Intergovernmental transfers: an evaluation of mechanism and design of transfers in Pakistan.
Khawaja, Idrees ; ud Din, Musleh
The 7th National Finance Commission (NFC) Award has seemingly put
an end to the deadlock over revenue distribution among the constituents
of the federation in Pakistan. This paper argues that though the 7th NFC
Award's use of multiple indicator criteria for the distribution of
resources is a step forward in the right direction, the distribution
design still falls short on various counts. For example, the weight of
82 percent for the population share is on the higher side whereas the
demographic structure of the population, an important indicator of the
expenditure needs, does not figure up in the distribution design. Also,
the basis of weights assigned to the four elements of the revenue
distribution criteria is unknown and no rigorous exercise seems to have
been undertaken to determine these weights. Similarly, matching grants,
which are a key element of the distribution design elsewhere, are
altogether absent in Pakistan. Furthermore, provinces still rely on
large transfers from the centre which undermines the incentives of the
provinces to generate their own revenues. The paper emphasises that
there is a need to rethink the mechanisms for resource sharing as well
as the institutional structure of the NFC itself.
JEL Classification: 1177
Keywords: Intergovernmental Transfers
1. INTRODUCTION
Pakistan is a federal country with two constitutional tiers of the
government--the federal government and the provincial governments;
moreover there are some Federally Administered Tribal Areas (FATA) and
the State of Azad Kashmir. As in many other countries, the federal
government in Pakistan generates more revenues than its needs.
Correspondingly, the provinces generate only a small percentage of the
revenue required to meet their expenditure needs. This calls for
transfers from the federal government to enable the provinces to carry
out their functional responsibilities. The National Finance Commission
is the institution responsible for devising the revenue sharing
arrangement between the federal government and the provinces. The
Commission recommends the sharing of the federal revenue with the four
provinces namely Punjab, Sindh, Khyber Pakhtunkhwa and Balochistan. (1)
The Commission is constituted every five years and has representation
(official as well as non-governmental) of all the stakeholders. The
Commission does not have a permanent existence, however, it is allowed
adequate time to work on the Award (i.e., announcement of revenue
sharing arrangement). The provincial ministers of finance and other
experts in the areas of finance and fiscal management typically man the
NFC which is headed by the federal minister for finance. The
'Unanimity rule'--all the provinces and the federal government
must agree--is the principle that the Commission follows in making its
recommendations to the government. Several previous Commissions either
faced deadlock or were forced to adopt the sharing arrangement
prevailing at the time due to the failure to develop a consensus on any
new sharing arrangement. The 'divisible pool' i.e., the
revenue sources available for sharing is specified in the constitution
and the President can add revenue sources to the divisible pool while
notifying the establishment of the Commission. Two kinds of conflicts
have often marred the proceedings of the Commissions. One, what should
be the share of the federal government in the divisible pool? Two, what
should be the elements of the criteria to be used for sharing the
divisible pool among the provinces.
The seventh NFC Award was announced in December 2009 and became
effective on July 01, 2010. The Award is seen as a landmark in the sense
that it broke the deadlock that had constrained the National Finance
Commissions, constituted in 2001 and 2006, to announce the awards. Two
major changes contributed to ending the deadlock: reduction in the share
of the federal government in the divisible pool by 10 percentage points
and the introduction of a multiple indicator criteria (MIC) for the
distribution of the divisible pool in place of the earlier criterion
that was solely based on population. The distribution criteria
prescribed by the 7th NFC is given in Table 1. For comparison, the
criterion used by the immediately preceding Award has also been included
in the table.
Idrees Khawaja <idreeskhawaja@pide.org.pk> is Associate
Professor at the Pakistan Institute of Development Economics, Islamabad.
Musleh ud Din <muslehuddin@pide.org.pk> is Joint Director at the
Pakistan Institute of Development Economics, Islamabad.
At present the divisible pool includes the following revenue
sources:
* Personal Income Tax
* Tax on corporate income
* Wealth Tax
* Capital Value Tax
* Taxes on sales and purchase of goods
* Custom duties
* Federal Excise Duty (excluding on Gas)
The reduction in the share of the federal government in the
divisible pool has enabled the NFC to recommend transfer of greater
funds to all the provinces. Even the province of Punjab, which in the
past had shown preference for the retention of the population share
criterion, has not been a loser despite the change in the distribution
formula. The end of the deadlock coupled with the transfer of more funds
to all the provinces have earned the award almost universal
appreciation. This study aims at a critical evaluation of the 7th NFC
Award in particular and the distribution design in general. The analysis
will primarily focus upon the institutional arrangement for the
distribution of funds and the formula for the distribution of available
resources among the provinces.
2. INSTITUTIONAL ARRANGEMENT FOR DISTRIBUTION OF RESOURCES
Though the 7th NFC Award has managed to break the deadlock that
marred the proceedings of the previous two commissions, an important
question that arises is: will the institutional structure of the NFC
prevent deadlocks in the future? To answer this question it is important
to put in perspective the reasons that led to previous deadlocks and the
factors that helped to break out of the stalemate.
2.1. The Deadlock
Historically the divisible pool has been shared among the provinces
on per capita basis. However since 1996 three out of four provinces have
been demanding the inclusion of more elements in the distribution
criteria. Each province, demanded the inclusion of such elements that
would entitle it to greater transfers from the divisible pool. With
Karachi being the country's hub of business activities and the
provincial capital, Sindh demanded that revenue generation effort be
made a part of the distribution criteria. On the other hand,
Balochistan, the largest province in terms of geographic area, contended
that its cost of public service delivery was relatively high due to low
population density, therefore the element of geographic area (reflected
in inverse population density) must be included in the criteria. Both
Khyber Pakhtunkhwa and Balochistan argued that the higher poverty levels
prevailing in the two provinces required greater transfers to alleviate
their poverty. Punjab, the largest province in terms of population
housed more than 60 percent of the country's population when the
deadlock arose in 1996. The province stood to gain from the distribution
solely on per capita basis and understandably argued for the retention
of this criterion. With each province insisting on including a different
element in the distribution criteria a deadlock in the proceedings of
the Commissions was imminent especially when the Commissions followed
the 'unanimity rule'.
When the 6th NFC Award was being negotiated in 1996, objections
were raised on the distribution criteria prevailing then, but these
objections were not responded to and the Award remained pegged to the
old formula. This was mainly due to the fact that the political party
then in power at the centre drew its strength from the Punjab which was
adamant on retaining the prevailing formula. The objections that were
raised at the time of the negotiation of the 6th Award surfaced again
with such intensity later on that these caused deadlocks over the NFCs
of 2001 and 2006. Perhaps even the 6th award reflected the strength of
the political party in power at the centre and the Punjab over the
preferences of the other provinces. It is also noteworthy that the four
Commissions that failed to announce the awards (or adopted the previous
awards without any changes) were constituted during the military regimes
(1979, 1985, 2001 and 2006) which demonstrates the capability of
democratic regimes which provide a better environment for striking a
compromise in situations where interests conflict.
2.1.1. How the Deadlock Ended?
Under the 7th NFC Award, each province is to get more transfers
from the federal government than what it would have received under the
previous distribution criteria. This was the key to opening the
deadlock: with each province getting more funds, all the provinces were
willing to go along even if some structural issues remained unaddressed.
Moreover the 7th NFC Award accepted the long standing demand of the
three smaller provinces for inclusion of their preferred elements in the
distribution criteria. The 7th NFC Award managed to placate the province
of Punjab by introducing a minimal change in the weight of the
population share--82 percent, down from 100 percent. Moreover, this time
around, the smaller provinces which had been demanding a change in the
distribution criteria were a part of the ruling coalition at the centre.
Efforts to keep the otherwise fragile coalition intact would also have
played some role in putting the deadlock to an end. Finally, given the
previous failures, all the stakeholders were under pressure to resolve
the conflict. All these factors together helped to resolve the long
standing conflict over the revenue sharing arrangement. Though the 7th
NFC Award has ended the stalemate and moved forward, a key structural
issue namely the 'unanimity rule' remains unaddressed making
deadlocks in the future possible.
2.1.2. The Unanimity Rule
As mentioned earlier, the National Finance Commissions constituted
in 2001 and 2006 failed to reach a consensus over the distribution
formula and, perhaps, over the magnitude of the federal share in the
divisible pool. The problem apparently lies in 'the unanimity
rule' that the Commission follows in adopting its recommendations.
What is the solution? Will the 'majority rule' solve the
decision making problem. Perhaps not. The majority vote can avoid the
deadlock in a narrow legal or administrative perspective but this may
raise problems for the federation. Smaller provinces may complain of
being the victim of federal hegemony. Alternatively, if the smaller
provinces get together as a group in the NFC, then the federal
government or the larger province (i.e. the Punjab) may feel deprived of
their respective share in the resources. Therefore the 'unanimity
rule' under the present institutional arrangement is not a choice
but a necessity. The point is that notwithstanding the spirit of
compromise shown by the federal government and the provinces while
negotiating the 6th and the 7th NFC awards, the structure of the NFC
itself has no mechanism to prevent a deadlock. As discussed later in the
study, it is possible to address this problem by devising an appropriate
institutional mechanism.
2.2. The Need to Rethink the Institutional Arrangement for the
Distribution of Resources
The NFCs have a history of failures. Even though the constitution
requires that there be an NFC Award every five years, only 8 Awards have
been announced since independence. The 5th NFC Award, due in 1979 came
in 1990--12 years after it should have been announced; similarly the 7th
Award due in 2001 was delayed by 9 years. The Commissions were duly
constituted in the intervening periods but these failed to reach a
consensus over the recommendations. Clearly, something needs to be done
to avoid possible deadlocks in the future. In this respect, useful
insights can be gained from research on subjects like assignment of
revenue resources to different tiers of the government and determination
of weights for the different elements of the distribution criteria.
There is not much evidence to suggest that overtime the NFCs in Pakistan
have either conducted research on the questions at hand or have made
enough use of research available on the subject.
The present institutional set-up of the NFC lacking the mechanism
to prevent a deadlock situation and the need for research on the issues
involved in designing a suitable distribution mechanism call for
revisiting the institutional structure of the body which is responsible
for declaring the Award. The next section reviews the institutional
arrangements used in different countries for making resource transfers
to the subnationals with a view to drawing guidelines for devising an
appropriate institutional arrangement for resource sharing among federal
units in Pakistan.
2.3. Institutional Arrangements Used in Different Countries
Institutional arrangements used in different countries for devising
the distribution criteria and making transfers from the federal
government to the constituent units can be broadly classified into the
following three categories:
(i) Central agency (central government's ministry),
(ii) Intergovernmental Forum,
(iii) Independent Agency.
2.3.1. Central Agency
The federal government on its own takes the decision regarding the
distribution of revenue resources among the constituent units.
Typically, the office of the president or prime minister or the ministry
of home affairs or the ministry of finance assumes the sole or partial
responsibility for the fiscal transfers to the constituent units.
Countries that are relying upon a central agency to determine the amount
of transfers include Kyrgyz Republic, Tanzania, China, Italy,
Kazakhstan, Netherlands, Poland, Switzerland, Ukraine, Ghana, Zambia and
Japan. The rationale for the central agency is that as the federal
government is responsible for managing the national objectives,
therefore the transfer decisions should be taken by the federal
government. However, this approach negates the essence of
decentralisation. Shah (2007) suggests that the constitutional
restrictions on the ability of the federal government to override
provincial preferences can limit the negative effects of this approach.
Shah further suggests that as an alternative to the federal
government's direct role in the distribution of federal revenues, a
separate body could be entrusted the task of designing the fiscal
relations among the various tiers of the government. The proposed body
could either be independent or an intergovernmental forum or may be an
intergovernmental-cum-civil society forum. Pakistan has such a forum. It
is appointed by the President every five years.
2.3.2. Intergovernmental Forums
The intergovernmental forums are formed to recommend the
distribution of the federal revenue among the federal government and the
constituent units. These forums typically enjoy representation from all
the stakeholders and provide room for some bargaining over the
distribution criteria. The limits of the bargaining are defined by the
constitution e.g., in Pakistan the revenue sources available for
distribution are defined by the constitution. Shah (2007) prefers a
simple distribution criterion which may render only approximate justice
to each constituent unit over a complex criterion with complete justice.
The study argues that quite often the constituent units have conflicting
interests and a forum with conflicting interests cannot handle a complex
distribution criterion. Countries that rely solely on intergovernmental
forums include Germany, Indonesia and Nigeria. Pakistan also relies on
such a intergovernmental forum with the difference that the Commission
members also include experts from the civil society of each province.
Countries like South Africa and India make use of an independent agency
in addition to the intergovernmental forum.
2.3.3. Independent Agency
An independent agency is established by the central government to
make recommendations to the government or the legislature on resource
transfers to the constituent units. The members of the agency are
experts in fiscal management. Some countries, for example India, draw a
member from the judiciary as well. Typically, this kind of agency has an
advisory position. Australia was the first to establish an agency for
recommending resource transfers in 1933. Since then this institution has
become popular in a number of countries including India and South
Africa. The independent agency was established in Australia after some
states had expressed dissatisfaction with the process of bilateral
negotiations with the federal government on requests for special grants.
A secession threat by Western Australia proved instrumental in the
decision to set up an independent agency. Thus the origin of the
independent agency has lessons for countries where one of the
constituent units is dissatisfied with the resource distribution.
The objective of setting up an independent agency is to let the
experts recommend the distribution criteria based on professional
knowledge and rigorous analysis of the prevailing economic environment.
The rationale for an independent agency is that it can disengage the
distribution criteria from politics. Shah (2007), however, does not
favour the independent agency on the ground that it tends to offer a
complex solution to an otherwise simple task thus increasing the cost of
devising the resource distribution criteria. Moreover, given the
complexity of the distribution formula it becomes difficult for the
ordinary citizens to monitor the performance of the agency.
2.4. Proposed Institutional Arrangement for Distribution of
Revenues
The foregoing suggests that both the intergovernmental forum as
well the independent agency have their merits and demerits. With the
regional representatives on board, the intergovernmental forums can
protect the regional interests more effectively but if these forums
follow the unanimity rule then their proceedings are prone to deadlock.
The independent agencies, comprising experts, can bring the required
rigour into the revenue sharing exercise but these agencies tend to
complicate problems that may have a simple solution. A better
institutional structure for revenue sharing would be one where the two
can supplement each other. Therefore we suggest that a two-tier
institutional structure may be set-up in Pakistan to design revenue
sharing among the constituents of the federation. The proposed two tiers
are: (i) an independent body of experts and (ii) an intergovernmental
forum.
An independent body, comprising fiscal experts, practitioners as
well as academicians, would constitute the first layer of this two-tier
structure. Experts will be selected without regard to provincial
affiliations and they would be full time/part time employees of the
independent agency. The agency would have the mandate to recommend not
only the sharing of the divisible pool but also to determine the revenue
sources that should comprise the divisible pool. The agency would also
have the mandate to recommend assignment of specific revenue sources to
the federal or provincial government. The agency would have resources to
conduct or commission research on the issues under consideration as well
as to make use of existing research available on the subjects. The
agency will commence its task two years before an Award is due and will
have 16-18 months to conduct research, deliberate upon possible options
and then make its recommendations. The agency will send its
recommendations to the upper tier, the intergovernmental forum. The
recommendations of the agency would carry detailed justification in
their support especially if the advice deviates from the established
formula. The recommendations would also be made public to encourage
debate on the subject. The independent agency will not insist on
arriving at a unanimous set of recommendations, a principle which in the
past has caused deadlock in NFC proceedings. The notes of the dissenting
members would form part of the independent agency's report.
The upper tier, the intergovernmental forum, will comprise the
federal and provincial ministers of finance only. The experts need not
be on the forum because the the expert work has already been done by the
independent agency. The forum would review the recommendation of the
independent agency and may or may not accept all or some of these. The
forum would pay due regard to the political factors and other
sensibilities that the independent agency would not have taken into
account. Moreover the forum will also take into consideration the public
debate on the recommendations of the independent agency. If the forum
decides not to accept some or all of the recommendations of the
independent agency, the forum and its individual members would have to
offer their reasons or justifications for their point of views. The
intergovernmental forum will then send its recommendation to the
government for final approval and announcement of the Award.
This two step approach is likely to put an end to the deadlocks
which have beset revenue distribution among constituent units in
Pakistan. The proposed two-tier institutional structure is an
improvement over the existing one for the following reasons:
* The experts drawn from the profession and the academia without
regard to provincial affiliations and put in the position of a
'judge' are less likely to take a biased position.
* The experts being paid employees of the independent agency would
do their assigned work according to the charter of the body rather than
work as a lobby for a particular constituent unit.
* The knowledge that the recommendations of the agency will be
debated publicly will induce the members to offer sound and practical
recommendations.
* The experts' reliance on research will:
--enable the independent agency to offer sound and practical
recommendations.
--make it difficult for the agency or the individual members to
take unjustified positions.
* It would not be possible for the intergovernamental forum to
easily ignore the recommendations of the independent body for the
following reasons:
--These would have the backing of eminent experts.
--The recommendations would have attracted sufficient public debate
by the time intergovernmental forum takes a decision on these.
--The forum and its individual members will have to record reasons
if they decide not to accept the recommendations of the independent
agency.
3. DISTRIBUTION OF REVENUE RESOURCES
The federal revenues available for distribution among provinces
have historically been distributed on a per capita basis. The 7th NFC
Award accepted a long standing demand of the three provinces for the
introduction of a multiple indicator criteria. The rest of this study
examines the new distribution design in the light of the revenue
distribution practices followed in other countries.
3.1. Resource Distribution Practices Adopted Internationally
Transfers from federal government to the sub-nationals take several
forms. These are formula-based as well as discretionary and could be
block unconditional, conditional or matching. The transfer programmes
often aim at fiscal equalisation among the constituent units i.e., to
enable the constituent units to provide the same kind of service with
comparable level of taxation.
In Canada, transfers from the federal to provincial governments are
unconditional and are given to only those provinces whose revenue
raising capacity is below the national average. It is noteworthy here
that revenue generation is highly decentralised with the share of
provincial own-source revenue standing close to 80 percent of the total
national revenue. It is only under this kind of revenue decentralisation
that some provinces can manage to function without any equalisation
transfers. The Indian system essentially involves distribution of funds
on the basis of estimated expenditure needs and, to an extent, on the
potential of the subnationals to generate revenues from their own
sources. The finance commission of India primarily uses the gap filling
approach for equalisation of fiscal capacity across states. The states
are allocated shares in central taxes based on a formula and the
difference between a state's budgetary expenditures and its
revenues is filled through the grants-in-aid. It is argued that the gap
filling methodology not only acts as a disincentive for the subnationals
to raise own-source revenue but is a source of inequity as well. In
Australia, the comprehensive nature of equalisation allows assessment of
all the circumstances that affect the relative cost differences a state
is faced with in delivering standard services. These include additional
costs faced by a sub-national government in meeting requirements of
large cities as well as in providing services in rural areas and remote
locations. If a state's differential per capita revenue or
expenditures is considered beyond the control of the state, for example,
due to geography, it is compensated for that. The Australian approach to
equalisation requires voluminous data across states at a high level of
disaggregation. The Australian equalisation programme has been
criticised on the grounds of efficiency, complexity and reliance on
internal standards rather than best practices. It is argued that
reliance on average internal standards in a sense rewards some states
for maintaining lower standards. However, by and large, there is a
general acceptance of the system. It is precisely because of carrying
out a very thorough equalisation programme that federal government
(known as the Commonwealth government) has been able to keep the states
satisfied despite continuing with the large vertical fiscal imbalance
(difference between revenue generated by the federal government and
states).
In the United States, unlike other federal countries, there is no
general form of revenue sharing. However around 600 grant programmes
exist for state and local governments. The different forms in which
grants are provided include project, categorical, and block grants.
While some grants have matching component, others have structured
formulas. Barring federal transfers for some specific purposes, the
overall grant system is small relative to other countries. Though a
degree of equalisation is built into grant programmes, however, in
general, the intergovernmental transfers do not aim at equalisation
despite wide differences in taxable capacity across states.
In Germany, the intergovernmental transfer system is highly
egalitarian. The unique feature of the German system is that richer
states transfer money to the poorer states. In practice, the states,
whose taxable capacity is below the national average, receive transfers
from the states with taxable capacity above the national average. The
transfer programme is designed in a manner that fiscal capacity of the
below-average state is brought to 90 percent of the national average.
These interstate transfers are unconditional.
The transfers from the federal governments to the provinces
typically attempt to equalise fiscal capacity and in some cases fiscal
needs as well (United States is an exception). The amount of transfers
in a number of countries is determined on the basis of some formula.
Indicators like population share, poverty, demographics, fiscal effort
and population density are typically used to determine fiscal needs and
capacities. 'Population share' is not considered a good
indicator of fiscal needs and is used only in a handful of countries.
Even the countries that use population share as the criterion for
revenue distribution typically accord a rather low weight to it in the
distribution formula e.g., India. Nigeria, with transfers based solely
on the basis of population, is an exception. Pakistan, with 82 percent
weight for population share, stands close to Nigeria.
Transfers are also used to achieve certain national objectives, for
example, education and healthcare for all. One of the typical
characteristic features of the transfer programmes is the use of
conditional and matching transfers for the provision of healthcare,
education and social security. The use of conditional/matching transfers
for these services reflects the importance attached nationally to the
provision of these services. The aim is to provide the specified
services to all up to a certain minimum level defined by the society.
Such choices are made through a variety of collective choice mechanisms
such as voting for electoral promises of the political
parties/candidates.
In Canada, besides the equalisation transfers, the other major
forms of transfers are the equal per capita transfers which are
nominally divided into two components--the Canada Health Transfer (CHT)
and the Canada Social Transfer (CST) which include welfare and post
secondary education. Only minimal conditions are attached to the
payments. To be eligible, the provinces cannot impose residency
condition on welfare payments and health insurance programmes have to
follow general criteria including access, affordability and
comprehensiveness.
In Australia, huge transfers from the federal government to the
states are made under the special purpose programmes (SPPs). These SPPs
are intended to support the implementation of some national priority and
these are in addition to the transfers from the united pool of funds
determined in the manner described earlier. The largest SPPs are in the
areas of education, health, social security, transportation and housing.
SPPs constitute a significant proportion of the total assistance from
the federal government to the states. This proportion has varied from 25
percent of the total federal assistance in early 1970s to 50 percent in
1990s. The majority of the SPPs are subject to conditions--the
conditions designed to ensure the achievement of national objectives.
These conditions include general policy conditions that the amounts so
transferred be spent on designated purposes only. Sometimes the
transfers require matching expenditures from the state's own
sources for the same purpose. Such grants are determined through
bilateral negotiations between the federal government and the concerned
state as well as negotiations at some forum where all states are
represented. In the United States, grants for health and income security
constitute the major purposes for which transfers are made to the state
and local governments. These grant programmes are discretionary at the
national level and are determined through the annual budget process. The
interstate highway system is financed jointly by the federal and state
governments with federal government typically funding 90 percent of the
construction cost. Other major grant categories include education and
transportation. In South Africa, in recent years the share of
conditional specific purpose grants, which are discretionary in nature,
have exhibited sizable growth in the total transfers to the provinces.
The discretionary nature of the conditional grants has made the
transfers system less transparent.
3.2.1. Fiscal Equalisation: What Method to Use?
The subnationals typically encounter a fiscal gap--the difference
between expenditure needs and the revenue means. The gap may arise
either because a region does not inherently enjoy the potential to
generate revenues or because the taxing powers are centralised with the
federal government. Whatever the reason for the fiscal gap, leaving the
gap unattended has economic as well as political ramifications. The gap
may cause large fiscal disparities among the regions which could be
politically divisive for the federation. This threat cannot be taken
lightly. Since 1975 more than 40 countries have been created and a
deeper analysis of the independence/liberation movements would reveal
that fiscal disparity, among the regions of a nation, was at the heart
of many if not all movements. Evidence suggests that addressing the
fiscal gap helps curb the feeling of deprivation and therefore
forestalls cessation threats. Australia and Canada have successfully
thwarted cessation attempts by bridging the fiscal gap of the
sub-nationals and through various autonomy measures. The primary tool of
fiscal equalization, are intergovernmental transfers, in the form of
revenue sharing and grants [Bilin (2005)]. The typical methods of
determining the size of transfers from the federation to the
subnationals include:
(1) Equalisation of fiscal capacities and fiscal needs,
(2) Fiscal capacity equalisation,
(3) Need criterion,
(4) Population share criterion.
The method of equalising the fiscal needs as well as the fiscal
capacity recognises that both may vary across regions. This method of
equalisation seeks to address the net variation in the fiscal need and
fiscal capacities of the regions. The method of equalising fiscal
capacity only assumes that the per capita fiscal needs are more or less
equal across regions. This method aims to transfer more funds to the
region whose fiscal capacity is below the national average. Both these
methods require voluminous data on revenue generation, actual as well as
potential, as well as minute details of the expenditure needs. The two
methods are therefore difficult to use in developing countries.
The need indicator criterion recognises that fiscal needs may vary
across regions. This criterion seeks to estimate the expenditures of the
subnationals on certain major fiscal needs using statistical and
econometric techniques. These estimates are then used to compute the
total fiscal need of the region. To estimate the expenditure on a
certain need, say healthcare, the need index is developed using possible
factors that may influence the healthcare such as the demographic
profile of the region, the historical evidence on common ailments and
the expenditures thereon. A certain weight is then assigned to
healthcare needs keeping in view the value of the index and the
historical share of the healthcare expenditure in the total expenditure.
Need indicators typically used to estimate fiscal needs of the
subnationals include: population, per capita income, unemployment rate,
population density, geographical area, infant mortality, life
expectancy, school enrolment rate and infrastructure. The multiple
indicator criterion (MIC) adopted by the 7th NFC is similar in spirit to
the need indicator criterion. However the MIC includes fewer indicators
than are typically included in the criterion. The weight determination
exercise for the individual elements of the MIC does not seem to be
supported by a detailed and rigorous exercise and the weight of the
population share is too large.
The last of the four fiscal equalisation methods mentioned above is
the population share criterion which has been in vogue in Pakistan until
2009. Ma (1997) argues that the use of population share criterion is
least effective at securing equalisation of fiscal needs across regions.
The population share criterion assumes that per capita expenditure needs
are equal across regions. However, in practice, the per capita
expenditure needs may vary, due to differences in population density,
geography, history, resource endowments and the level of development.
Moreover, the remote location or the difficult terrain of an area may
increase the cost of delivering public services. The metropolitan
character of a city may also call for incurring above average
expenditures.
3.2.2. The Absence of Matching Grants from the Distribution Design
The provinces in Pakistan are free to use the transfers from the
federal government in the manner they deem fit. Such block unconditional
transfers, though in accord with the spirit of the provincial autonomy,
do not provide any guarantee that funds will be used to provide a
minimum level of public service, especially in respect of essential
needs like healthcare and education. Thus, with unconditional block
transfers the level of public service in respect of essential needs may
vary across jurisdictions. The question then is, what is more
important?--provincial autonomy or homogeneous minimum national
standards across provinces for essential social services.
The merits of provincial autonomy notwithstanding, there are strong
arguments for setting uniform minimum national standards for essential
social services like healthcare and education. The conventional wisdom
that inequality is essential for economic growth [Kuznets (1955), Lewis
(1954)] has been convincingly challenged in recent decades [Galor and
Zeira (1993); Easterly (2007)]. Raising the living standards of lagging
regions is now considered important for aggregate economic prosperity as
well as for political stability. Moreover the homogeneous national
standards encourage mobility of goods, services, labour and capital
across jurisdictions. The uniform standards also increase the market for
the goods of any region and allow the regions to gain from their
respective comparative advantage. Establishing relatively homogeneous
standards calls for incurring greater expenditures in regions that are
below the national average. Conditional or matching grants can be used
to achieve uniform standards across jurisdictions. A region that lags
behind, say on healthcare indicators, can be induced by the federal
government to improve healthcare services by conditioning the transfers
with the kind of measurable improvement that is desired.
Conditional or matching grants, especially for social needs like
healthcare and education, are a key element of the transfer programme in
the developed countries. This is despite the fact that the revenue
mobilisation is fairly decentralised in these countries--United States,
Canada and a number of other countries. The rationale for conditional
transfers, besides the uniform national standards, is that the
subnationals in an effort to woo businesses into their jurisdiction may
impose lower tax burden on them. This may ultimately result in under
provision of essential public services. Conditional grants ensure that
essential services will be provided to the required minimum level.
Conditional grants could be administered in a variety of ways.
Conditions may be imposed on the subnationals either with respect to
inputs (i.e. expenditures) or outputs (i.e. desired results). The input
grants may encourage the sub-nationals to engage in wasteful expenditure
to show higher numbers. This kind of adverse incentive cannot be related
to output based grants. Therefore the output based grants are preferable
unless the measurement of output is highly difficult. Conditions would
be imposed not on the specific use of grants but on attainment of
standards in quality, access and level of service. Matching grants allow
the subnationals to access transfers if they spend a certain specified
percentage on a specific service from their own sources. Such grants are
termed open-ended when there is no limit to transfers from the federal
government on this count. Close-ended programmes, on the other hand, put
a maximum cap on matching transfers. The close-ended programmes are
favoured over open-ended grants because these can be designed while
taking into account the budget constraint of the federal government.
The literacy rate of Khyber Pakhtunkhwa and Balochistan is
significantly lower than that of Punjab and Sindh (Table 2). It is
obvious that the two lagging provinces need to spend more on education
to bring their literacy rate closer to the other two provinces.
The use of elements like poverty and inverse population density as
indicators in the distribution formula is based on the fact that some
provinces lag behind others in the level of development. Greater funds
have been transferred to the provinces under the 7th NFC Award on
grounds of higher cost of delivery (reflected by Inverse Population
Density) and poverty. Increase in literacy could be an ideal way to
alleviate poverty on a long term basis. The block unconditional
transfers do not guarantee that the additional funds will be used to
alleviate poverty or, for example, will be spent on increasing access to
education in the sparsely populated Balochistan. The two provinces could
have been made to spend more on social services had the incremental
transfers been conditioned upon certain improvement in literacy rate,
enrolment rate or the patient-doctor ratio. A mix of general purpose and
matching grants would better serve the cause of development in Pakistan.
3.2.3. Demographic Structure and Distribution Criteria
The demand for public services for different age groups is
different. For example, the population aged 5-20 needs education while
the elderly require greater healthcare. If the age structure of the
population varies across regions then, to provide equal level of
services, the expenditure will vary across regions. The estimated
province-wise age structure of population in Pakistan, as of 2006, shown
in Table 3, depicts that the school age population is relatively greater
in the provinces of Khyber Pakhtunkhwa (KP) and Balochistan while
elderly population is greater in Punjab and Sindh.
Given the province-wise demographic structure of the population, it
is clear that the need to spend on education is greater in KP and
Balochistan while the need to spend on healthcare is greater in the
remaining two provinces. This will be true even if we assume equal per
capita expenditure on these services across provinces. The foregoing
suggests that to provide more accurately for the expenditure needs of
the provinces the demographic structure should be accounted for in the
distribution formula. However, the demographic structure is not an
element of the multiple indicator distribution criteria adopted by the
7th NFC Award.
3.2.4. Weights of the Multiple Indicator Criteria
The 7th NFC Award has assigned certain weights to the four elements
of the multiple indicator criteria. There could be no two opinions that
the methodology for the determination of weights should be widely known
in the interest of transparency and public debate. However, this is not
the case. It is unknown what role the historical expenditure patterns,
statistical tools and research have played in the weight determination
exercise and to what extent rough calculations and political manoeuvres
have influenced the weights. The weights influenced by political
compromises are likely to prove less stable as there could be a demand
for revision with the change in the power configuration.
To illustrate how the weights should be computed one could compute
the per pupil cost of education for a school located in some remote area
of Balochistan and compare this with the corresponding cost for some
school located in the central Punjab. The difference in the two costs
could form the basis for the weight of inverse population density. This
example is only illustrative and of course the cost differential would
have to be examined in greater detail to construct the weight. Similar
exercises could be undertaken to compute the weights of other elements
of the criteria.
3.2.5. Weight of Population Share
The previous distribution criterion was criticised primarily on the
ground of revenue sharing solely on the basis of population. With the
assignment of 82 percent weight to the population share no major change
has been effected in the distribution formula. Thus all the arguments
put forth to criticise the previous formula are still valid. Very few
countries make transfers to subnationals on the basis of population
share and the ones that do accord it a small weight, for example 10-20
percent in India. A problem with the use of population share criterion
is that the provinces may question the credibility of the population
census. Nigeria, where transfers are solely on the basis of population,
has encountered such problems. Perhaps in an effort to avoid the
problems of the sort, India is still using the population figures of
1971 to distribute revenue according to weightage assigned to
population.
3.2.6. Poverty as an Element of Multiple Indicator Criteria
It is generally argued that revenue distribution should not be
based on indicators that are likely to generate perverse incentives.
Poverty level is one such indicator. The use of poverty as an indicator
acts as a disincentive for the provinces to alleviate poverty because
the poorer a province, the greater its entitlement under the NFC Award.
Moreover, the estimates of poverty levels in Pakistan have been
questioned for accuracy. This has prompted the 7th NFC to use the
average of the estimates generated by the three different agencies.
Including 'poverty' as an element in the revenue distribution
criteria will make the provinces stakeholders in the poverty estimation
exercise. How this would influence the estimates is difficult to tell.
It may add to the controversy about the accuracy of the estimates but,
on a positive note, the possibility is that the estimation exercise may
become more transparent and less questionable, given the potential gains
and losses of the different stakeholders.
3.2.7. Provincial Resource Mobilisation
The National Finance Commission presently does not enjoy the
mandate to offer advice on the provincial revenue generation but still
designing the revenue distribution mechanisms hinges on the extent of
own-source revenue generated by the provinces--if the provinces generate
more own-source revenues the reliance on federal transfers decreases.
The intergovernmental fiscal relationship in Pakistan is highly
imbalanced. The provinces account for around 35 percent of all
government expenditures but they generate merely 8 percent of the
consolidated national total tax revenue which is only 0.5 percent of the
GDP. The need to improve provincial resource mobilisation is but
obvious. (A comparison of the intergovernmental fiscal imbalance is
given in Table 4. Though six years old the comparison still shows that
decentralised revenue generation in Pakistan is among the lowest in the
world).
In the context of fiscal relationship between the federal
government and the provinces the primary issue is how the fiscal needs
of the provinces should be met? Whether the federal government should
collect a larger part of the revenue and then transfer it to the
provinces through some transfer mechanism or the provinces should be
allowed to generate more revenues on their own and rely on the federal
government only to cover the shortfall. The latter approach may have
several advantages as discussed below. (2)
The low revenue mobilisation on the part of the provinces should be
viewed in the perspective of the national tax effort. The aggregate
tax-to-GDP ratio in 2009-10 was 10.5 percent and has been on the decline
for over a decade (it was 12.5 percent in 1996). This is significantly
lower than the average for developing countries (15 percent) as well as
developed countries (35 percent). The tax-to-GDP ratio is much better
even in the South Asian countries like Sri Lanka (16 percent) and India
(14.5 percent) [Nabi and Shaikh (2011)]. According to Bhal, et al.
(2008), the present state of revenue decentralisation and its future
prospects present a dismal picture. Though the provinces have access to
as many as 15 tax bases, the effective yields are very low. The tax
bases are considerably eroded due to exemptions and are undervalued,
incomplete and dated. Moreover, while the broad based taxes like
personal income tax, tax on corporate profits, sales tax on goods and
custom duties are with the federal government, the hard to collect taxes
are with the provinces: sales tax on services and the tax on
agricultural income--the former is administratively difficult and the
latter is politically sensitive.
Though most of the broad based taxes have been assigned to the
federal government but the incentive to mobilise revenue at the central
level may not be as much as it could be at the provincial level. The
fact that 57 percent of what is collected does not remain with the
federal government may dampen its incentive to increase collection from
the revenue sources that are to be shared with provinces. Moreover, with
access to money creation and foreign aid, the federal government may not
be as hard pressed for cash as the provinces are--provinces cannot
create money and they have only recently been allowed to borrow abroad,
but only under restrictive conditions.
Though the conventional wisdom suggests that broad based taxes like
the personal income tax and the tax on corporate profits should be with
the federal government, some federal governments in developed countries
are successfully sharing these taxes with the subnationals. The federal
government in Pakistan shares the tax revenue with the provinces but
only through the NFC Award, not the tax bases.
The devolution of taxes has several advantages. If the provinces
are allowed to share the broad based tax bases like personal income tax
and tax on corporate profits with the federal government this would
solve the free rider problem. The provinces would make an effort to
generate more from the two tax bases because the revenue would belong to
them. Moreover, better revenue generation by one province can generate a
strong demonstration effect, encouraging other provinces to emulate the
example set by the high revenue generating province. To accomplish the
sharing of the tax bases, the federal government may reduce its tax rate
on corporate profits and personal income to make room for the provinces
to levy tax on these bases. For example, a reduction of 10 percentage
point in corporate tax rate will allow the provinces to tax corporate
profits at the rate of 10 percent. The revenue loss that federal
government will incur would be offset by the reduced transfers to the
provinces under the NFC Award. Overall the national tax revenue is
likely to increase due to this kind of sharing because of the greater
incentive of the provincial governments to collect more taxes.
The devolution of taxes like income tax to the provinces in
Pakistan is criticised on the ground that the provinces do not have the
requisite administrative capacity for the purpose, which with all its
machinery and power even the federation finds difficult to collect.
Increase in collection cost due to the loss of scale economies is yet
another argument against decentralisation of revenue generation. Here
one can learn from the Canadian example. In Canada though the taxes have
been devolved in the sense that the provinces are free to set their own
rate structure, a single Canadian Revenue Agency collects the income tax
on behalf of the provinces [Boadway (2007)]. A system on these lines can
also be developed in Pakistan. The collection of provincial revenue
against the tax bases being shared with the federal government can be
assigned to the federal government for a certain charge. The collection
of taxes by the Federal Board of Revenue, on behalf of the provinces
will take care of the supposedly low collection capacity in the
provinces and a higher aggregate collection cost under the devolution.
3.2.8. Revenue Generation Effort
The 7th NFC has included revenue generation (more commonly known as
tax effort) as an element of the resource distribution criteria. This is
a welcome development. However, the 10 percent weight assigned to
revenue generation is not enough to induce the provinces to increase
their tax effort. The effort made by the provinces to generate tax
revenue is accounted for in a number of countries while determining the
size of transfers. The objective is to encourage the provinces to
generate more own-source revenue by rewarding the existing revenue
generation. Own-source revenue generation has a number of advantages. It
reduces dependency on the federal government and improves governance at
the regional level. Moreover, each province can levy taxes in accordance
with the preferences of the electorate for the level and kind of public
service required.
Moore (2000) argues that nations that mostly rely on unearned
income (defined as foreign aid or income from natural resources) are
typically poorly governed. The reason is simple: with easy access to
money the rulers do not have to enter into a 'bargain with the
citizens'--taxation revenues in return for good governance and
better service delivery. Unconditional transfers from the federal
government to the subnational governments are like aid to a country from
a foreign nation--this reduces the need to raise revenue from the
citizens and thus saves the rulers from a more difficult task--providing
good governance and better service delivery. (3) If the federal
government conditions the transfers to the provinces with sufficient
demonstration of own-source revenue generation effort, then the
provinces would have no choice but to mobilise more own-source revenue.
3.2.9. Revenue Generation Effort as Element of Distribution
Criteria
Though the inclusion of revenue generation as an element of the
distribution criteria is a step in the right direction, there are some
issues in its implementation. Revenue generation as an element of
distribution criteria means the total tax revenue generated in a
province i.e., the tax collected in a province against the tax bases
assigned
to the federal government as well as to the provinces. As argued
below, a better approach for this purpose would be to consider only the
revenue generated against provincial tax bases (own-source revenue).
If the objective of the inclusion of 'revenue generation
effort' in the distribution criteria is to encourage generation of
own-source revenue by the provinces, then it is not clear how
distribution on the basis of federal revenue generated in a province
would encourage generation of own-source revenue. Moreover, collection
of revenues by the provinces against tax bases assigned to the federal
government would not yield (and has not so far yielded) the benefits of
own-source revenue generation for two reasons. First, the machinery for
tax collection is federal rather than provincial and secondly, the
citizens do not expect the provincial governments to provide better
services in return for federal taxes. Thus the improvement in governance
at the provincial level would not result merely because more federal
revenue is being generated by a province.
Another problem with the use of the revenue collected in a province
against federal tax bases is that numerous firms do business and
generate income in more than one province. Logically, the tax should be
payable in the province where the income is generated. However, for
administrative convenience the firms are required to pay tax on their
consolidated national income in the province where the head office of
the firm is located. Since the income tax is a federal tax, therefore
the provinces, as well as the federal government, were till now
indifferent to whether the tax payable from income generated in province
X is actually deposited in province X or province Y. However now, that
the revenue generation is an element of the distribution criteria, the
administrative convenience referred to above gives an undue advantage to
the province that might be host to head offices of a greater number of
firms. For example the banking sector--the largest tax payer, generates
income from all over the country, but pays income tax mostly in Sindh on
its consolidated income in the country. The reason is that the head
offices of most of the banks are located in Karachi--the capital city of
Sindh. A more realistic approach therefore would be to include only the
revenue generated against provincial tax bases for determining the tax
effort of the province.
If at all it is essential to include the revenue generated against
federal tax bases then the income generated by multi-provincial firms in
each province should be estimated so that the tax liability against the
province-wise income of the firm can be assessed for the purpose of the
distribution criteria. Whereas estimating regional profits for a
multi-provincial firm may be a difficult exercise, the practices adopted
by different countries can be examined to estimate the regional
earnings.
If the changes discussed above are incorporated in the distribution
design, then the weight of 5 percent assigned to revenue generation
should be increased significantly to encourage own-source revenue
generation by the provinces. This would encourage the provinces to
increase revenue generation from the provincial tax bases. It may be
mentioned here that some important tax bases assigned to the provinces
include property tax, tax on agricultural income and GST on services.
3.2.10. Specification of the Divisible Pool: A Disincentive for
Resource Mobilisation
The process of distribution of revenues between the federal
government and the provinces begins with the specification of the
divisible pool--the revenue sources which the federal government can
share with the provinces. Most but not all revenue sources are included
in the divisible pool, for example personal and corporate taxes are a
part of the divisible pool while Petroleum Development Levy (PDL) is
not. Exclusion of some revenue sources from the divisible pool
encourages the federal government to concentrate on increasing revenues
from the excluded sources because the revenue from these does not have
to be shared with the provinces. The specification of divisible pool
creates a disincentive for the federal government to increase revenues
from the sources which comprise the divisible pool. To illustrate,
suppose that the federal government wants to raise its own revenue by Rs
100. To raise the required amount through tax on corporate profits the
federal government would have to increase the corporate tax rate by such
percentage that an additional amount of Rs 236 is mobilised. The federal
government needs to mobilise more than the revenue that it requires
because 57.5 percent of the additional revenue i.e. Rs 136 would go to
the provinces, leaving the federal government with the required Rs 100.
An alternative for the federal government is to increase the PDL by such
percentage so as needed raise an additional Rs 100 only. The PDL
requires lesser increase because it is not a part of the divisible pool
i.e. the revenues from PDL are not to be shared with the provinces.
How can these disincentives be avoided? Table 5 provides the
answer. At present 44 percent of the gross national revenue is being
transferred to the provinces. Instead of specifying an elaborate list of
revenue sources which would form the divisible pool, it can be simply
stated that 44 percent of the gross federal revenue would constitute the
pool of resources divisible among the provinces. This would take care of
the federal disincentive to increase revenue from the sources that are
divisible. This of course would have to be qualified with details like
excluding royalties from oil and gas, which are to be transferred in
full to the province concerned.
3.2.11. Evaluation of Distribution Design Against Best Practice
The broad principles of resource distribution design derived from
the review of relevant literature are given in Box 1.
Box 1
Autonomy: The transfers should allow the subnational governments to
determine their own expenditure priorities.
Predictability: The amount transfers should be known well in
advance so that the provinces may budget their expenditures with a
modicum of certainty.
Simplicity: The transfer criteria should be objective and be fairly
easy to understand.
Equity: The transfers should take care of the fiscal needs of each
subnational government.
Revenue Adequacy: Transfers should take care of the imbalance in
resource availability between the federation and the provinces as
well as amongst the provinces.
Incentives: transfers should encourage constituent units to raise
revenues and control expenditures.
Accountability: The grantor must be accountable for the design and
operation of the programme. The recipients must be accountable to
the grantor and the citizens for financial integrity and better
utilisation.
Source: Primarily Adapted from Pulling Back from the Abyss: Third
Annual Report, Institute of Public Policy, Beaconhouse National
University. (The last point 'Accountability' is an addition to the
criteria included in the report.)
The revenue distribution design in Pakistan fares well on the
yardsticks of autonomy and simplicity. Predictability is not complete
but not bad either. However, it scores poorly on the scales of
'incentives' and 'accountability' whereas it is too
early to assess it in terms of 'revenue adequacy' and
'equity'. The performance of the distribution design is
discussed below in some detail.
The subnationals enjoy complete autonomy as to the use of the funds
available, the criteria is simple enough--the weights assigned to the
elements of the criteria are known in advance and one has to know only,
for example, the population of a province to figure out the grant
entitlement of the province against the population share. The absolute
amounts of transfers that a province is to receive are partially
predictable. The criterion describes the transfers in terms of
percentage share of the divisible pool. The absolute size of the
divisible pool in a financial year depends upon the revenue that the
federal government is able to generate against the sources included in
the divisible pool. The transfers are predictable in the sense that the
federal government sets the target for collection against each tax
source, thus the targeted amount of the divisible pool and the targeted
provincial shares are known at the beginning of the financial year.
However, the transfers are unpredictable in the sense that in recent
years the federal government has been missing the revenue target by a
significant margin. This introduces an element of uncertainty regarding
the size of the divisible pool and hence the provincial share of
transfers. More time is required to grade the new distribution design on
the criteria of 'revenue adequacy' and 'equity'. On
the one hand, a large number of new functions have been transferred to
the provinces under the 18th amendment, while greater funds are
being transferred under the 7th NFC Award. With more functions to
perform, and greater financial resources at the disposal of the
provinces, only time will tell whether the resources are enough to meet
the financial needs of the provinces and if these are equitably
distributed across provinces.
The revenue distribution design fares poorly in terms of
'incentives' to raise own-source revenue and
'accountability' for the appropriate use of the transfers. As
discussed in section 3.2.8, the 7th Award in fact had a dampening effect
on own-source revenue generation. The reason of course is greater
transfers from the federal government. Lesser transfers from the federal
government coupled with perhaps partial allocation of some attractive
tax bases, like income tax, to the provinces will encourage the
provinces to increase revenue generation from their own tax sources.
This is also likely to make the provinces more accountable to their own
electorate as successful taxation is essentially a bargain between the
citizens and the government.
To conclude, the distribution design is simple, allows autonomy to
the provinces as to the use of funds and the amount of transfers is
predictable, though with a degree of uncertainty. However, the
distribution design offers no encouragement to the provinces to raise
own-source revenue. The two principles, namely 'autonomy' and
'incentives to raise own-source revenue' may at times
conflict--the distribution design that offers greater incentive for
own-source revenue generation may not always allow complete autonomy to
the constituents as to the use of transfers. For example, the use of
matching grants in some developed countries restricts the use of
transfers for certain specific purposes but at the same time encourages
own-source revenue generation because the subnationals have to spend a
part of the amount from own sources. It is society in general and the
policy makers in particular who choose between greater provincial
autonomy in the use of transfers and more incentives to generate
own-source revenue. The latter has more benefits from the perspective of
economic development and good governance.
4. CONCLUSION
The 7th NFC Award in December 2009 ended a prolonged deadlock over
the design of the distribution of revenue resources. Yet the present
institutional structure of the NFC remains prone to potential deadlocks
in the future. The two-tier institutional structure proposed in this
study can be helpful in smooth functioning of the NFC with belter
distribution mechanisms. Though the new distribution design is an
improvement over the previous one but still it does not come close to
the international best practices. For example, the population
share--which is not a part of the distribution design in developed
countries and carries a small weight in some developing countries--has a
large weight of 82 percent in Pakistan. A good design should encourage
own-source revenue generation by the provinces as against the potential
dampening effect of the 7th NFC Award in this respect. The distribution
design, on its own, can only partially encourage own-source revenue
generation e.g., by including the revenue effort and matching grants in
the distribution design. While the weight of the revenue effort is not
large enough, the matching grants do not figure at all in the prevailing
distribution design. To further encourage own-source revenue generation,
the federal government needs to share broad base tax bases, like the
income and corporate taxes, with the provinces rather than transferring
the revenue from these through the distribution design.
REFERENCES
Boadway, Robin (2007) Canada. In Anwar Shah (ed.) A Global Dialogue
on Federalism. Vol. 4. Forum of Federations and IACFS, Montreal:
McGill-Queens University Press.
Bhal, R., W. Sally, and C. Musharraf (2008) The Challenges of
Intergovernmental Fiscal Relations in Pakistan: Property Tax Dimension.
International Studies Program. (Working Paper 08-25).
Galor, Oded, and Joseph Zeira (1993) Income Distribution and
Macroeconomics. The Review of Economic Studies 60:1, 35-52.
Institute of Public Policy (2010) State of Economy: Pulling Back
from the Abyss. Third Annual Report. Beaconhouse National University.
Kuznets, Simon (1955) Economic Growth and Income Inequality. The
American Economic Review 45:1, 1-28.
Lewis, W. Arthur (1954) Economic Development with Unlimited
Supplies of Labour. Manchester School 22:2, 139-91.
Ma, J. (1997) Intergovernmental Fiscal Transfers: A Comparison of
Nine Countries, Macroeconomic Management and Policy Division, Economic
Development Institute, Washington, DC, World Bank.
Moore, Mick (2000) Political Underdevelopment. Institute of
Development Studies. (Working Paper).
Morris, Alan (2007) Commonwealth of Australia. In Anwar Shah (ed.)
A Global Dialogue on Federalism. Vol. 4. Forum of Federations and IACFS,
Montreal: McGillQueens University Press.
Nabi, Ijaz and Hina Shaikh (2011) Reforming the Urban Property Tax
in Pakistan's, Punjab. Lahore University of Management Sciences.
(DPRC Working Paper).
Shah, Anwar (2006) Fiscal Decentralisation and Macroeconomic
Management. International Tax Public Finance 13, 4374-62.
BIBLIOGRAPHY
Ahmad, Ehtisham and B. Giorgio (2006) Introduction: Fiscal
Federalism--A Review of Developments in the Literature and Policy. In
Ehtisham Ahmad and Giorgio Brosio (eds.) Handbook of Federalism.
Cheltenham, UK: Edward Elgar Publishing.
Ambrosania, M. F. and Massimo Bordignon (2006) Normative versus
Positive Theories of Revenue Assignments in Federations. In Ehtisham
Ahmad and Giorgio Brosio (eds.) Handbook of Federalism. Cheltenham, UK:
Edward Elgar Publishing.
Anderson, George (2008) Federalism: An Introduction. Ontario: Forum
of Federations, Oxford University Press.
Anderson, George (2010) Fiscal Federalism: A Comparative
Introduction. Ontario: Forum of Federations, Oxford University Press.
Bilin, Neyapti (2005) Equalization via Fiscal Decentralization.
Turkish Economic Association. (Discussion Paper 2011/11).
http://www.tek.org.tr
Dafflon, Bernard (2006) The Assignment of Functions to
Decentralised Governments: from Theory to Practice. In Ehtisham Ahmad
and Giorgio Brosio (eds.) Handbook of Federalism. Cheltenham, UK: Edward
Elgar Publishing.
Easterly, William (2007) Inequality Does Cause Underdevelopment:
Insights from a New Instrument. Journal of Development Economics 84:2,
755-76.
Epko, Akpan U. (2007) Federal Republic of Nigeria. In Anwar Shah
(ed.) A Global Dialogue on Federalism. Vol. 4. Forum of Federations and
iacfs, Montreal: McGillQueens University Press.
Feld, Lars P. and V. H. Jurgen (2007) Federal Republic of Germany.
In Anwar Shah (ed.) A Global Dialogue on Federalism. Vol. 4. Forum of
Federations and iacfs, Montreal: McGill-Queens University Press.
Fox, William F. (2007) United States of America. In Anwar Shah
(ed.) A Global Dialogue on Federalism. Vol. 4. Forum of Federations and
iacfs, Montreal: McGillQueens University Press.
Galligan, Brian (2010) Institutions of Federalism. In Ehtisham
Ahmad and Giorgio Brosio (eds.) Handbook of Federalism. Cheltenham, UK:
Edward Elgar Publishing.
Hussain, I. and Sumbal Rana (2010) Comparison of Fiscal Effort by
Provincial Governments in Pakistan. The Pakistan Development Review
49:4.
Khumalo, Bongani and Mokate Renosi (2007) Republic of South Africa,
In Anwar Shah (ed.) A Global Dialogue on Federalism. Vol. 4. Forum of
Federations and iacfs, Montreal: McGill-Queens University Press.
Kirchgassner, G. (2007) Swiss Confederation. In Anwar Shah (ed.) A
Global Dialogue on Federalism. Vol. 4. Forum of Federations and iacfs,
Montreal: McGill-Queens University Press.
Mclure, C. E. Jr (1999) The Tax Assignment Problem: Conceptual and
Administrative Problems in Achieving Subnational Fiscal Autonomy,
Washington, DC, World Bank.
Musgrave, R. A. (1959) The Theory of Public Finance: A Study in
Public Economy. New York: McGraw-Hill.
Nabi, Ijaz and Hina Shaikh (2010) Fiscal Federalism in Pakistan: A
Radical Departure and Some New Challenges. Lahore University of
Management Sciences. (DPRC Working Paper).
Nambiar, Shankaran (2007) Malaysia. In Anwar Shah (ed.) A Global
Dialogue on Federalism. Vol. 4. Forum of Federations and iacfs,
Montreal: McGill-Queens University Press.
Oates, W. E. (1972) Fiscal Federalism. New York: Harcourt Brace
Jovanovich.
Oates, W. E. (1999) An Essay on Fiscal Federalism. Journal of
Economic Literature 37, 1120-49.
Oraya, Amir K. (2010) Concurrent List: 1973 Constitution and Recent
Political Developments. (Available at: www.forumfed.org/.../
Constitutions%20concurrency%
20&%20current%20events%20AKhan%20finalJan2010.doc)
Rao, M. Govinda (2007) Republic of India. In Anwar Shah (ed.) A
Global Dialogue on Federalism. Vol. 4. Forum of Federations and iacfs,
Montreal: McGill-Queens University Press.
Rezende, Fernando (2007) Federal Republic of Brazil. In Anwar Shah
(ed.) A Global Dialogue on Federalism. Vol. 4. Forum of Federations and
iacfs, Montreal: McGillQueens University Press.
Rodden, J. and W. Erik (2002) Beyond the Fiction of Federalism:
Macroeconomic Management in Mutitiered Systems. World Politics 54,
494-31.
Shah, Anwar (2007) Comparative Conclusions on Federalism. In Anwar
Shah (ed.) A Global Dialogue on Federalism. Vol. 4. Forum of Federations
and iacfs, Montreal: McGill-Queens University Press.
Shah, Anwar (2007) The Practice of Fiscal Federalism: Comparative
Perspectives. In Anwar Shah (ed.) A Global Dialogue on Federalism. Vol.
4. Forum of Federations and iacfs, Montreal: McGill-Queens University
Press.
Tanzi, Vito (1996) Fiscal Federalism and Decentralisation: A Review
of Some Efficiency and Macroeconomic Aspects, Annual World Bank
Conference on Development Economics.
Ter-Minassian, T. (1997) Decentralisation and Macroeconomic
Management. (IMF Working Paper, IMF/WP/97/155).
World Bank (2006) Property Taxes in the Punjab, Pakistan. The World
Bank Infrastructure and Energy Department South Asia Region. (Available
at: http://www.urbanunit.gov.pk/pubpublic/pubcomp/ 36.pdf)
(1) The province of Gilgit-Baltistan was created after the latest
Award (i.e., 7th NFC Award) had been announced and at present its
expenditure needs are being directly met by the federal government.
(2) This discussion draws on the chapter on 'Provincial
Resource Mobilisation' in Fiscal Decentralisation in Pakistan,
Pakistan Institute of Development Economics.
(3) One reason why very meagre amount of agricultural income tax is
collected in Pakistan is that the tax lies in the domain of the
provinces which have little incentive to mobilise own revenues owing to
their reliance on transfers from the federation.
Table 1
Criteria for Distribution of National Revenue
Presidential 7th NFC
Order 2006 Award
Provincial Share in Divisible Pool 46.25% 56% increasing
to 57.5%
Grants and Subventions 3.75% --
Indicators and Weights
Population 100% 82.0%
Poverty 10.3%
Revenue Generation 5.0%
Inverse Population Density 2.7%
Given the Weights indicated above, the provincial
share in the Divisible Pool works out as follows:
Punjab 53.01% 51.74%
Sindh 24.94% 24.55%
Khyber Pakhtunkhwa 14.88% 14.62%
Balochistan 7.17% 9.01%
Source: Adapted from "Pulling Back from the Abyss:
Third Annual Report", Institute of Public Policy,
Beaconhouse National University.
Table 2
Literacy Rate (Provincial Profile)
Age Punjab Sindh KP Balochistan
Literacy Rate 59 59 50 45
Source: Demographic and Health Survey of Pakistan (2006).
Table 3
School/College Age Population (Percent)
Age Punjab Sindh KP Balochistan
School Age (5-19 Years) 38.2 39.1 41.8 41.7
Elderly (60 Years and Above) 7.0 5.0 5.8 4.3
Source: Demographic and Health Survey of Pakistan (2006).
Table 4
Imbalance between Revenue and Expenditure
in Countries at the Sub-national level
Revenue Expenditure
Australia 31 46
Brazil 31 46
Canada 56 63
India 34 55
South Korea 5 50
Germany 35 63
Pakistan 8 28
Source: Adapted from Watts (2005), cited in
Beaconhouse National University (2010).
Table 5
Revenue Transferred to Provinces
Revenue Transferred to Provinces
as Percentage of Gross Total
NFC Financial Year Revenue of Federal Government (%)
NFC 1991 1991-92 26.0
1992-93 26.1
1993-94 27.9
1994-95 30.1
1995-96 31.8
NFC 1996 1996-97 33.8
1997-98 26.3
1998-99 24.2
1999-00 27.4
2000-01 30.4
2001-02 27.7
2002-03 27.5
2003-04 27.8
2004-05 28.0
2005-06 29.5
NFC 2006 2006-07
2007-08
2008-09 31.4
2009-10 31.9
NFC 2009 2010-11 44.6
2011-12 * 44.0
