Chapter

III Postwar Reconstruction, Public Finances, and Fiscal Sustainability

Author(s):
Thomas Helbling, and Sena Eken
Published Date:
June 1999
Share
  • ShareShare
Show Summary Details
Author(s)
Thomas Helbling

Since the end of the civil war, Lebanon has embarked on an ambitious program of economic reconstruction and stabilization. A rapid rehabilitation and enhancement of the country’s infrastructure was an essential component of the government’s policy strategy. Accordingly, the government’s program envisaged sharp increases in capital expenditure and large primary budget deficits, which would be largely financed through capital markets. The primary deficits were expected to reverse rapidly to a primary surplus as (1) the infrastructure rehabilitation would ignite private-sector-led growth with a correspondingly expanding tax base, (2) the strengthening of the administrative infrastructure would foster fast recovery of the tax administration and its revenue-collection capacity, and (3) the increases in capital expenditure would be transitory and reversed upon completion of the infrastructure rehabilitation. Therefore, it was anticipated that the upper limit of the debt-to-GDP ratio would be reached soon during the reconstruction.

A policy strategy that involves borrowing in anticipation of higher, future permanent income is reasonable during a reconstruction period, particularly for the government, which should suffer less from financial market imperfections than private households and firms. As pointed out by Barro (1979) and others, borrowing allows for tax smoothing, that is, the temporary debt financing of spending during a period of revenue below its long-term average. In the context of Lebanon’s reconstruction, debt financing has allowed for significant capital expenditure and for the gradual rehabilitation of the revenue system and administration.

During 1991–97, the government succeeded in rehabilitating a significant part of the country’s infrastructure, and the reconstruction program has been one of the forces underlying the rapid economic growth during the period. Managing the fiscal implications of the reconstruction, however, has been a major challenge in the context of the stabilization policy and given problems associated with larger-than-expected current expenditure and a slower-than-anticipated recovery of the revenue administration and collection capacity. Under these circumstances, the anticipated rapid reversal in the primary budget balance did not materialize and budget deficits remained large–varying between 9 percent of GDP in 1993 and about 26 percent in 1997. Consequently, net public debt increased from a low of 38 percent of GDP at the end of 1993 to about 97 percent at the end of 1997.

In light of these large deficits and the associated rapid increase in public debt, questions about the implications of the current fiscal policy stance on the debt dynamics and macroeconomic stability in Lebanon have arisen. The delay in the adjustment of the primary balance has increased the adjustment needed to stabilize the debt dynamics as depicted in the stylized scenario in Figure 3.1. The delay not only shifts the debt path upward but also changes its slope as a result of the increasing interest burden incurred by the growing debt. A fiscal adjustment path that would have been sufficient to ensure stabilization of the debt dynamics earlier can become inadequate later.

Figure 3.1.Fiscal Shocks and Delayed Adjustment

(In percent of GDP)

Source: IMF staff calculations.

1 The panels show the simulated outcome of a fiscal adjustment program covering 20 years that is delayed for a period of 5 years. In the left panel, the same adjustment path for the primary balance is simulated for each year between 1993 and 1997. These paths reflect the envisaged adjustment in the primary balance. The actual primary balance is assumed to remain at a deficit of 10 percent of GDP during the period 1993–97. In the right panel, the implications of the primary deficit path on the public debt are shown.

The government has recognized the challenges posed by the delayed adjustment in the primary government balance and the growth in the public debt. With the 1997 budget, which envisaged a reduction in the primary budget balance by about 7 percentage points from the 7.6 percent of GDP registered in 1996, another attempt was made toward fiscal adjustment. The actual outcome, however, was once more disappointing as the budget deficit in percent of GDP increased instead of decreasing and the rapid accumulation of public debt continued. The medium-term adjustment to ensure fiscal sustainability thus became a major policy issue in late 1997 when the 1998 budget was being prepared.

In 1998, the government made a determined attempt to turn around the fiscal situation and stabilize the public debt dynamics. Learning from past problems, the government included in the budget, for the first time since 1995, not only expenditure but also revenue measures so as to initiate the adjustment needed to ensure medium-term fiscal sustainability. Based on information about the actual budgetary outcome during the first half of the year, chances for achieving the budgetary targets of 1998 are good. Doing so would prepare the ground for successful fiscal adjustment in the medium term.

Lebanon’s Public Finances—An Overview

For many years, large budget deficits and a rapidly growing public debt have been the most important macroeconomic problems faced by the authorities. For a full appreciation of the current policy issues, the origins and evolution of the current fiscal problems need to be ascertained. As shown below, these problems can be traced to the fiscal disarray that resulted from the war years and the delays in implementing the planned fiscal adjustment after 1993.

Fiscal Policies and Budgetary Performance, 1972–97

Before the war, the government had a limited role in the economy.1 The country was well known for its conservative fiscal policies with low expenditure and low tax ratios, in a macroeconomic environment characterized by a stable currency, low inflation, sustained economic growth, and overall balance of payments surpluses. The civil war has had a profound impact on the level and structure of public finances in Lebanon, and, consequently, large fiscal deficits have become a chronic problem. Since the end of the war, the reconstruction needs coupled with a weak initial administrative capacity have only compounded the fiscal challenges that Lebanon has been facing.

The War Legacy

The war had four main effects on public finances. First, the general breakdown in government authority had a dramatic effect on the revenue-collection capability, and revenue plummeted from levels of more than 15 percent of GDP to levels that were generally below 10 percent (Table 3.1). The effects of the loss of control over the ports were particularly important because customs revenue was the main source of tax revenue.2 Second, the stagnant and irregular economic activity reinforced the decrease in collected revenue. Third, the accelerating inflation led to an erosion of real revenue as a result of the tax collection lags (often referred to as the Tanzi effect). Fourth, government expenditure, as a percentage of GDP, rose above 20 percent on account of the authorities’ attempt to maintain a minimum of public-services and social services–for example, through subsidies on various commodities and services or through transfers, higher military expenditures, and increased interest payments. As a result, sizable budget deficits emerged, which were to a significant extent financed by the central bank. At the end of the war, public finances and their administration were in a state of disarray: the administrative infrastructure of the ministry of finance was reduced to a bare minimum; the revenue structure was unbalanced in its reliance on a few indirect tax and nontax revenue sources; existing taxes were complicated and difficult to administer; and the newly formed government inherited a substantial amount of public debt (the net public debt amounted to 87 percent of GDP at the end of 1990).

Table 3.1.Government Operations1
Selected
Prewar YearsSelected War YearsPostwar Years
1972197419751976198019851988198919901991199219931994199519961997
(In percent of GDP)
Revenue12.115.610.73.713.77.31.86.89.715.912.014.114.616.817.316.4
Tax revenue8.011.57.52.09.52.00.61.42.15.45.49.39.411.114.012.6
Indirect taxes5.68.96.41.97.00.90.10.30.23.24.97.57.69.512.410.7
Of which: customs duties2.42.61.10.22.41.10.51.10.12.43.45.05.27.38.07.5
Direct taxes2.42.61.10.22.41.10.51.11.92.20.51.81.81.61.62.0
Nontax revenue4.14.13.11.74.35.31.25.57.610.56.64.85.25.73.33.8
Expenditure15.415.013.615.727.143.119.239.139.428.923.423.435.135.237.942.2
Current expenditures12.112.47.612.620.939.517.836.737.825.121.820.025.825.729.433.6
Wages and salaries8.57.44.57.210.69.06.99.911.210.411.110.8
Other current10.812.44.510.914.010.37.94.23.33.84.47.4
Interest payments1.59.95.911.310.85.05.56.09.710.413.014.8
Domestic1.59.85.811.210.34.94.85.79.69.712.114.1
Foreign0.10.10.10.10.50.00.70.20.10.70.90.7
Electricité du Liban fuel subsidy9.82.97.42.30.81.50.01.61.21.00.7
Capital expenditures3.32.66.13.16.33.61.42.41.73.91.33.49.39.48.58.6
Overall balance (excluding grants)-3.30.6-3.0-12.0-13.4-35.8-17.4-32.3-29.8-13.1-11.4-9.2-20.5-18.4-20.6-25.8
Financing3.3-0.63.012.013.435.817.433.326.615.710.39.820.518.624.226.8
Foreign4.74.88.1-0.4-0.2-0.10.00.0-1.12.67.84.94.23.5
Grants4.74.84.10.00.00.01.53.30.40.30.3
Domestic0.035.114.736.218.033.426.615.711.47.212.713.620.023.4
Banking system0.035.111.831.012.628.522.74.411.15.76.15.614.417.2
Banque du Liban0.219.64.92.9-7.59.916.9-10.3-7.0-0.9-10.1-0.2-5.911.8
Commercial banks-0.215.56.928.120.018.55.814.718.16.618.25.720.35.4
Nonbank private2.95.25.44.93.911.30.31.54.58.15.76.2
Sources: Ministry of Finance; Banque du Liban (BdL); Council for Development and Reconstruction (CDR):IMF staff estimates.

Includes the treasury and the foreign and domestically financed CDR capital expenditure. Excludes other public entities except for budgetary transfers. See Table A6 in the Statistical Appendix for details of public finance data.

Sources: Ministry of Finance; Banque du Liban (BdL); Council for Development and Reconstruction (CDR):IMF staff estimates.

Includes the treasury and the foreign and domestically financed CDR capital expenditure. Excludes other public entities except for budgetary transfers. See Table A6 in the Statistical Appendix for details of public finance data.

Postwar Normalization, 1991–92

After the reestablishment of a government of national unity at the end of 1990, the fiscal situation in 1991 and 1992 improved markedly for several reasons (Figure 3.2). First, revenue recovered as the government regained control over revenue sources, particularly with respect to customs revenue and nontax revenue. Second, the rapid growth resulting from the normalization of economic activity reinforced the revenue increase. Third, the elimination of war-related expenditures and an expenditure restraint (including a hiring freeze) led to a decline of expenditure in terms of GDP. As a result, the overall fiscal deficit, as a percent of GDP, dropped from 30 percent in 1990 to 13 percent in 1991 and to 11 percent in 1992.

Figure 3.2.Public Finances

Sources: Data provided by Lebanese authorities; and IMF staff estimates.

1 Excluding interest payments.

2 The structure of tax revenue and current expenditure excluding interest payments is not entirely comparable before and after 1993 as a result of a reclassification of budget items. In 1995, following the tariff reform, the share of indirect taxes on goods and services declined as a result of the inclusion of excises on imports in customs revenue. Another reclassification of revenue items, affecting the shares of other indirect taxes and nontax revenue, became effective in 1996.

3 Excluding interest payments and wages and salaries.

Revenue Developments and Policies, 1993–97

In contrast with the period 1991–92, during which revenue increases were primarily the result of the reestablished government authority over revenue sources and the normalization of economic activity, the authorities embarked on a number of reforms to mobilize revenue during the period 1993–97. Reflecting the favorable revenue effects of major reforms introduced during this period and the improvement in the tax administration, Lebanon’s tax ratio (total tax revenue as a percentage of GDP) increased from 5.4 percent in 1992 to 12.6 percent in 1997 (Table 3.2), and tax revenue registered a buoyancy ratio of 1.8 during 1993–97. The principal reform measures included the following steps.

Table 3.2.Revenue in Percent of GDP and Buoyancies of Major Categories of Taxes and Other Revenue 1
Buoyancy
199319941995199619971993–951993–961993–97
(In percent of GDP)
Taxes on income, profits, and property1.81.81.61.62.00.70.71.3
Taxes on goods and services22.52.52.12.02.30.50.41.0
Duties and taxes on imports25.05.27.38.07.52.72.62.2
In percent of imports9.19.313.214.43.65.3
Other taxes32.50.8
Total tax revenue9.39.411.114.012.61.72.41.8
Nontax revenue34.85.25.73.33.81.70.10.5
Total revenue14.114.616.817.316.41.81.61.4
Sources: Ministry of Finance; and IMF staff calculations.

See Table A6 in the Statistical Appendix for details of the revenue data.

With the 1995 tariff reform, the principal excise and some other taxes on goods and services became part of customs duties and are recorded in duties and taxes on imports.

Under the revised budget classification scheme of 1996, some revenue, such as fiscal stamp duties, which were classified as nontax revenue until 1995, are now included in the new item “other taxes.”

Sources: Ministry of Finance; and IMF staff calculations.

See Table A6 in the Statistical Appendix for details of the revenue data.

With the 1995 tariff reform, the principal excise and some other taxes on goods and services became part of customs duties and are recorded in duties and taxes on imports.

Under the revised budget classification scheme of 1996, some revenue, such as fiscal stamp duties, which were classified as nontax revenue until 1995, are now included in the new item “other taxes.”

Duties and taxes on imports. This category of taxes for the period 1993–97 registered the highest buoyancy ratio, 2.2, reflecting the favorable revenue effects of the 1995 tariff reform and the temporary increase in imports required for reconstruction. The reform measures included the introduction of a minimum tariff rate of 2 percent, the reduction of the number of tariff rates and the number of tariff lines, the consolidation of various taxes collected by different ministries at the customs into a single, unified tariff, and the inclusion of excise taxes on imported goods in the tariff. They also included the elimination of the so-called customs exchange rate for valuation purposes, which was overvalued compared with market exchange rates.3 Following these reforms, customs revenues in nominal terms almost doubled between 1994 and 1995, despite the relative decline of the share of imports in GDP. However, the inclusion of excises and various other taxes and fees in the tariff has led to some decline in revenue from taxes on goods and services and nontax revenue. Therefore, the overall revenue effect should be assessed using overall tax revenue rather than customs revenue alone.

Taxes on income and profits. This tax category, despite some progression in nominal rates, registered a buoyancy ratio of less than unity until 1996, reflecting the major income tax reform of 1993. The reform included the following measures: (1) the reduction in the nominal tax rates from 26 percent to 10 percent on corporate profits and from 15 percent to 5 percent on dividends and on other corporate distributions; (2) the reduction of the top marginal rates of individual income taxes from 32 percent to 10 percent for wages and salaries, and from 50 percent to 10 percent for individual business profits; (3) the adoption of an amnesty program that added 9,700 taxpayers; and (4) accelerating the payment of taxes withheld at the source by having the withholders release the funds quarterly (before the reform, these taxes were paid by withholders at the time of filing the declaration). The reduction of the nominal tax rates was intended to encourage the flow of international capital and direct investments and improve the voluntary compliance by taxpayers. In 1997, after significant strengthening of the tax administration, the revenue yield of their tax category improved and the buoyancy ratio reached a value of 1.3.

Taxes on goods and services. During the period 1993–95, the excise rates on tobacco and cigarettes were raised from 5 percent to 30 percent, and the prices of petroleum products were also raised a number of times to narrow the gap with the international prices. The taxation of real estate transactions also yielded increasing revenue given the real estate boom of 1993–95. Despite these developments, this category of taxes registered the lowest buoyancy ratio (1.0) during the period 1993–97. This can be attributed to the collection of principal excises at the stage of importing and recording of their revenue with that from customs duties, as explained earlier.

Analyzing the evolution of the structure of revenue over the period 1993–97 shows that the improvements in the tax revenue mobilization are primarily the result of the customs reform of 1995 (Figure 3.2). The decline in the share of taxes on income and profits until 1996 reflected the combination of the revenue effects of the 1993 reform and weaknesses in taxpayer compliance and tax collection (see below). The reform of income taxation has therefore not yet resulted in the expected increased yield even with the improvements registered in 1997. Today, Lebanon’s revenue structure is unbalanced, as it is relying heavily on three sources: imports, some excisable goods, and a plethora of taxable public-services and administrative fees. Most domestic transactions and a large part of income are not effectively subject to taxation.

The ministry of finance recognizes the need to enhance its administrative capacity and has embarked on an ambitious project of modernizing and computerizing the budget process and revenue administration. Regarding the latter, the creation of a database on taxpayers, particularly for taxes on income and profits, has been of utmost importance.4 The modernization project in the ministry’s revenue department has included a countrywide campaign to enforce taxpayer registration, the buildup of auditing capacity and expertise, and the streamlining and automatization of tax administration procedures. These measures have the potential to increase substantially the yield of revenue from taxes on income and profits over the next few years.

Expenditure Developments and Policies, 1993–97

A sharp increase in government expenditure has been the predominant characteristic of public finances during the period 1993–97. Government expenditure rose from 23 percent of GDP in 1993 to 42 percent in 1997 (Figure 3.2 and Table 3.1), with all expenditure categories, albeit to a different degree, contributing to the increase.

Current expenditure (excluding interest payments). This expenditure category consists primarily of salaries and wages, purchases of goods and services, and transfers and subsidies. The government has, in general, succeeded in containing the overall wage bill through wage freezes despite considerable political pressure to compensate government employees for the accumulated loss of purchasing power on wages during and after the civil war. In late 1995, a retroactive wage increase was granted (from January 1, 1995 on ward).5 On a cash basis, only one-third of the retroactive increase was effectively disbursed in 1995; the full effect of the wage increase only materialized in 1996. Since then, the government has adopted the policy that further wage increases will only be granted if they are covered by compensating revenue measures. Subsidies to public enterprises have remained important, reflecting the slow cost-recovery capability experienced by providers of basic public services. The ongoing conflict with Israel has also led to recurrent transfers to people and institutions in the south of Lebanon.

Interest payments. Interest payments have been a major factor underlying the growth dynamics of expenditure. Since 1991, the deficit has been financed almost entirely by issuing debt, largely denominated in domestic currency, to institutions and agents other than the central bank. In the context of the government’s exchange-rate-based nominal anchor policy, and given the significant domestic and regional political risks, nominal interest rates on domestic currency assets have been high and subject to dramatic adjustments. Together with the rapid accumulation of public debt, this has led to sharp increases in budgetary interest payments.

Capital expenditure. After the launch of the Horizon 2000 program, the government embarked on a large number of infrastructure rehabilitation and enhancement projects, and capital expenditure rose from 3.4 percent of GDP in 1993 to 9.3 percent in 1994 and 9.4 percent in 1995. In the preparation of the program, the government had envisaged spreading the capital expenditure in nominal terms equally across the 12-year planning period. Capital expenditure as a percentage of GDP would therefore decline over time, leading to some built-in fiscal adjustment. This effect was already visible in 1996 and 1997, when with almost unchanged levels in nominal terms, capital expenditure fell to 8.5 and 8.6 percent of GDP, respectively.

Fiscal Policies and Public Debt Dynamics, 1990–97

Since 1991, large budget deficits have been largely financed through issuing debt, particularly in the form of short-term treasury bills denominated in Lebanese pounds. During 1991–93, the ratio of net public debt to GDP decreased despite very sizable primary deficits and the nominal depreciation of the Lebanese pound on account of the rapid growth and the negative real interest rates on the public debt (Table 3.3,Figure 3.3).6 The negative real interest rates were associated with the substantial inflation during 1991–93 and the favorable financial market sentiments after the appointment of Prime Minister Hariri in the fall of 1992. In 1994, the restrictive monetary policy stance implied by the exchange-rate-based nominal anchor policy led to a single-digit inflation rate and real interest rates on the public debt turned positive. Moreover, political uncertainties were associated with pressures on the exchange rate peg, and the central bank had to increase domestic interest rates to prevent excessive reserve losses. Accordingly, the favorable interest rate dynamics reversed, and net public debt started to grow. From a low of 38 percent of GDP at the end of 1993, it increased to an estimated 97 percent of GDP at the end of 1997. Given the persistent primary deficits and the high interest rates, Lebanon’s fiscal situation has raised concern about the sustainability of the public debt dynamics.

Table 3.3.Public Debt and Public Debt Dynamics 1
19901991199219931994199519961997
(In percent of GDP)
Gross public debt98.466.249.048.569.478.198.9102.7
Domestic75.254.044.044.261.166.584.486.5
External23.212.35.34.38.311.514.116.2
Government deposits11.58.010.910.617.215.019.06.1
Net public debt86.958.338.137.952.263.179.996.6
Discounted net public debt347.555.565.075.8
Change in net public debt-28.6-20.2-0.214.210.916.917.0
Contribution of:3
Primary deficit8.15.93.310.88.07.611.1
Domestic interest factor-35.0-26.1-5.54.92.15.05.9
External interest factor-12.1-6.3-1.1-0.5-0.5-0.4-0.8
Exchange rate valuation0.55.8-0.2-0.1-0.2-0.3-0.2
Discrepancy9.90.43.5-0.81.64.71.1
Memorandum items
Implied real interest rate on net public debt-38.1-78.3-3.021.912.914.412.3
Growth adjusted-76.3-82.8-10.013.96.410.48.3
Implied interest factor on net public debt0.60.50.91.11.01.11.0
Implied interest factor on external debt0.50.50.50.90.91.00.9
Growth factor2.12.31.41.21.21.11.1
Exchange rate factor1.02.10.91.01.01.01.0
Discount factor20.90.90.80.8
Sources: Banque du Liban, various publications: IMF, International Financial Statistics; and IMF staff estimates and calculations.

See Table A7 in the Statistical Appendix for details of the public debt.

See footnote 23 in the appendix for details of the calculations.

The decomposition of the changes in the net public debt (as a percent of GDP) in this table and in Figure 3.2 is based on the following identity:

where b denotes the outstanding net public debt in percent of GDP; pb is the primary balance in percent of GDP:r represents the real interest rate on net government debt; y is real GDP: S stands for the end-of-period nominal exchange rate; f for the foreign currency debt in percent of GDP; π is the inflation rate; d is a discrepancy; t is a time index; and a hat over a variable denotes a percentage change. The second term is referred to as the interest factor in this table and Figure 3.2, while the third term is the exchange rate valuation factor. The discrepancy arises as a result of including accrued interest in the data on outstanding long-term public debt, a significant share of treasury bills in total public debt, cash accounting of interest payments in the budget, the exchange rate valuation of foreign currency related debt flows during the year, and deficiencies in the recording of expenditure by public institutions not included in the central government budget. The real interest rate on the public debt in period t is calculated as the ratio of interest payment in periodt over the end-of-period stock of net debt in period t-1 minus the current period inflation rate.

Sources: Banque du Liban, various publications: IMF, International Financial Statistics; and IMF staff estimates and calculations.

See Table A7 in the Statistical Appendix for details of the public debt.

See footnote 23 in the appendix for details of the calculations.

The decomposition of the changes in the net public debt (as a percent of GDP) in this table and in Figure 3.2 is based on the following identity:

where b denotes the outstanding net public debt in percent of GDP; pb is the primary balance in percent of GDP:r represents the real interest rate on net government debt; y is real GDP: S stands for the end-of-period nominal exchange rate; f for the foreign currency debt in percent of GDP; π is the inflation rate; d is a discrepancy; t is a time index; and a hat over a variable denotes a percentage change. The second term is referred to as the interest factor in this table and Figure 3.2, while the third term is the exchange rate valuation factor. The discrepancy arises as a result of including accrued interest in the data on outstanding long-term public debt, a significant share of treasury bills in total public debt, cash accounting of interest payments in the budget, the exchange rate valuation of foreign currency related debt flows during the year, and deficiencies in the recording of expenditure by public institutions not included in the central government budget. The real interest rate on the public debt in period t is calculated as the ratio of interest payment in periodt over the end-of-period stock of net debt in period t-1 minus the current period inflation rate.

Figure 3.3.Public Debt

(In percent of GDP unless noted otherwise)

Sources: Data provided by the Lebanese authorities; and IMF staff estimates and calculations.

1 See text for details of the calculations.

Fiscal Adjustment and the 1998 Budget

In light of the large deviations of the actual budget deficits and the actual debt path from targets that had been registered from 1994 onward, the Lebanese government began recognizing the need for additional fiscal adjustment beyond the measures described above. In 1996 and 1997, the fiscal adjustment was mainly underpinned by expenditure restraint and increased revenues resulting from economic growth and improved tax administration (Table 3.4). This strategy proved to be unsuccessful, however. Expenditure restraint was difficult, partly because of the large accumulated carryovers in capital expenditure approved in previous budgets. Once approved, these expenditures could be spent at any time and were difficult to control. Similarly, given the debt and interest rate dynamics, interest payments also turned out to be difficult to predict and control. Revenue as a percent of GDP improved but not enough to meet the ambitious budget targets. The problems associated with controlling the fiscal situation became obvious during 1997, when the annual deficit target in nominal terms was almost reached by the end of July. To contain the deviation from budgetary targets for the remainder of 1997, the authorities enacted a number of revenue measures: excise duties on automobiles were increased in July; the government’s share of revenues collected from cellular phone calls was increased effective August 1; and customs duties and prices on tobacco products were increased in August. These measures are estimated to have generated revenues amounting to LL 250 billion on an annual basis (1.1 percent of GDP) and close to LL 100 billion in 1997 (0.4 percent of GDP).

Table 3.4.Budget Targets and Actuals 1(In billions of Lebanese pounds)
199619971998
BudgetActualsJan.–June2BudgetActualsJan.–JuneBudgetJan.–June
Revenue4,0223,5341,6304,1003,7531,7974,6002,293
Of which: customs duties1,8001,6327691,8001,7227732,190977
Expenditure6,4587,2252,9896,4339,1623,6837,9203,528
Of which: interest expenditure2,2502,4687972,7003,3781,5583,2001,583
Overall balance3-2,436-3,691-1,359-2,333-5,409-187-3,320-1,235
Primary balance4-186-1,223-562367-2,031-329-120348
Memorandum items
Overall balance3
In percent of expenditure-37.7-51.1-45.5-36.3-59.0-51.2-41.9-35.0
In percent of GDP-11.5-18.1-9.8-23.6-12.8
GDP21,12220,41723.84022,87825,946
Sources: Ministry of Finance; Banque du Liban.

Includes the treasury and the domestically financed capital expenditure by the Council for Development and Reconstruction (CDR). Foreign financed CDR expenditures are excluded to faciliate the comparison with monthly data published by the authorities. See Table A6 in the Statistical Appendix for details of public finance data.

Cumulation based on monthly data.

Excluding foreign grants

Excluding foreign financed capital expenditures.

Sources: Ministry of Finance; Banque du Liban.

Includes the treasury and the domestically financed capital expenditure by the Council for Development and Reconstruction (CDR). Foreign financed CDR expenditures are excluded to faciliate the comparison with monthly data published by the authorities. See Table A6 in the Statistical Appendix for details of public finance data.

Cumulation based on monthly data.

Excluding foreign grants

Excluding foreign financed capital expenditures.

In the preparation for the 1998 budget, the government decided to address the fiscal problems more forcefully by relying on both revenue and expenditure measures. The overall target was to initiate front-loaded fiscal adjustment that would make a significant contribution to stabilizing the debt dynamics. On the revenue side, the budget incorporated a 2 percentage point increase in import duties on virtually all imports and a new 5 percent service tax on hotels and restaurants; both measures are seen as a first step to the introduction of a general sales tax (GST). Other revenue measures included (1) the simplification of and increases in road-user taxes; (2) increases in the yearly tax on private cargo trucks and the cement tax; (3) the introduction of or increases in other fees and charges (on residence permits, passport issuance and deed renewal). On the expenditure side, the budget envisaged reducing the number of school teachers and professors,7 as well as contractual employees in the ministry of information, while increasing the overall expenditure on social sectors; incorporated a provision to cancel all uncommitted carryover from the year 1995 and earlier; and included for the first time in the budget presentation all nonbudget treasury expenditures to ensure that actual cash outlays would be in line with budget projections. These measures, taking into account the envisaged capital expenditures and interest payments on public debt are projected to reduce the overall budget deficit to about 15 percent of GDP and the primary deficit to about 2.2 percent of GDP.

The budgetary outcomes for the first half of 1998 confirm the government’s more determined efforts to achieve fiscal adjustment. In terms of expenditure, the overall budget balance during January–June 1998 amounted to 35 percent, well below the out comes observed in 1996 and 1997.8 Moreover, excluding foreign-financed capital expenditure, the primary balance was positive, implying some repayment of net domestic public debt. In April 1998, the government enacted further revenue increases in anticipation of a forthcoming wage increase effective in the beginning of 1999. This wage increase had been agreed upon in March after government wages had been frozen since the beginning of 1996. Despite strong political pressures for an immediate increase, the increase was delayed, and any retroactivity was made conditional on further revenue measures beyond those enacted in April. Compensatory revenue measures put into effect at the end of April included a 2 percentage point increase in tariffs on all imports and a 10 percentage point increase in taxes on cigarettes and tobacco. These increases provided further guarantees for the 1998 budget targets to be met.

Theoretical Aspects of Fiscal Sustainability

Overall, the fiscal adjustment envisaged by the government aims at achieving fiscal sustainability. In the remainder of this section, the government’s adjustment plans are reviewed in light of this notion. According to Razin (1996), a fiscal policy program is considered to be sustainable if it meets the solvency requirements,9 can be continued into the indefinite future, is consistent with high and sustained medium-term economic growth, and is not prone to an abrupt and discrete change when the economy is hit by a shock. 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.

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. Ideally, the degree of adjustment needed to ensure the sustainability of fiscal policies should be evaluated on the basis of a comprehensive macroeconometric model. For Lebanon, as for many others, such models are not available, and a simpler approach is needed. To evaluate the level and structure of Lebanon’s fiscal adjustment need, two steps are proposed: first, assess the level of the adjustment needed to stabilize the debt dynamics; second, ascertain the structure of adjustment needed in expenditure and revenue.

The Level of Adjustment Needed to Stabilize the Debt Dynamics

When assessing of the level of adjustment needed to stabilize the debt dynamics, so-called primary gap measures are useful tools.10 The T-period primary gap is defined as the difference between the current primary deficit and the average annual primary balance needed to achieve a certain target debt-to-GDP ratiob in T years ahead. A positive primary gap implies that the primary balance needs to be increased to achieve the debt target. If the current primary balance pb is assumed to remain constant throughout the period t to t+T. the T-period primary gap GAPT is defined as:11

where rg denotes the growth adjusted real interest rate,b is the debt-to-GDP ratio, and t stands for a time index.12 The intuition underlying the gap-formula in equation (1) is as follows: The first term on the right hand side captures the difference between the current debt-to-GDP ratio and the net present value of the target value of the debt-to-GDP ratio on an annual basis. The difference is the annual value for the primary balance needed to achieve the debt target. The second term provides the net present value of the current primary balance, which is assumed to prevail throughout period t to T, also on an annual basis. The first term minus the second term in equation (1) then equals the gap, that is, the difference between the average primary balance needed and the planned primary balance. If the primary balance under current policies is not constant during the period t to T, the gap measure has to be rewritten as:

In practice, fiscal consolidation tends to be gradual since administrative, economic and political factors constrain the adjustment that can be achieved on a year-by-year basis. If such a gradual adjustment in the primary balance could be incorporated in the primary gap measures, the latter would provide a useful benchmark for assessing actual medium-term fiscal adjustment plans. Assuming a typical, smooth adjustment path, the gradual convergence of the primary balance to its medium-term target be modeled as a first-order autoregressive (AR(1)) process:

where ρ is a measure for speed of the gradual adjustment. The speed with which the primary balance approaches its targeted medium-term level is inversely related to the coefficientρ. In this case, the primary gap would be defined as:

The application of primary gap measures is typically a difficult undertaking since data on variables, which are linked through various channels, are required for an analysis based on empirically validated relationships. For example, the growth-adjusted real interest rate, the debt-to-GDP ratio, and the fiscal adjustment path are all linked through the crowding out mechanism: An increase in the debt-to-GDP ratio leads, ceteris paribus, to an increased absorption of private savings, which requires higher real interest rates for the bond market to clear. This increase in real interest rates decreases the medium-term growth rate through its negative impact on investment, which, in turn, reduces the growth path of the capital stock and thus potential output.13 Any implementation of primary gap measures should take these channels into account. It is exactly at this stage that a fully specified macroeconometric model would be useful.

In the absence of such a model, a simpler approach is required. Frequently, a constant growth-adjusted real interest rate is used. Some authors have used a parameterized interest rate function, which relates the real interest rate to the debt-to-GDP ratio.14 The latter approach is associated with considerable uncertainty, given that econometric estimates of the relationship between real interest rates and the level of government debt provide, at best, reduced-form parameters. If one is interested, for policy analysis, in determining a path for the primary balance that would allow a country to reach a certain target debt-to-GDP ratio, the application of constant growth-adjusted real interest rates based on historical averages or a priori limits, or both, based on the experience of other countries appear preferable.

Another difficulty arising in this context is the choice of the target debt-to-GDP ratio. Theoretical considerations suggest to opt for a low ratio in light of the crowding out mechanism. Moreover, lower values for the debt-to-GDP ratio facilitate the macro-economic management in general since they reduce the problems associated with the refinancing of maturing debt and the incentives to reduce the real value of the government debt through surprise inflation. The experience of many countries suggests that debt-to-GDP ratios above 100 percent to 120 percent of GDP can lead to difficulties in macroeconomic policy management if they are maintained over a long period.15

The Structure of the Adjustment

Achieving a sustainable fiscal policy program typically involves adjustments not only in the level of the primary deficit but also in the structure of expenditure and revenue. The latter are often the result of the expenditure and revenue measures enacted to reduce the fiscal imbalance. In general, these measures should aim at increasing government saving, reducing expenditure for services that can be provided by the private sector or which are unproductive, better targeting of recipients of transfers and subsidies, and minimizing the distortionary effects of taxes.

Regarding expenditure adjustments, recent research on Fiscal policy and growth has provided some indications on expenditure components that tend to support growth. On the expenditure side, adjustment should be such that government expenditures that are essential for enhancing growth are preserved at their optimal level.16 These expenditures include capital expenditure and expenditure on health and education, all of which contribute to the accumulation of physical and human capital. Unfortunately, the optimal level of these expenditures is difficult to ascertain for two reasons. First, it depends on the specific structures of the growth model used, in particular the specification of the production function.17 Second, reliable empirical estimates for key parameters of aggregate production and consumption functions are often not available. Moreover, the optimal expenditure level is also a function of the tax system if taxation has distortionary effects. A reconstruction economy faces not only the problem of determining the optimal level of infrastructure capital, but also that of the optimal convergence to this level from a public capital stock destroyed by war. The determination of the optimal adjustment path is subject to caveats that are similar to those regarding the optimal level. However, from a perspective of maximizing long-run growth, standard “turnpike” considerations suggest that it would be optimal to move to the long-run equilibrium level of the capital stock as fast as possible. In the end, it is likely that macroeconomic policy constraints, fiscal sustainability considerations, and the project implementation capacity will determine the path of capital and other growth-enhancing expenditure during a reconstruction.

Concerning revenue, recent research has recommended adjustments in revenue structure such that the yield on taxes is constant or increasing with GDP growth to allow for tax smoothing. Moreover, the tax system should be balanced so that tax revenue is not overly dependent on one source. In this way, the system is less prone to shocks to the tax base of a particular tax category. Recent research on Fiscal policy and growth has also emphasized the role of broad-based consumption taxes as a means to raise revenue without imposing excessive distortions on an economy.

Fiscal Adjustment and Aspects of Fiscal Sustainability

With the 1998 budget, the government has made determined efforts to embark on a front-loaded fiscal adjustment program that aims at ensuring medium-term fiscal sustainability. As discussed above, a program aiming at fiscal sustainability should be reviewed in the light of two fundamental questions. First, is the envisaged adjustment in the primary deficit sufficiently large to ensure that the debt dynamics is consistent with solvency and medium-term macroeconomic stability? Second, is the adjustment in the structure of expenditure and revenue consistent with growth and reducing their vulnerability to exogenous shocks and structural changes in the economy?

The Adjustment Need Implied by Primary Gap Measures

The computation of primary gap measures for Lebanon is difficult because some of the basic macroeconomic relationships needed in the calculation of empirically based gaps cannot be estimated because of missing data or structural breaks, or both. For this reason, two sets of a priori primary gap measures, for different values of the growth-adjusted real interest rate and the target debt-to-GDP ratio, are computed. The first set comprises 10-year primary gap measures (Table 3.5, left panel). The 10-year primary gap is the difference between the average, annual primary balance needed to reach a target net debt-to-GDP ratio at the end of year 10 and the current primary deficit, that is, 11.1 percent of GDP in 1997. The second set consists of 15-year primary gap measures, which are also based on the 1997 primary deficit of 11.1 percent of GDP (Table 3.5, right panel). For the growth-adjusted real interest rates, values of −5 percent to 10 percent were selected. While standard economic theory suggests that the growth-adjusted real interest rate is strictly positive in the long run, the possibility that it is close to zero or even negative during a sharp growth acceleration at some phases of reconstruction cannot be excluded. For the target net-debt-to-GDP ratio, values ranging from 60 percent—the Maastricht criteria—to 150 percent, which certainly is an upper limit to a debt-to-GDP ratio that is sustainable in the medium term are used in the calculations.

Table 3.5.Primary Gap Measures 1(In percent of GDP)
Ten-Year Primary GapsFifteen-Year Primary Gaps
Target NetGrowth-adjustedGrowth-adjusted
Debt-to-GDPreal interest rate (in percent)real interest rate (in percent)
Ratio-50510-50510
605.79.513.618.05.28.912.817.2
655.19.013.217.74.88.512.617.0
704.58.512.817.44.38.212.416.9
753.98.012.417.13.87.912.116.7
803.27.512.016.83.47.511.916.6
852.67.011.616.52.97.211.716.4
902.06.511.216.22.46.911.416.2
951.46.010.815.82.06.511.216.1
1000.75.510.415.51.56.211.015.9
1050.15.010.015.21.15.910.815.8
110-0.54.59.614.90.65.510.515.6
115-1.14.09.214.60.15.210.315.5
120-1.73.58.814.3-0.34.910.115.3
125-2.43.08.414.0-0.84.59.815.1
130-3.02.58.113.7-1.34.29.615.0
135-3.62.07.713.3-1.73.99.414.8
140-4.21.57.313.0-2.23.59.114.7
145-4.91.06.912.7-2.73.28.914.5
150-5.50.56.512.4-3.12.98.714.4
Source: IMF staff calculations based on data provided by the Lebanese authorities.

See text for details of the calculations, which are based on the primary deficit recorded in 1997.

Source: IMF staff calculations based on data provided by the Lebanese authorities.

See text for details of the calculations, which are based on the primary deficit recorded in 1997.

For the understanding of the implications of the primary gaps, it is useful to start with the case of 95 percent, that is, a stabilization of the net debt around the end-of-1997 level. The 10-year primary gap at a zero growth-adjusted real interest rate amounts to 9.5 percent. Hence, a stabilization of the debt dynamics at the current level requires achieving a primary deficit of about 1.6 percent of GDP over the entire 10 years provided that the interest cost of the debt grows in line with GDP. However, if the real interest rate exceeds the growth rate of GDP by a margin of 5 percentage points, additional adjustment of about 4.8 percentage points of GDP a year is required. In other words, an annual primary surplus of 3.2 percent of GDP over the next 10 years is needed to maintain a stable debt-to-GDP ratio. The 15-year primary gap amounts to 10.0 percent of GDP in the case of a zero growth-adjusted real interest rate, implying a primary deficit of 1.1 percent of GDP to achieve the debt target. With a growth-adjusted real interest rate of 5 percent, however, the 15-year primary gap measure implies that a primary surplus of 3.6 percent of GDP is needed for a debt target of 95 percent of GDP to be realized.

The figures in Table 3.5 show that the primary gaps are positively correlated with the growth-adjusted real interest rate, reflecting the larger interest rate costs of the debt stock. The correlation with the target net-debt-to-GDP ratio is negative because a larger target value relative to the initial value requires less fiscal adjustment. A comparison of the first panel in Table 3.5 with the second panel shows that the correlations for the 10-year primary gap vary more in both dimensions than in the case of the 15-year primary gap. This larger variation is the result of two mechanisms. If the target net-debt-to-GDP is lower than the debt-to-GDP ratio in the initial period, the fiscal adjustment required to achieve the lower target within a shorter time span is larger. For a target ratio that is larger than the current ratio, less fiscal adjustment is required within the 10-year time span than within the 15-year time span because the accumulated interest cost of the debt is smaller.

Overall, the primary gap measures shown in Table 3.5 indicate that the primary deficit target in the 1998 budget is not yet consistent with the stabilization of the ratio of net-public debt to GDP at a value close to the end-of-1997 level as long as the growth-adjusted real interest rates remain larger than zero. This result, however, should be interpreted with caution as the primary gap measures analyzed so far are based on the assumption that the adjustment in the primary balance is undertaken in one step and maintained throughout the 10- or 15-year period. Given the dramatic improvement in the primary balance needed in the case of Lebanon, it would be unreasonable to assume that such an adjustment could be undertaken in one step. A gradual primary balance adjustment as described by the autoregressive process in equation (3), however, could be feasible in light of the discussion of expenditure and revenue reforms following below. The computation of primary gap measures according to equation (4), therefore, provides one with useful benchmarks for comparison.

Fifteen-year primary gap measures that already incorporate some fiscal adjustment according to equation (3) indicate additional adjustment needs for some parameters combinations (Table 3.6). To illustrate the adjustment, the time path of the primary balance for a period of 15 years is shown in Figure 3.4 for values of the adjustment coefficientρ ranging from 0.9 to 0.1 and for a medium-term primary surplus target of 2 percent of GDP. For example, starting from a primary deficit of 11.1 percent in 1997, an adjustment coefficient of 0.7 would imply that a zero primary balance would be reached in 2003 while a coefficient of 0.5 would lead to zero primary balance in 2000. Under the assumptions of a zero growth-adjusted real interest rate and positive medium-term primary surplus targets, the primary gaps reported in Table 3.6 show that ambitious adjustment paths, that is, paths that aim at reducing the primary deficit by at least 5 percentage points of GDP in the first year (equivalent to values ofρ of 0.5 or less), are consistent with a target net-debt-to-GDP ratio of 80 percent of GDP. With a 5 percent growth-adjusted real interest rate, the primary gap measures indicate a need for further adjustment to attain the same target debt-to-GDP ratio.

Table 3.6.Fifteen-Year Primary Gaps with Autoregressive(l)-Adjustment 1(In percent of GDP)
AR-Adjustment CoefficientTarget Long-Run Primary Balance
-2.0-1.5-1.0-0.50.00.51.01.52.02.53.0
Target net-debt-to-GDP ratio: 100 percent; growth-adjusted real interest rate: 0 percent
0.93.93.12.41.60.90.2-0.6-1.3-2.0-2.8-3.5
0.70.80.2-0.4-0.9-1.5-2.1-2.7-3.3-3.8-4.4-5.0
0.5-0.1-0.6-1.1-1.7-2.2-2.7-3.3-3.8-4.3-4.9-5.4
0.3-0.4-0.9-1.5-2.0-2.5-3.0-3.5-4.0-4.5-5.1-5.6
0.1-0.6-1.1-1.6-2.1-2.6-3.1-3.6-4.2-4.7-5.2-5.7
Target net-debt-to-GDP ratio: 80 percent; growth-adjusted real interest rate:0 percent
0.96.55.85.14.33.62.82.11.40.6-0.1-0.9
0.73.52.92.31.71.10.60.0-0.6-1.2-1.7-2.3
0.52.62.11.51.00.5-0.1-0.6-1.1-1.7-2.2-2.7
0.32.21.71.20.70.2-0.30.8-1.4-1.9-2.4-2.9
0.12.01.51.00.50.0-0.5-1.0-1.5-2.0-2.5-3.0
Target net-debt-to-GDP ratio: 100 percent; growth-adjusted real interest rate: 5 percent
0.98.37.66.86.15.34.63.83.12.31.50.8
0.75.24.64.03.42.82.31.71.10.5-0.1-0.7
0.54.23.73.22.62.11.51.00.5-0.1-0.6-1.2
0.33.83.32.82.21.71.20.70.2-0.4-0.9-1.4
0.13.53.02.52.01.51.00.50.0-0.5-1.0-1.5
Target net-debt-to-GDP ratio: 80 percent; growth-adjusted real interest rate: 5 percent
0.910.39.68.88.17.36.65.35.14.33.52.8
0.77.26.66.05.44.84.33.73.12.51.91.3
0.56.25.75.24.64.13.53.02.41.91.40.8
0.35.85.34.84.23.73.22.72.21.61.10.6
0.15.55.04.54.03.53.02.52.01.51.00.5
Source: IMF staff calculations bused on data provided by the authorities.

See text for details of the calculations, which are based on the primary deficit recorded in 1997.

Source: IMF staff calculations bused on data provided by the authorities.

See text for details of the calculations, which are based on the primary deficit recorded in 1997.

Figure 3.4.Adjustment Path of Primary Balance

(In percent of GDP)

Source: IMF staff calculations.

What are the implications of these primary gap measures for Lebanon? In the 1998 budget, the targeted primary deficit amounts to 2.4 percentange of GDP with the envisaged front-loaded adjustment. For the years thereafter, the government plans to achieve primary surpluses. This adjustment path implies low values (below 0.4) for the adjustment coefficientρ discussed above. As evident from Table 3.6, such an adjustment path would be consistent with stabilizing the net public debt at the end-1997 level for growth-adjusted real interest rates in the range from zero to 5 percent.

Adjustments in the Structure of Expenditure and Revenue

There are also issues regarding the structure of expenditure and revenue during fiscal adjustment. For expenditure, the questions are whether capital expenditure can and should be adjusted and whether reductions in other expenditures could contribute to adjustment. For revenue, the questions are whether the current structure is consistent with a constant or even rising yield with respect to GDP growth and what effects the ongoing modernization of the tax administration will have on the revenue yield of income taxes.

Reconstruction and Capital Expenditure

As discussed earlier, government capital expenditure plays an important role in growth. The optimal level of capital expenditure for Lebanon is difficult to determine given the lack of empirical evidence. A cross-country comparison shows that Lebanon’s capital expenditure as a percentage of GDP has indeed been high by international standards (Figure 3.5). Over the period 1991–95, the average capital expenditure in a sample of 25 industrial and developing countries amounted to 3.7 percent of GDP (median: 2.7 percent) while that of Lebanon reached 9.1 percent. In a sub sample of 16 developing countries, the average level of capital expenditure was about 4.8 percent of GDP over the same period (median: 4.1 percent). A comparison over a longer period (1981–95) yields an average of 4.5 percent of GDP for all countries (median: 3.4 percent), while it was 6.1 percent of GDP for the developing countries in the sample (median: 4.2 percent).18 This comparison shows that substantial fiscal adjustment in the order of 3 percent to 5 percent of GDP can be expected from capital expenditure in the medium term once it starts to converge to internationally comparable levels following with the gradual completion of the reconstruction program. In the 1998 budget, a planned decrease in capital expenditure by 2.4 percentage points of GDP contributes indeed to the deficit reduction.

Figure 3.5.Cross-Country Comparison of Capital Expenditure

Sources: IMF, Government Finance Statistics and International Financial Statistics

1 For some countries, the sample period covers only parts of 1991’95.

2 Average 1994’96.

3 In percent of GDP as shown on right scale.

4 For some countries, the sample period covers only parts of 1981’95.

Other Expenditure

The reduction of other expenditure is also difficult, particularly in the short term. In 1997, about 76 percent of current expenditure was accounted for by expenditure that was subject to short-term rigidities: payments for wages and salaries and interest payments. The contribution to fiscal adjustment from these two expenditure items in the medium term depends on progress in fiscal adjustment in the interim (interest payments) and on the pace of civil service reform, which could be a delicate issue in the current sociopolitical environment. While civil service reform could imply substantial short-term costs, its contribution could be substantial in the medium term. In the short term, restraints on the aggregate wage bill could provide important support to fiscal consolidation efforts. Indeed, a newly unchanged nominal aggregate wage bill contributes about 1.2 percentage points of GDP to the overall adjustment. Other current expenditure, including purchases of goods and services for current consumption and transfers (including social spending) account for about 20 percent of expenditure. Reductions in this area could be feasible but are likely to be limited by the need to improve the social safety net and the recurrent expenditure implied by the reconstruction program. Greater private sector participation and cost recovery through user fees (electricity, water) would contribute to reducing the budgetary burden associated with such recurrent expenditure in the future.

Revenue Structure and Modernization of Tax Administration

Lebanon’s fiscal consolidation efforts need to involve measures to increase the ratio of revenue to GDP as well as to improve the current revenue structure, which is very dependent on revenue from customs duties as discussed earlier. Currently, most domestically produced goods and services are not subject to indirect taxes. Moreover, the ratio of income tax revenue to GDP has been low, and the income tax base has not yet been expanding in line with private sector growth. Therefore, without further revenue measures as a result of the expected structural changes (accelerated growth in domestic production, gradual decline in external imbalances) and shocks, the overall tax revenue, as a percent of GDP, might fall rather than increase in the future

Revenue measures should aim at expanding the tax base to incorporate the most dynamic sectors, on both the demand and income side. On the demand side, a broad-based consumption tax, possibly a general sales tax, could expand the tax base considerably. In 1996, it was estimated that imports amounted to about 50 percent of GDP while private consumption was about 100 percent of GDP. A broad-based consumption tax covering the domestically produced goods and services, therefore, has the potential to almost double the tax base compared with the current base, which only covers the imported part of consumption. Such a tax could also allow the authorities to lower the average import tariff rate and to participate in trade liberalization efforts. On the income side, continued efforts at improving the capacity of the income tax administration and expanding the base of registered taxpayers could also yield additional revenue given the anticipated high private sector growth in the medium term.

Conclusions

In recent years, Lebanon’s sizable budget deficits and its large and growing public debt have raised the issue of fiscal sustainability. Rough calculations using primary gap measures based on the 1997 fiscal outcome confirm a need for substantial fiscal adjustment in the medium term to stabilize the ratio of public debt to GDP at levels consistent with macro-economic and financial stability. If the debt-to-GDP ratio is to be maintained at current levels or to be reduced over the medium term, surpluses in the primary budget balance would have to be run over the years to come. The size of the surpluses critically depends on the growth-adjusted real interest rate. A high GDP growth rate would facilitate the task of fiscal adjustment considerably since the growth-adjusted real interest rate is quite likely to be inversely related to GDP growth.

The timing of fiscal consolidation should be such that front-loaded fiscal adjustment is achieved as further delay would only increase the adjustment need in the future. The 1998 budget target is indeed aimed at converging to such an adjustment path, as a significant drop in the primary budget deficit is envisaged. In 1999 or 2000, a primary surplus is targeted. Such front-loaded adjustment also has the advantage that it puts less strain on the overall macro-economic policy mix and is likely to support a decline in growth-adjusted real interest rates as it contributes to the credibility of the planned fiscal policy path.

Fiscal adjustment in Lebanon will have to incorporate efforts to generate additional revenue as (1) further reductions in capital expenditure are likely to materialize only in the medium term, (2) a large share of the current expenditures is either subject to short-term rigidities (e.g., interest expenditure), (3) other current spending (excluding interest payments and wages) is small and its further reduction would entail substantial losses in terms of social spending and infrastructure maintenance, and (4) further increases in the tax buoyancy are unlikely given the recent decline in imports as a share of GDP. Moreover, it would also be advantageous to achieve a more balanced tax structure so as to lower the vulnerability to shocks and avoid distortions.

Further revenue measures, including cost-recovery measures related to public infrastructure services, are also consistent with tax-smoothing considerations. Tax-smoothing principles would suggest that the maturity of the debt issued to finance reconstruction projects should be equal to the lifetime of the project and that taxes for the debt service should be levied during the entire lifetime of the project. Accordingly, some cost recovery through taxation or levy of user fees, or both, should already be carried out in the beginning of reconstruction.

Revenue measures could focus on two areas. First, a general sales tax and the extension of excises levied on imports to domestically produced goods would ensure that the yield of taxes on goods and services would be linked to the expected acceleration in the domestic production of goods and services. Second, continued efforts in strengthening the tax administration and the database on registered taxpayers would help increase the returns from taxes on income and property in the medium term. With the expected accelerated private sector growth, these measures would also contribute to ensuring that the overall tax yield is consistent with future sources of growth in the economy.

Regarding the sequencing of measures, revenue measures should be implement as soon as possible given the rigidities on the expenditure side discussed above. A tight expenditure stance on the aggregate wage bill and other current noninterest expenditure should support the revenue measures. To ensure the front-loaded adjustment, the revenue measures could include temporary measures, such as a surcharge applied to the least distortionary revenue sources (e.g., on a general sales tax or excises), which would be in effect until the medium-term benefits of a civil service reform, the convergence of capital expenditure to their medium-term equilibrium level, and the effects of tax reform would materialize. The recent revenue measures enacted by the government are consistent with these recommendations.

Appendix. Intertemporal Government Budget Constraint and Solvency

The concept of government solvency defines a minimum requirement for sustainability. While it is inappropriate for Lebanon under the current circumstances, the concept is nevertheless useful for understanding the broader concept of fiscal sustainability.19 In the simplest possible form, its essentials can be summarized as follows.20 In each period, the government’s flow budget constraint implies that under the assumption of pure debt financing the end-of-period public debt is given by:

where B denotes the outstanding public debt, TR is government revenue,E represents noninterest government expenditure. R is the nominal interest rate on government liabilities, and t is a time subscript. For the subsequent analysis, it is useful to rewrite the flow constraint in equation (6) in terms of the current period GDP:

where a lowercase letter variable denotes a capital letter variable as a fraction of the current period GDP (except for r, the real interest rate), and ý the growth rate of real GDP. The intertemporal solvency constraint requires that the flow budget constraint in equation (6) is expected to hold in every period in the future, which leads to the condition:

where rg denotes the growth-adjusted real interest rate, (that is. 1 + r)/(l+ ý) - 1. The sum of the expected present value of all present and future expenditure and the current level of debt has therefore to be equal to the net present value of current and future revenue. The condition in equation (7) can be rewritten using the primary balance pb rather than expenditure and revenue:

If the growth-adjusted real interest rate is positive on average, as suggested by standard models of economic growth, the condition in equation (8) leads to the familiar requirement that the government has to run primary surpluses in the future if it has some outstanding liabilities today. In the derivation of equations (7) and (8), it was assumed that the so-called transversality condition:

holds. This condition can be interpreted as a limit on the average increase in the debt-to-GDP ratio in the future, which has to be lower than the average growth-adjusted real interest rate.21 Together, the solvency constraints in equations (7) or (8) and the transversality condition in equation (9) ensure the intertemporal consistency of a fiscal program. They provide the accounting framework for the requirements that the debt is serviced in every period and will eventually be repaid. In other words, they ensure that the government’s net worth on a present value basis is positive.

The conditions in equations (7) or (8) and (9) have constituted the core of many empirical studies on fiscal sustainability.22 In some studies, the focus lies on testing whether the primary deficit or the debt-to-GDP ratio, or the discounted debt-to-GDP ratio are stationary over a sufficiently long time period.23 As discussed below, stationarity of these variables is necessary for solvency but not a sufficient condition for fiscal sustainability.24 In other studies, some arbitrary steady-state, debt-to-GDP ratio is defined, which is then used to assess the sustainability of the current fiscal policy on the basis of the current level of debt and average growth rates of expenditure, revenue, interest rates, and GDP growth.

In many circumstances, the concept of fiscal solvency is not terribly useful for policy analysis for the following reasons:

  • While it seems reasonable and pragmatic to require that the debt-to-GDP ratio be a stationary time series in a very large data sample, the requirement is weak in that it is consistent with almost any positive mean value for this variable. Even a shift in the long-run debt-to-GDP value from, say, 40 percent to 100 percent, does not violate the solvency conditions. From a general macroeconomic perspective, however, such changes would not be minor, since their implications for growth and macroeconomic policies are likely to be substantial.

  • In many countries, the time span covered by the data sample is insufficient to allow for a meaningful distinction between stationary and nonstationary time series for the primary deficit and the debt-to-GDP ratio.

  • Macroeconomic variables, such as the real interest rate and growth, are usually taken as given in the derivation and testing of equations (3) and (4). However, fiscal policies, in particular if they are perceived to violate solvency conditions, have repercussions on financial market prices and growth. The solvency and sustainability of certain fiscal policies, therefore, can not be assessed without taking into account macroeconomic policies in general as well as the interaction between macroeconomic variables and policies. For example, if a government embarks on a fiscal program that, in the perception of financial markets, implies an explosive path for the government debt, financial market participants would require higher and higher interest rates to be compensated for the risks and would eventually refuse to acquire or hold government bonds. The government would then be forced to abandon its unsustainable program.

  • Another problem of the interaction of fiscal policies and other macroeconomic variables is that the range of fiscal policies consistent with solvency can depend on the overall macroeconomic policy mix. For example, the exchange rate regime, which, through its implications for monetary policy, can narrow the set of fiscal sustainable policies, as often demonstrated by many episodes of debt-related currency crises in the case of fixed exchange rates.

For Lebanon, the problems associated with the notion of fiscal solvency are particularly relevant. It is obvious that increases of the debt-to-GDP ratio of 8 to 10 percentage points or more a year would result if the large primary deficits during 1991–97 were maintained for a long time and if the growth-adjusted real interest rate were positive. Over time, such a policy would be perceived to violate the conditions in equations (3) and (4) and would thus become unsustainable, as investors would refuse to purchase government bonds. However, since these large primary deficits are temporary during a reconstruction, the standard methods of assessing fiscal solvency are impractical. For Lebanon, the principal issue in the assessment of the sustainability of its fiscal policies is the degree of adjustment—defined as the reduction in the primary deficit—that is needed to ensure that a manageable debt-to-GDP ratio level can be maintained over the medium term.

See Saidi (1989) and Eken (1995 Section III) on fiscal developments in Lebanon since 1974.

See Makdisi (1987) for a detailed discussion.

The adjustment of the customs valuation exchange rate to prevailing market rates was combined with a proportional reduction in tariff rates. At the time of the adjustment, the net revenue effect of the two measures was zero. Over lime, however, the introduction of a market-based customs valuation exchange rate has ensured that customs revenue will not be eroded by inflation.

Prewar taxpayer records are outdated, and the gap between registered and potential taxpayers has become large. It is estimated that less than half of all households and firms that would have to file for income and profit taxes under current tax laws are registered. Moreover, about 95 percent of all income and profit tax revenue in 1996 and before was based on declared income only as the auditing and administrative capacity of the revenue department at the ministry of finance was limited.

For the lowest wage categories, the increase amounted to 20 percent. Middle- and upper-wage categories received increases between 10 percent and 20 percent.

Throughout the section, the concept of net public debt is used. Net public debt is defined as total public debt minus government deposits with the banking system. The latter have emerged, beyond normal levels implied by seasonal revenue fluctuations, as a result of sterilization operations of the central bank that served both monetary policy and debt management purposes (see Section V). While these operations imply quasi-fiscal costs given the positive interest rate differential with respect to foreign currency assets, they do not create any net debt liabilities for the treasury and are therefore ignored here.

The reduction was intended to reduce over-staffing and was not expected to affect the quality of services delivered.

In their budget presentation and monthly reporting, the Lebanese authorities do not include foreign financed capital expenditure, that is, capital expenditure that is directly financed by nonresidents in the context of the reconstruction program. To allow for a comparison, the format of the budget presentiation in Table 3.4 excludes this expenditure category but it is included in all other fiscal tables presented in this paper.

Solvency requires that the sum of the expected present value of all present and future expenditure and the current level of debt has to be equal to the net present value of current and future revenue. The solvency concept and its implications for fiscal sustainability are discussed in the appendix.

See Blanchard (1990) and Buiter (1997) for a detailed discussion.

The following exposition abstracts from valuation problems associated with foreign currency debt, given the forward-looking perspective. Table 3.3 shows that, with the exception of 1992, valuation changes have not contributed much to debt dynamics in the recent past.

The growth-adjusted real interest rate is defined as rg = (1+r)/(1+ý)-1. where ý denotes the growth rate of real GDP.

The strength of these effects depends on whether Ricardian equivalence holds. At the limit, under conditions of full Ricardian equivalence, they would be nonexistent. See Barro (1974) and Bernheim (1987). among others, on this issue.

See Ford and Laxton (1995), Helbling and Wescott (1995), Mongelli (1996), and Tanzi and Fanizza (1995). among others, on the relationship between real interests rates and government debt. In all these studies, the real interest rate rather than the growth adjusted real interest rate is used as a dependent variable.

See, for example, Dornbusch and Draghi (1990) and Obstfeld (1994).

Tanzi and Zee (1996) discuss the linkages between the budgetary structure and long-run growth from the perspective of both recent empirical and theoretical research.

See Barro and Sala-i-Martin (1992,1995) on public-finance related aspects of growth.

This sample was chosen on the basis of data availability. Data on capital expenditure were taken from the IMF’s Government Finance Statistics, while the nominal GDP data were taken from the IMF’s International Financial Statistics.

See Buiter (1985, 1997) and Blanchard (1990) for a more detailed discussion.

This exposition abstracts from valuation problems associated with foreign currency debt. Table 3.3 shows that, with the exception of 1992, valuation changes have not contributed much to the debt dynamics.

Note that the transversality condition is irrelevant if the real interest rate is, on average, lower than the GDP growth rate. However, if this case were relevant, the economy could be dynamically inefficient (see Abel and others, 1989).

See Hamilton and Flavin (1986), Wilcox (1989), and Buiter and Patel (1992). among others.

The discounted public debt is defined by the continuous application of the transversality condition in equation (5), starting from the first data point in the sample. In Table 3.3 the discounted net public debt for Lebanon is shown. Since the transversality condition in equation (5) is only relevant if the real interests rate exceeds the growth rate, it is only reported for the period 1994–97,

Note, however, that the debt-to-GDP ratio does not need to be stationary to satisfy the transversality condition, which only limits its growth rate.

References

    AbelAndrew1989Assessing Dynamic Efficiency: Theory and Evidence,Review of Economic Studies. Vol. 56 (January) pp. 9120.

    BarroRobert J.1974Are Government Bonds Net Wealth.Journal of Political EconomyVol. 82 (December) pp. 10951117.

    BarroRobert J.1979On the Determination of the Public Debt,Journal of Political Economy Vol. 87 (October) pp. 94071.

    BarroRobert J. and XavierSala-i-Martin. 1992.Public Finance in Models of Economic Growth,Review of Economic Studies Vol. 59 (October) pp. 64561.

    BarroRobert J. and XavierSala-i-Martin. 1995Economic Growth (New York: McGraw-Hill).

    BernheimB. Douglas1987Ricardian Equivalence: An Evaluation of Theory and Evidence,NBER Working Paper 2330 (Cambridge Massachusetts: National Bureau of Economic Research). pp. 263303.

    BlanchardOlivier J.1990Suggestions for a New Set of Fiscal Indicators,OECD Working Paper No. 79 (Paris: OECD).

    BlanchardOlivier J.Jean-Claude P. RobertChouraquiHagemannNicolaSartor1990The Sustainability of Fiscal Policy: New Answers to an Old Question.OECD Economic Studies. No. 15(Autumn) pp. 736.

    BuiterWillem1985A Guide to Public Sector Debt and Deficits,Economic Policy Vol. 1 (November) pp. 1379.

    BuiterWillem1997Aspects of Fiscal Performance in Some Transition Economies Under Fund-Supported Programs,IMF Working Paper9731 (Washington: International Monetary Fund).

    BuiterWillemUrjitR. Patel1992Debt, Deficits, and Inflation: An Application to the Public Finances of India,Journal of Public Economics Vol 47 (March) pp. 171205.

    DornbuschRudigerMarioDraghi. 1990.Public Debt Management: Theory and History (Cambridge; New York: Cambridge University Press).

    EkenSenaPaulCashin. Nuri ErbasS. JoseMartelino. AdnanMazarei.1995Economic Dislocation and Recovery in Lebanon IMF Occasional Paper No 120 (Washington:International Monetary Fund).

    FordRobertDouglasLaxton. 1995World Public Debt and Real Interest Rates,IMF Working Paper 95/30 (Washington: International Monetary Fund).

    HamiltonJames D.FlavinMarjorie A.1986.On the Limitations of Government Borrowing: A Framework for Empirical Testing,American Economic Review Vol 76 (September) pp. 80819.

    HelblingThomasRobertWescott1995The Global Real Interest Rate,Staff Studies for the World Economic Outlook World Economic and Financial Surveys (Washington: International Monetary Fund) pp. 127.

    MakdisiSamir A.1987Political Conflict and Economic Performance in Lebanon, 1975-1987.Bulletin Trimestriel. Banque du Liban (Second and Third Quarters) pp. 412.

    MongelliFrancesco P.1916The Effects of the European Economic and Monetary Union (EMU) on National Fiscal Sustainability,IMF Working Paper 96/72 (Washington: International Monetary Fund).

    ObstfeldMaurice1994The Logic of Currency Crises.NBER Working Paper No. 4640 (Cambridge, Massachusetts: National Bureau of Economic Research)

    RazinAssaf1996.Notes on Fiscal and External Sustainability” (unpublished; Washington: International Monetary Fund).

    SaidiNasser1989. “Deficits Inflation and Depreciation: Lebanon’s Experience 1964-88” in Politics and the Economy of Lebanon ed. by Nadim Shebadi and Bridget Harney (London: The Center for Lebanese Studies, Oxford University, and the School of Oriental and African Studies).

    TanziVitoDomenicoFanizza1995Fiscal Deficit and Public Debt in Industrial Countries, 1970-94,IMF Working Paper 95/49 (Washington: International Monetary Fund).

    TanziVitoZeeHowell H.1996Fiscal Policy and Long-Run Growth.IMF Working Paper 96/119 (Washington: International Monetary Fund).

    WilcoxDavid W.1989The Sustain liability of Government Deficits: Implications of the Present-Value Borrowing Constraint,Journal of Money Credit and Banking Vol 21 (August) pp. 291306.

    Other Resources Citing This Publication