Pakistan: Selected Issues and Statistical Appendix

This Selected Issues and Statistical Appendix paper presents cross-country regression results that identify investment in physical capital and improvements in institutional quality as having the largest pay-off in terms of increased growth. The paper employs three approaches to forecast inflation in Pakistan. A leading indicator model outperforms a univariate autoregressive moving average model as well as a vector autoregressive model in terms of forecast quality. The paper presents three case studies of Pakistani public sector enterprises that have recently witnessed strong improvements in their financial performance.

Abstract

This Selected Issues and Statistical Appendix paper presents cross-country regression results that identify investment in physical capital and improvements in institutional quality as having the largest pay-off in terms of increased growth. The paper employs three approaches to forecast inflation in Pakistan. A leading indicator model outperforms a univariate autoregressive moving average model as well as a vector autoregressive model in terms of forecast quality. The paper presents three case studies of Pakistani public sector enterprises that have recently witnessed strong improvements in their financial performance.

IV. Revenue Generation in Pakistan: Performance, Policies, and Prospects16

A. Introduction

67. Faced with a high incidence of poverty and large public debt, domestic resource mobilization is at the heart of Pakistan’s reform agenda. Pakistan’s social indicators lag behind most countries in the region, reflecting in part relatively low social expenditure, and a range of compounding socio-political factors. A low revenue ratio restricts the fiscal room for maneuver, and revenue collection is constrained by a small taxpayer base and administrative shortcomings.

68. In order to successfully meet the authorities’ social and poverty alleviation objectives, revenues, in real terms, need to increase sizably in the short and medium term. In this context, Pakistan’s Poverty Reduction and Strategy Paper (PRSP), published in late 2003, envisions the tax-to-GDP ratio to increase by 1 percentage point of GDP in five years. The revenues gained from the enhanced effort, as well as lower interest expenditures in the coming years, are expected to create the fiscal space necessary to reach the government’s poverty, social, and development spending objectives.

69. The PRSP also lays out a strategic reform framework that builds on recent initiatives and takes additional steps to raise revenue. A number of initiatives, through tax policy and administration reforms, have already been undertaken in recent years to improve revenue collection. This paper examines the past trends and performances to obtain a set of useful lessons for the road ahead, examining in particular:

  • The recent history of revenue measures in Pakistan;

  • Pakistan’s revenue ratio against a group of comparable countries;

  • The gap, if any, between past revenue generating measures, targets, and outcomes, and lessons learnt; and

  • A realistic revenue target for the medium term and a roadmap of policy and administrative reforms to get there.

70. Since taxes account for more than three-fourth of budgetary revenues, and the Central Board of Revenue (CBR) is directly or indirectly responsible for collecting virtually all federal revenues, the papers analysis focuses on CBR-related reforms and performances.

71. In general, the paper finds that Pakistan’s tax system, while still yielding low levels of revenue, appears to be moving in the right direction. Following the implementation of several key policy and administrative reforms in recent years, the structure of taxes has improved, the taxpayer environment has become friendlier, and there are some indications of increased efficiency of the tax system. Also, tax collection by the CBR in the four core tax categories have shown a modest upward trend of late. The authorities, however, still need to ensure that the reform momentum is maintained, including by addressing still-persisting administrative inefficiencies and shortcomings, and by implementing some further improvement in tax policy by broadening the tax base, which remains rather narrow.

B. Recent Steps

72. The tax policy and administrative reforms in recent years have aimed at creating a fairer and friendlier tax environment, as well as widening the tax net. These include:

  • Rationalization of both the personal income tax rate and the corporate tax rate;

  • Revision of tax laws to enforce uniformity of tax treatment (e.g., between residents and nonresidents, banks, and public limited companies);

  • Introduction of self-assessment for filing personal income and corporate tax;

  • Levy of withholding tax on stock market transactions;

  • Refinement on the presumptive taxation scheme; and

  • Expansion of general sales tax (GST) on agriculture inputs.

Also, customs and excise duties on a large number of items have been rationalized, and various income tax, duty, and GST exemptions have been removed. On the administrative front, the CBR has opened large- and medium-taxpayer units and a customs house toward taxpayer facilitation. Efforts are underway to revamp the human resource management of CBR and to implement functional integration of various CBR units to improve the tax information base.

73. Many of these measures were corrective steps to remove distortions from the tax system, but not necessarily revenue enhancing in nature. Moreover, tax administration measures aimed at taxpayer facilitation may not immediately provide for substantial revenue gains, although they make the tax environment more conducive to compliance. Steps that aim at base broadening, increased tax administration efficiency and better implementation skills, would ultimately pave the way for a buoyant revenue environment.

C. Tax Composition, Trends, and Comparisons

Pakistan’s key tax revenue sources are direct tax (on personal and business income), consumption taxes (sales tax and federal excise duty), and customs duties. The CBR is directly responsible for collection and administration of these taxes.17 Following a number of trade liberalization measures in the mid- and late-1990s, Pakistan’s structure of taxation has changed considerably, with a gradual reduction in the dependence on foreign trade taxes (collected through customs) and a concurrent increase in GST and direct tax collection. These changes indicate a move toward a modern tax structure.

Figure IV.1
Figure IV.1

Pakistan: Major Tax Components

(1993/94-2003/04, as a share of GDP)

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Source: Central Board of Revenue.

75. Direct taxes as a share of total CBR collection have increased over the past decade, from about 25 percent in the early 1990s to 32 percent in 2003/04. A number of measures, including the introduction of self-assessment, removal of a large number of income tax exemptions, levying withholding tax on some targeted goods and transactions, several tax amnesty schemes, and widening the scope of dividend taxation, were implemented during this period. Some revenue reducing measures, principally cuts in corporate income tax rates have also been put in place in recent years. The increase in overall CBR share notwithstanding, direct tax receipts as a share of GDP have been basically stagnant since the mid-1990s, hovering at around 3 percent. 18

76. Induced largely by trade liberalization measures, customs collection declined sharply over the past decade, with the exception of an import rebound-led bounce in the past two years. As a share of GDP, customs collections declined from 3.4 percent in 1993/94 to 1.1 percent in 2001/02. In the early 1990s, customs accounted for 35-40 percent of CBRs total collection, when high and differentiated levels of imports tariffs characterized the regime. Since then, customs duties have been reduced and rationalized on a large number of products, and some exemptions have been removed. The policy measures, along with improved administration, appears to have yielded results of late, as collections have shown a pickup in buoyancy in the last two years. During this period, on the heels of double-digit imports growth, customs duty collection improved by 0.6 percentage points of GDP, growing by 44 and 32 percent in 2002/03 and 2003/04.19

Figure IV.2
Figure IV.2

Pakistan: Imports, Customs, and GST

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Sources: IMF, Central Board of Revenue
Figure IV.3
Figure IV.3

Pakistan: Customs and Import GST

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Source: Central Board of Revenue.

77. Some of the policy-induced decline in customs receipts has been offset by increases in import-related GST. This reflects a major tax policy change implemented in the 1994/95 budget, which brought 125 imported goods under the GST net. Subsequently, some targeted rises in GST and turnover tax also boosted import-related GST collection, which accounted for a quarter of CBRs collection in 2003/04, compared to about 8 percent in 1993/94.

78. Excise duties accounted for about a fifth of CBRs collection until 1998/99, targeting levies on utility services, bank advances, and other goods and services at the point of production.20 Since then, some administrative difficulties, gradual cuts in excise duties, as well as removal of selected items from the excise net, have led to a sharp decline in collection, with excises currently comprising of only 9 percent of total CBR revenues or 0.8 percentage points of GDP.

79. The general sales tax, introduced in the value-added mode under the Sales Tax Act of 1990, is the major source of revenue in Pakistan, currently accounting for more than 42 percent of total CBR revenues. Over the past decade and a half, hundreds of previously exempt items have been brought under the GST net, while the coverage has been extended to agriculture inputs, wholesale and retail sales, and selected services. In addition, the standard rate was raised from 12.5 to 15 percent in 1993/94, turnover and further tax were levied on certain products, retailers were registered, and various efficiency enhancement measures were put in place (e.g., presumptive assessment). The result, however, has been modest, with the domestic component of GST accounting for slightly over 1.5 percentage points of GDP during the last five years. The import-component of GST thus accounts for most of the increase in GST intake in recent years.21

Figure IV.4
Figure IV.4

Tax Ratio in Percent of GDP

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Source: Central Board of Revenue

80. Considering the tax categories together, it is seen that Pakistan’s tax revenue relative to GDP has been stagnant at around 11 percent for a number of years. Indeed, five-year average tax ratios have been on a slightly declining path—from 11.4 percent of GDP during 1987-91, to about 11 percent by 2000-04. Reform initiatives in recent years have at least stabilized the ratio somewhat, and data from 2002 onward appear to indicate the beginning of a slight upward trend. In particular, after bottoming in the late 1990s, CBR revenue ratios have increased steadily.

81. How does Pakistan measure up against countries with similar characteristics? Comparisons can be made with countries in the same region, or with comparable levels of development (proxied by per capita GDP). First, Pakistan’s tax collection since 1987 is examined against regional counterparts: Bangladesh, India, Nepal, and Sri Lanka—countries with broadly similar tax policies and administration structures. The unweighted average of these four economies tax ratio over the sample period is about 1 percentage point of GDP higher than that of Pakistan’s. If Sri Lanka, which enjoys a substantially higher per capita GDP than the other South Asian economies, is taken out of the sample, the average ratio is broadly comparable to Pakistan’s.

82. Comparisons with nonregional comparators are unambiguously unfavorable. Using latest available data, a worldwide ranking of per capita GDP on a purchasing power parity basis was prepared, and tax ratios of countries that were no more than one-half standard deviation away from Pakistan in the distribution were selected as the comparator group.22 Unweighted average of the these economies tax ratios are about 7-8 percentage points of GDP higher than Pakistan’s.

Table IV.1.

Tax Revenue Against Comparators

(1987-2004)

article image
Source: IMF, based on general government data (where available) from the Government Financial Statistics, converted to calendar year basis. 2004 values are based on projectionss

Bangladesh, India, Nepal, and Sri Lanka (simple average).

Bolivia, Cameroon, Cote dIvoire, Egypt, Indonesia, Lesotho, Moldova, Mongolia, Nicaragua, Senegal, and Zimbabwe (simple average). Selection criterion is having per capita GDP (ppp) within 0.5 standard deviation of Pakistan in a worldwide ranking.

83. Looking at broader comparators, the average tax ratio of non-OECD countries, at 15.2 percent, is substantially higher than Pakistan’s. In this cohort, direct taxes and VAT revenues amount to about 4.5 and 5 percent of GDP, respectively, compared to Pakistan’s 3 percent and 4 percent, respectively. Excises average about 2 percent of GDP, more than twice that of Pakistan’s excise collection.

84. Among developing economies with Pakistan’s level of wealth, relative stagnation in revenue ratios is common (particularly in South Asia), reflecting a variety of factors associated with low-income countries, including capacity constraints and poorly designed tax systems. Pakistan’s revenue trend however looks unsatisfactory compared to a selected number of dynamic economies. For example, China, Indonesia, and Thailand show clear revenue buoyancy in recent years, as their economic growth rates have accelerated

Figure IV.5
Figure IV.5

Selected Countries: Tax Revenue Evolution, 1997-2004

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Sources: IMF; general government data.

D. Behind the Low Ratio

85. Factors underlying Pakistan’s low revenue yield are commonly associated withlow-income countries. They include problems in tax administration, including a taxpayerunfriendlyenvironment, discretionary powers of tax officials, complex tax rules, and weaksupervision of staff which undermines willingness of taxpayers to comply. A pervasive lack of ownership regarding the link between taxation and public service delivery has also contributed to tax evasion and noncompliance. A narrow and inflexible tax base hampers revenue generation, while a large informal economy continues to escape the tax net. Pervasive smuggling remains a major problem, with associated revenue losses. On the tax policy end, some tax exemptions aimed at special interest groups persist. This section focuseson some of these issues, including sales tax efficiency and the tax base, and on taxexpenditures incurred through various exemptions.

86. Pakistan’s major source of tax revenue, the GST, has been subject to a variety of reforms in recent years, but continues to yield rather low levels of revenues. Since its introduction, GST efficiency has admittedly increased, but from a very low base, and at a very slow pace. In fact, since 1999/2000, GST efficiency, in particular the domestic component, has been virtually flat.

Figure IV.6.
Figure IV.6.

Pakistan: GST Efficiency 1/

(1993/94-2003/04)

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Source: Central Board of Revenue.1/GST collection in percent of GDP divided by 15, which is the GST standard rate.

87. Pakistan’s GST efficiency is low by international standards. Indeed, of the 86 countries that currently have a VAT type sales tax system, Pakistan’s GST (or VAT) efficiency ratio is at the bottom 10 percent. Also, the ratio is the lowest in South Asia. Despite the implementation of various measures, including removal of exemptions, enhancing taxpayer facilitation, and widening the tax net, there appear to be shortcomings in Pakistan’s GST collection. Reinforcing existing policies and enhancing compliance are usual first steps to resolving this problem.

Table IV.2.

VAT Rates and Revenues

(Selected Countries)

article image
Source: IMF

Revenue efficiency = Total VAT revenue as percentage of GDP divided by the VAT standard rates.

88. With regard to direct taxes, a narrow base, noncompliance, and tax evasion are the major underlying factors behind the relatively disappointing performance. 23 The economy has a large informal, virtually nontaxed component, estimated by most studies to be at least over a quarter of GDP. Also, the service and agriculture sectors, accounting for 52 and 23 percent of GDP respectively, are taxed nominally, with tax administration efforts focusing mostly on the industrial sector.

Table IV.3.

Pakistan: Taxpayer Information 1/

(In millions, unless otherwise stated)

article image
Sources: Central Board of Revenue; World Bank (2004).

Based on 2003/04 data.

After adjusting for double-counting and invalid entries in the taxpayer database.

89. A gradual reduction in personal and corporate income taxes in recent years partly explains the lack of pickup in direct tax collection, but major gaps remain in the proper assessment and collection of income taxes, and many more taxpayers should be under the revenue net. About a third of the registered taxpayers do not file for taxes in Pakistan. Moreover, given the consumption level and pattern of goods and services that are relatively costly, conservative estimates suggest that the number of taxpayers should be at least 3 million.24

90. Despite some administrative efforts, including occasional amnesties, the number of returns filed since 1999/2000 has been stagnant, in fact declining as a share of the population. 25 National taxpayer identification numbers, introduced to enhance the collection and profiling of taxpayer information, are somewhat out of date and the database contains many errors. Automation of the taxpaying process, which would increase efficiency and allow more resources to be devoted to compliance and enforcement, is yet to be accomplished. Since 2003, the CBR has been attempting to identify additional taxpayers by exploiting the various consumer information available in its electronic database (e.g., automobile ownership, utility charges, etc.) from GST administration. To improve the quality of audits, it has also been using a software to link the unique national taxpayer identification number with the information indicating consumption patterns. Universal self-assessment for tax filing has been introduced, and the CBR is refining its system of assessment and auditing under this new environment. It remains to be seen if these efforts will bear fruit in the coming years.

91. An additional factor accounting for Pakistan’s low revenue ratio is the practice of exempting special interest groups from various taxes. In this regard, Pakistan has been moving in the right direction, with a large number of tax exemptions removed in recent years. Tax expenditures, estimated to account for nearly 0.8 percent of GDP in 2000/01, were reduced by a half through 2003/04. The authorities justify some of the exemptions as means to promote investment, exports, and growth. Nevertheless, closer examination of remaining tax exemptions should be considered with a view to reducing tax expenditures.

Table IV.4.

Pakistan: Tax Expenditures, 2000/01–2003/04

(In percent of GDP)

article image
Sources: Economic Survey, Government of Pakistan, various issues.

E. Revenue Generating Measures, Targets, and Outcomes

92. Given the mostly flat tax ratio trend, it may appear that taxes have at least grown in line with the economy in recent years. A close look at Pakistan’s tax buoyancy, measured as a ratio of nominal tax collection growth rate relative to the growth rate of nominal growth, indicates a more complex picture. Tax buoyancy has been rather volatile in the past decade and a half, although there are some signs of an upward trend in recent years. Some of the fluctuations can be explained by economic shocks, and various policy and administrative changes undertaken. It also appears, however, that until very recently, there havent been many gains from revenue measures. The volatile tax buoyancy also reflects to some extent uneven administration of taxes and policy-induced cuts in levies.26

Figure IV.7
Figure IV.7

Pakistan: Tax Buoyancy, 1990/91-2003/04

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

93. Another way to look at the gap between tax measures and budgetary outcome is the difference between initial estimate and actual collection. Encouragingly, year-end collections exceeded budgeted amounts both in 2002/03 and 2003/04, a notable reversal of earlier trends. During these two years, no downward revisions of revenue estimates were made either.

Figure IV.8
Figure IV.8

Central Board of Revenue: Ratio of Budgeted L05 -, and Actual Collection, 1990/91-2002/03

Citation: IMF Staff Country Reports 2004, 415; 10.5089/9781451830644.002.A004

Source: Central Board of Revenue

94. The years prior to the latest two years were, however, not encouraging. Looking at aggregate figures from 1990 onward, it is seen that over the past decade and a half, Pakistan’s actual revenue outturn was almost always lower than the initial budget estimates, as the measures outlined in the budget usually fell short of achieving their expected goals. There was also a consistent pattern of making downward revisions to the initial projections, as administrative shortcomings and, occasionally, economic factors, combined to force the adjustment. Among administrative shortcomings that have been responsible for this pattern, weak planning and insufficient resources in the implementation of the budget measures, inadequate information analysis of the impact of the measures, and occasional rescinding of initial steps owing to political considerations are notable.

95. An ex post examination of tax by tax outcomes over the last five years sheds further light on this issue (see Table IV.5). 27 The figures show that year-end outcomes fell short of what should have been expected under the realized growth factors and budgeted revenue measures in about half the cases in the sample. Excise collection was consistently lower than what could have been expected, accounting for the bulk of the aggregate shortfall, possibly indicating underestimates of the impact of the relief measures contained in the initial budget assumptions.

Table IV.5.

Pakistan: Ex Post Analysis of Revenue Performance, 1999/2000-2003/04

(In Pakistani rupees billion, other than growth rates)

article image
Note: Baseline defined as previous years collection, Net measures reflect CBR estimates during the budget preparation stage. Expected revenues equal the sum of revenue based on relevant nominal growth factor and net measures. Nominal growth factors are: direct tax: nominal GDP growth; sales tax (import): nominal import growth; sales tax (domestic): large manufacturing growth and inflation; customs: dutiable imports growth and inflation; excise: large manufacturing growth of selected items and inflation.Sources: Central Board of Revenue, Pakistan Economic Survey, IMF.

96. There are no clear patterns on the other tax categories. Indeed, although GST (domestic component) and direct tax collections relative to GDP have been stagnant, the outcomes have been mostly in line with the growth factors and budgeted measures. The attached table also shows that the net estimated impact of the various direct tax measures was mostly negative, partly accounting for the lack of buoyancy. Collections of customs duties were substantially short of the expected figures through 2002/03, but the situation has reversed over the last two years, indicating some pickup in buoyancy. In 2001/02 and 2002/03, import-related GST saw substantial shortfalls.

97. The ex post analysis can be a useful guide to evaluating the performance of taxes in relation to macroeconomic developments. Indeed, tax by tax performances should be seen in light of evolving economic conditions, and not only in relation to the initial estimates, as the impact of macro factors (which are seldom the same as the initial forecast), and tax measures are better understood through the course of the year.

F. The Road Ahead

98. Pakistan’s PRSP envisages raising CBR revenues by 1 percentage point of GDP between 2003/04 and 2007/08. The PRSPs fiscal policy priorities include debt reduction and increasing the volume and effectiveness of social and development expenditures over the medium term. Raising the tax ratio by at least 1 percentage point—largely through continued tax administration reform—would contribute significantly toward achieving these objectives. Implementation of the existing reform agenda and continued economic growth should allow for that target to be reached readily. Nonetheless, reaching the five year target of a tax revenue ratio of around 11.5 percent of GDP would still leave Pakistan with a lower revenue base than many comparator countries, underscoring the need for further efforts in the medium term to mobilize resources necessary to achieve its social goals. Indeed, given the favorable growth outlook and the ambitious reform agenda in place, it should be reasonable to expect greater tax buoyancy in the coming years, with tax ratios rising higher than envisaged in the PRSP.

99. Pakistan’s relatively flat revenue ratio should also be looked at in the context of a rather unstable political and economic past. The economy underwent several episodes of exogenous shocks and balance of payments difficulties through the past decade and a half, and has only very recently begun to experience accelerated growth. The last couple of years have demonstrated that the gains from a combination of favorable economic environment and reforms can be formidable. Indeed, three or five-year averages of tax buoyancy are well over one through 2003/04, and at around their highest levels in the past decade and a half.

100. Gains from reforms should become more visible provided the reform momentum is maintained. Major tariff rationalization and other tax cut measures have already been implemented, so the coming years should not necessitate many more revenue reducing adjustments. As per capita income increases, Pakistan is likely to experience a combination of buoyant consumer demand leading to rising imports and retail sales, as well as increases in taxable income. The experiences of Korea and Thailand should be instructive in this context. Both countries saw their tax-to-GDP ratio improve by over 1 percentage point during the five years when their per capita GDP (on a purchasing power basis) rose from $2,000 to $3,000. Pakistan’s PRSP envisages a comparable rise in per capita income and tax ratio over the next half decade.

101. Along the road ahead, key reform steps should include:

  • Move toward a functional organization of tax administration that relies on voluntary compliance, self-assessment, risk-based audits, and minimal face-to-face contacts between taxpayers and officials. Encouraging progress has been made in this regard in recent years, and the momentum of the reforms must be maintained.

  • Administrative and governance reforms to ensure better documentation of economic transactions, ownership of assets, and sources of income, to enable the government to implement existing tax laws more effectively as well as to widen the tax net in the future. In particular, the databank of existing economic transactions should be refined, updated, and utilized;

  • Introduce a national taxpayer database that combines the information existing in the income tax and sales tax databases. Also, functional integration of income tax and GST administration and automation of the taxpaying process in the CBR need to be advanced;

  • Streamline the tax system through steady reduction of exemptions and increasing the number of taxpayers by bringing nonfilers into the tax net;

  • Extend the GST further into the service sector, for example, in the areas of consulting, accounting, broadcasting, and medical, and law practices;

  • Assess the scope for raising the revenue yield of agriculture taxation in the context of fiscal devolution.28 This could strengthen own source revenue of lower levels of government, complementing expenditure devolution. Given the lack of success on taxation on the basis of land and income, an alternative approach could be to introduce presumptive tax on selected agriculture output. The appeal of this measure is that it would be rather be relatively easy to operationalize, given the small number of withholding agents that would carry out these transactions. Well-known social and political constraints are associated with taxing the agriculture sector, thus the focus should be on implementing simple, transparent, and gradual measures.

G. Conclusion

102. Pakistan’s tax ratio remains low by international standards, but encouragingly its tax policy and administration are moving in the right direction. Despite various shortcomings, a review of recent measures clearly point toward a tax environment that is becoming less distortionary and more conducive to taxpayers compliance. Pakistan’s tax authorities need to mobilize sufficient resources for meeting the countrys social objectives. Reinforcing the reforms already in place and taking additional steps consistent with the reform agenda in the near and medium term would assist significantly along the way.

References

  • Ebrill, Liam, Michael Keen, Jean-Paul Bodin, and Victoria Summers, 2001, The Modern VAT, (Washington, D.C.: International Monetary Fund).

  • Government of Pakistan, 2003, Accelerating Economic Growth and Reducing Poverty: The Road Ahead, Poverty Reduction Strategy Paper.

  • World Bank, 2004, Pakistan: Public Expenditure Management, Strategic Issues and Reform Agenda, Volume1, Country Report No. 25665-PK

  • United Nations Development Program, 2004, Cultural Liberty in Today’s Diverse World, Human Development Report.

16

Prepared by Taimur Baig and M.S. Lal, Member, Policy and Tax Reforms, Central Board of Revenue, Government of Pakistan.

17

The CBR branches across the country also receive gas and petroleum surcharges, amounting to about 1½ percentage points of GDP.

18

This paper uses the newly rebased GDP of Pakistan for calculating various ratios. The revised GDP data is currently available for only five years going back. For data prior to that, necessary adjustments were made to maintain continuity in the time series. The new data, owing to substantial upward revision, reduces Pakistan’s tax-to-GDP ratio by about 2% percentage points of GDP.

19

An additional contributing factor toward the large jump in customs revenues in 2002/03 was the large reduction in duty drawback rates to account for the tariff reductions implemented in the previous years.

20

Typically, excises are imposed on efficiency grounds, on goods that have low elasticity of demand (e.g., cigarettes), are easy to administer since they are levied at the point of production or importation, and may internalize negative externalities that consumption of goods such as tobacco and petroleum impose on the rest of the society. However, excises can also be regressive. For example, tobacco products tend to represent a higher proportion of spending on poorer households.

21

A further caveat is that refunds for GST at the import stage are netted from domestic GST, thus the import component of GST can be characterized as “gross” receipts, whereas the domestic component is “net” in nature.

22

Despite having comparable per capita income, Bangladesh, Nepal, and India were excluded as they had already been compared in the previous column. Including them reduces the averages by about 1 percentage point.

23

Pakistan’s direct tax collection, at 3 percent of GDP, is substantially lower than the non-OECD average of 4.5 percent.

24

After adjusting for income tax threshold and excluding labor force in the agriculture sector.

25

Slightly over 0.97 million income tax returns were filed in 1999/2000; the number was 1.05 million in 2003/04.

26

Note that in 2003/04, the decline in revenue buoyancy was driven by declines in gas and petroleum surcharges. CBR revenues, representing the core tax categories, grew in line with GDP, and exceeded the initial budget projection by 2.4 percent.

27

The analysis incorporates relevant nominal growth factors and expected impact of various budget measures (made by the CBR).

28

Agriculture tax is a provincial tax.

Pakistan: Selected Issues and Statistical Appendix
Author: International Monetary Fund