that of debt service payments on the outstanding debt (at the outset of the comparison).

79. The increasing principal payments, which more than doubled both in terms of GDP and current foreign exchange receipts, were the first important factor underlying the sharp rise in the debt service burden during the 1990s. The increase in overall principal payments must be attributed largely to the increased share of private debt, which must have been more short-term on average compared to public debt. Nevertheless, in the early 1990s, a small rise in the ratio of principal payments on public debt to the stock of public debt (by about 1 percentage point as shown in Chart II-2) contributed also to the increased principal payments. This suggests that the structure of public and publicly guaranteed debt with regard to maturity or concessionality (interest rate, grace period) became somewhat less favorable between 1992–95. Since then, the structure of public debt appears to have remained broadly unchanged. As evinced by the increase in the ratio of principal payments on private debt to the stock of private debt, the effective maturity of private debt must have decreased considerably from the mid-1990s. On the basis of this measure, the effective maturity of private debt fell from roughly 10 years in 1996 to about two years in 2000.33 Again, it is important to note that this general trend began to emerge before the debt and balance of payments problems in 1999. Between 1992 and 1998, amortization payments on private debt rose from about 9 to 36 percent of the stock of private debt.34

80. A second factor was the increase in the implied interest rate on private external debt during the early to mid-1990s (Chart II-2). In contrast, interest payments on public debt rose broadly in line with the debt stock. This only supports the conclusion that the increased share of more expensive and more short-term private external debt was an important reason behind the deterioration in the debt service burden during the last decade.

81. A third and important factor behind the deterioration in the debt service ratios as shown in Chart II-2 was the sharp slowdown in the growth of current foreign exchange receipts from 1992. In the circumstances, debt service became more of a burden with sometimes large external current account deficits and a growing debt stock. Hence, the unfavorable export performance affected the debt service ratios also through the very unfavorable “denominator” effect on the debt service ratios, in addition to the effects on the debt dynamics through sometimes large external current account deficits (before interest payments).

Dimensions of Pakistan’s debt burden

82. In practice, the assessment of external debt sustainability relies heavily on the analysis of past debt dynamics on the one hand, and experience, cross-country comparisons, and rules of thumbs on the other.

A cross-country comparison of Pakistan’s debt burden

83. Table II-6 shows a cross-country comparison of indicators for the debt burden of a country, covering the period from 1990 to 1998 so that it remains unaffected by the exceptional financing received during 1999–2000. The main results of the comparison are as follows:

  • As a percent of current foreign exchange receipts, Pakistan’s stock of external debt was about at the average level for low-income developing countries in 1998 but above average when compared to developing countries as a group, middle-income developing countries, or to South Asia.35 Compared to heavily indebted countries, the country’s external debt remains below average.

  • As a percent of GDP, Pakistan’s stock of external debt was above average compared to all but heavily indebted countries in 1998.

  • Comparing the debt stock indicators for 1990 and 1998 shows a noticeable deterioration in Pakistan’s relative debt burden.

  • Pakistan’s debt service indicators are above average compared to all country groups in 1998. In 1990, total debt service was already above average, while interest payments on external debt were in the upper echelon but not at the highest level compared to other country groups.

Table II-6.

Cross-Country Comparison of Debt Burden Indicators 1/

article image
Source: World Bank, Global Development Finance 2000 (except for Pakistan); and data provided by the Pakistan authorities.

Calendar year basis except for Pakistan, for which data is on a fiscal year basis.

84. The cross-country comparison confirms the conclusion that Pakistan suffers from a heavy external debt service burden. It also illustrates how the already high debt service burden in the early 1990s was further aggravated by large external current account deficits and a deterioration in export performance.

85. An interesting outcome of the comparison is that the Pakistan’s debt service burden was above average in 1998 when evaluated against low-income developing countries, while the country’s debt stock ratios were about at average levels. In light of the previous analysis of the dynamics of the external debt burden, an obvious explanation would point to above-average shares of nonconcessional public debt and private debt with shorter maturities and higher interest rates in Pakistan’s external debt, which reduces both its grant element and its average maturity.

86. The hypothesis of a less favorable debt structure is confirmed by World Bank estimates.36 In 1997, the average grant element in the debt of low-income countries was about 44 percent, while that of Pakistan was only 21 percent. In contrast, in 1990, the average grant element in Pakistan’s external debt amounted to 35 percent, only slightly below the low-income countries’ average of about 41 percent. Similarly, in 1997, the average maturity of Pakistan’s external debt was only 12 years compared to an average of about 25 years in low-income developing countries, while in 1990, the same Chart was about 22 years for Pakistan’s external debt, and about 24 percent for low-income developing countries. At the same time, the share of private external debt in Pakistan rose from roughly 1 percent to about 20 percent of total external debt. The decreasing grant elements as estimated by the World Bank implies that in terms of the net present value of the debt, Pakistan’s debt is more of a burden when measured by the net present value of the debt stock as a percent of current foreign exchange receipts.

Rules of thumb: solvency index

87. Cohen (1988) proposed a solvency index to gauge the extent of the external debt burden. The solvency index s is the share of noninterest current foreign exchange receipts in period t that is needed to keep the external debt f (as a ratio of noninterest current foreign exchange receipts or “exports”) constant at the period t level in the future:

S=(rx^1+x^)f

where the r denotes the interest rate on external debt and where x stands for current foreign exchange receipts (a hat over a variables denotes a growth rate).37 High values of the solvency index would suggest that a significant share of foreign exchange earnings is needed for interest payments on external debt, which could indicate unsustainable debt levels.

The solvency index can be computed in two ways:

  • The first is to compute the index based on recent averages of interest and export growth rates, and on the most recent debt to export ratio. The resulting solvency index can then be compared to the actual external debt service, which shows the extent to which the debt service burden is a function of principal payments.

  • The second way is to compute the interest rate on external debt, adjusted by the export growth rate, which would be consistent with maintaining the current debt and debt service ratios in the indefinite future. The computed rate can then be compared with recent actual rates, which provides for an assessment from a slightly different angle.

88. Table II-7 shows external solvency indices for Pakistan, both on a year-on-year basis and on an average basis. The averages are shown for both 1992–98 and 1992–2000 to avoid that biased conclusions are drawn because of the debt restructuring and rescheduling during the last two years.

Table II-7.

Pakistan: External Solvency Index, 1992–2000 1/

article image
Source: Fund Staff calculations based on data provided by the State Bank of Pakistan and Ministry of Finance.

Fiscal year basis. Fiscal year runs from July 1 to June 30. For example, 1992 refers to the fiscal year running from July 1, 1991 to June 30, 1992.

89. In the first two lines of Table II-7, the solvency index based on actual interest rate and export growth data is compared with the actual debt service ratio. The solvency index is typically small when compared to the actual debt service ratio. For example, during 1992–98, on average only 0.7 percent of current foreign exchange receipts would have been needed to keep the external debt stock constant with actual interest rate and export growth rates.

90. Lines three and four of Table II-7 compare the interest rate implied by the solvency index with the actual growth-adjusted interest rate on external debt. The calculations suggest that with the actual interest payments on external debt, which averaged about 0.4 percent during 1992–98 and 2.1 percent during 1992–2000, it would have easily been feasible to keep the debt stock constant as a percent of exports. With actual export growth rates, interest rates between about 13 to 15 percent would have been consistent with keeping the debt stock ratio constant. Such interest rate levels are, of course, well above the levels registered during the 1990s. Overall, the solvency indices suggest that Pakistan’s external debt problem is primarily a liquidity problem arising from an unfavorable amortization profile.

Assessment

91. The cross-country comparison and the analysis of solvency indices suggest that the external debt problems are not related to the interest payments on external debt. Although the interest burden appears higher than that borne by other countries, especially low-income and South Asian countries in general, nominal interest rates on external debt are nevertheless sufficiently low to keep the intrinsic interest dynamics benign. Moreover, interest payments as a percent of current foreign exchange receipts are relatively low, as shown by the solvency index calculation.

92. What emerges clearly from the above analysis is that the external debt burden must be attributed to large amortization payments, which were the results of a large stock of external debt and unfavorable changes in the structure of external debt, as well as to a deterioration in the export performance. In recent years, scheduled debt service payments were extraordinary when compared to the standard benchmark in cross-country comparisons—current foreign exchange receipts. Moreover, they became even more burdensome in light of the turnaround in gross debt-related capital flows in the second half of the 1990s. It should be noted that this turnaround began before the balance of payments crisis that began to unfold after May 1998. With hindsight, one could conclude that the debt crisis that began to unfold after May 1998 had been looming during much of the 1990s.

93. A final verdict on the sustainability of Pakistan’s external debt would be premature without more detailed, forward-looking analysis of the structure of the external debt and the medium-term debt service profile. The events of the last two years obviously demonstrate that the external debt service was not sustainable, a conclusion that also follows from a comparison of Pakistan’s debt service ratios with standard benchmarks. With debt flow restructuring and rescheduling, this liquidity problem was resolved. In the short term, the available resources remain insufficient to meet the scheduled debt service obligations unless imports and per-capita consumption contracted dramatically, with corresponding socioeconomic and sociopolitical implications. It is for this reason that the new program to be supported by a Stand-By Arrangement includes another round of flow restructuring and rescheduling. The matter is less clear-cut when it comes to the sustainability of the outstanding debt stock. While the net present value of the debt is reduced with the flow restructurings and reschedulings, the answer to the question of whether the reduction is sufficient to allow debt ratios to return to levels considered as sustainable awaits further analysis.

C. Public Debt Issues and Sustainability

94. This section analyzes the sustainability and other issues related to public debt. Public debt issues are related to fiscal sustainability, that is, the capacity of the federal and provincial governments to service the debt in an orderly manner. Orderly debt service requires that the government is solvent, that is, that the net present value of government assets is at least equal to the net present value of government liabilities.38 If a government is indebted, as is typically the case, solvency requires that the net present value of future primary balances (including seignorage revenue) equals or exceeds the current stock of government (net) debt.39 Hence, on average, solvent governments will have to run surpluses in the future if real interest rates exceed real GDP growth.40

95. This solvency requirement emanates from accounting identities and is by itself not very meaningful from a policy point of view. A more meaningful approach is to examine whether fiscal policies are sustainable, that is, whether their implications are consistent with basic goals of economic policy. Following Razin (1996), fiscal policies can be considered sustainable if they are consistent with the regular servicing of the public debt as well as with high medium-term economic growth and macroeconomic stability and if they are robust to possibly persistent perturbations to the anticipated path of key variables such as growth and interest rates. In terms of instruments, the issue is the level and structure of adjustment in expenditure and revenue needed to ensure that a fiscal policy program becomes sustainable.

96. From this perspective, the sustainability of a fiscal policy program depends, among other factors, on the overall macroeconomic policy mix, the current level and structure of expenditure and revenue, and the current level and structure of the debt-to-GDP ratio. Against this background, the assessment of fiscal sustainability and robustness calls in principle for a comprehensive macroeconometric model, which unfortunately is not available for Pakistan. In the circumstances, the analysis of past debt dynamics and simple scenario analysis based on sustainability indicators are often useful for the assessment of fiscal sustainability.

Pakistan’s public debt—some stylized facts

Level of public debt

97. At 92 percent of GDP at end-June 2000, the level of net public debt in Pakistan is high by international standards (Table II-8). Net public debt is defined as gross public debt minus government deposits and its variations correspond to the budgetary financing of the consolidated federal and provincial government budgets.41 Unfortunately, it is impossible to eliminate what is usually considered a small amount of publicly guaranteed external debt that is not serviced by the budget.42 It should also be noted that the concept of net public debt used in this section excludes external central bank liabilities (including IMF credit) or other contingent government liabilities, as changes in these items do not affect budgetary financing. Other contingent government liabilities are also not included.43 The debt-to-GDP ratio has been increasing steadily since the mid-1980s, when it was 76 percent of GDP (Chart II-3). The same picture of a steadily increasing debt ratio also emerges when the debt is normalized by government revenue.

Table II-8.

Pakistan: Net Public Debt, 1975–2000 1/

article image
Source: State Bank of Pakistan; Ministry of Finance.

As of June 30 in each year.

Includes some publicly guaranteed debt but excludes external liabilities of the central bank and IMF credit (net).

Chart II-3.
Pakistan: Net Public Debt, Budget Balance, and Interest Rates
A02ct03
Source: Staff calculations based on the data provided by the authorities.
Federal and provincial debt

98. The outstanding public debt is by and large debt issued by the federal government; provinces have very little debt outstanding. The constitution allows provincial governments to mobilize external and domestic loans and grants. If they have outstanding loans guaranteed by the federal government, however, their borrowing must be approved by the federal government. As about 95 percent of total provincial debt is reported to be borrowing from or guaranteed by the federal government, the provinces can not make independent borrowing decisions in practice.

Composition of domestic debit44

99. Roughly 50 percent of the public debt is denominated in domestic currency and is held almost exclusively by residents. The share of domestic currency denominated debt (referred to as domestic public debt from hereon) in total public debt has remained remarkably stable in recent years and amounted to 54 percent of GDP in terms of gross debt and 45 percent in terms of net debt at end-2000 (Table II-9). The authorities classify the issued debt instruments into three main categories, permanent debt, floating debt, and unfunded debt:

  • The so-called permanent debt comprises largely medium- to long-term securities held by banks and nonbanks. Prize bonds, that is, bonds with a lottery element are also part of this category. This debt component can be considered to be medium to long-term debt on an original maturity basis.

  • Floating debt includes short-term government papers, especially 3–12 month treasury bills, which are mostly held by banks.

  • Unfunded debt refers to resources mobilized through the national savings schemes (NSS) administered by the federal government’s Central Directorate of National Savings. This funding category comprises a number of instruments. Most of them are of medium to long-term maturity although their effective maturity can vary in some instances as they include put option-like elements for the holder. For example, defense savings certificates have a 10-year maturity but allow for early redemption without penalty after one year.45 Unlike most of the permanent and floating debt instruments, they are issued on tap.

Table II-9.

Pakistan: Domestic Public Debt, 1990–2000 1/

article image
Source: State Bank of Pakistan; Ministry of Finance.

As of June 30 in each year.

100. The category other debt consists mainly of unbacked bank advances or transitory financing items such as security deposits. Bank advances are relevant mainly for the so-called commodity operations. Both the federal and the provincial governments engage in these operations, which involve the procurement of agricultural commodities such as wheat.

101. In recent years, the share of permanent decreased while that of the other two categories increased. The increase in the share of unfunded debt is especially striking. At end-June 2000, regular medium-to long-term debt accounted for less than 20 percent of domestic debt while the share of the other two categories was about 40 and 45 percent, respectively. While the average maturity structure of unfunded debt is difficult to pinpoint, the 40 percent of floating debt nevertheless constitute only a lower bound or the share of short-term debt in total net domestic public debt.

102. The increase in the share of unfunded debt during the 1990s is often attributed to the emphasis on raising the share of nonbank budgetary financing to reduce inflationary pressures arising from bank-financing of the budget. Given the segmented structure of Pakistan’s financial market, NSS instruments have in practice been the only debt instruments available to private investors and non-bank financial institutions.

103. Roughly 45 half of the net domestic public debt is held by the banking system. These holdings, equivalent to net credit to the government, in turn account for slightly less than one half of broad money M2. About one fourth of the net domestic public debt is held by the central bank, the holdings of which account for over 70 percent of reserve money. With such large shares of (net) government credit in total credit, it is clear that government financing needs and government debt have become an important and constraining factor for monetary policy.

Interest burden of the public debt

104. Overall, the average interest rate on Pakistan’s net public debt has been steadily increasing over the last 25 years (Chart II-3), both in nominal and real, growth-adjusted terms.46 Accordingly, the interest burden on public debt, as measured by budgetary interest payments as a percent of GDP, has been rising (Table II-10). In 1998, the budgetary interest burden reached, for the first time, more than 7 percent of GDP.

Table II-10.

Pakistan: Interest Payments on Public Debt, 1990–2000 1/

article image
Source: State Bank of Pakistan; Ministry of Finance.

Data are on a fiscal year basis. For example, 2000 stands for the fiscal year running from July 1, 1999 to June 30, 2000.

Implied interest rate minus percentage change in GDP deflator (GDP at market prices).

Inflation-adjusted implied interest rate minus real GDP growth rate (GDP at market prices).

105. The rise in the average interest rate on the net public debt is primarily the result of increases in the average interest rate on domestic currency denominated public debt, which, in connection with decreasing inflation rates and lower real GDP growth, accentuated the intrinsic debt-interest dynamics (Table II-10). This contrasts markedly with the interest burden of external public debt. The implied interest rate on this debt category remained well below inflation and real GDP growth rates, so that the relevant interest rate for the interest burden (the growth-adjusted real interest rate) remained negative.

106. The striking difference in the interest dynamics in recent years reflects the differences in the composition of external and domestic debt. As noted above, the former is owed mostly to official creditors, which often involves concessional terms so that the related interest dynamics was somewhat less affected by economic factors. Interest rates on domestic public debt on the other hand are partly market-determined and increased significantly with some financial liberalization and the trend increase in the stock of public debt. This trend is aptly illustrated in Chart II-4, which shows the 6-month treasury bill rate and the interest on defense savings certificates—an important instrument in the category of unfunded domestic debt—in nominal and real terms.47

Chart II-4.
Pakistan: Interest Rates on Domestic Public Debt Instruments 1/
A02ct04
Source: Data provided by the authorities and staff calculations.1/ Period averages on a fiscal year basis. For example, 1999 refers to the values in the fiscal year 1998/99 running from July 1, 1998 to June 30, 1999.2/ Nominal interest rates minus average annual inflation rate.

107. Besides general macroeconomic developments and the rising debt stock, the increase in the interest burden has sometimes been attributed in part to current practices in public debt management. In particular, NSS instruments have been issued on tap, and their rates of return used to be set with little consideration for general financial market developments or money market benchmark rates. During the 1990s, socio-political objectives such as the promotion of savings in rural areas also appear to have played a role in the determination of NSS rates of returns. Over time, a substantial after-tax return differential on NSS instruments (compared to treasury bills or bank deposits) emerged. In the second half of the 1990s, the return differential contributed to the increased absorption of private savings through the NSS, which has led to some financial disintermediation and has hampered the developments of a general market for government bonds. Mindful of these developments, the authorities have begun to reduce rates of returns on NSS instruments in 2000.

108. The soaring real interest rates on domestic debt have become a major challenge for policy makers in Pakistan. With the disappointing progress in revenue mobilization during the 1990s, their efforts to control budget deficits required the reduction of development expenditure (as a percent of GDP) and limited the scope for providing essential social and education services. Many economist would support the hypothesis that these changes in the structure of government expenditure were among the factors underlying the decrease in the average growth rate observed during the 1990s.

Public debt dynamics, 1975–2000

109. The identification of the factors that contributed to the evolution of public debt ratios in the past and present provides useful information on policy issues and constraints as well as some guidance with regard to the prospects. As shown in Appendix II-III, the change in the debt-to-GDP ratio between t and t+1 is determined by three main factors:48

  • Primary balance, which determines the budgetary financing needs for regular government operations. Primary surpluses alleviate the debt dynamics whereas primary deficits worsen it.

  • The intrinsic debt-interest dynamics, which emanates from the difference between thereal interest rate and real GDP growth. On average, this difference will be positive (except, possibly, for concessional external debt).49 Positive, so-called growth-adjusted real interest rates underlie the dynamics on interest compounding and are the key behind the potential for unstable debt dynamics.

  • Valuation effects, which arise from the effects of nominal exchange rate changes on the foreign currency denominated debt. A depreciation of the domestic currency against currencies in which external public debt was contracted raises the value of this debt in domestic currency term and, with less than proportional effects on domestic prices, in terms of GDP.

110. The actual decomposition of the debt dynamics in Pakistan into the main components can be found in Table II-11.50 As in the section on the external debt, the exposition is in terms of push and pull factors.

Table II-11.

Pakistan: Public Debt Dynamics, 1978–2000 1/

article image
Source: Fund staff calculations based on data provided by the Pakistan authorities.

Years covered in table are fiscal years. For example, 1998 denotes the fiscal year running from July 1, 1997 to June 30, 1998. See Appendix II-III for details of the underlying calculations.

Based on average revenue and primary expenditure ratios during 1991–98. A negative sign denotes a surplus.

Output gaps were calculated on the basis of 5-year averages of real GDP. A positive sign denotes a negative output gap, i.e., trend GDP exceeds the actual GDP.

Primary balance

111. Historically, primary deficits were arguably the most important push factor behind the increase in the debt-to-GDP ratio. During the 1990s, however, the contribution of primary deficits to the public dynamics began to decrease noticeably, especially from 1995. In 1998/99, a primary surplus was recorded for the first time.

112. The decreasing contribution of the primary balance to increases in the debt ratio was the result of the more determined fiscal adjustment efforts in recent years. In the more detailed decompositi