MID-COURSE REVIEW OF 
FISCAL DECENTRALIZATION: 
THE UNFINISHED AGENDA
Prepared by 
James S. McCullough, RTI
Emil Savov, LGI
Stefan Ivanov, Club Economika 2000
June 2003, Sofia
The Local Government Initiative is implemented by R.T.I. under USAID Contract 
number EEU-I-00-99-00014-00, The Local Government Assistance Initiative Services, 
IQC, task order 802.
2
MID-COURSE REVIEW OF FISCAL DECENTRALIZATION: 
THE UNFINISHED AGENDA
TABLE OF CONTENTS
1. Overview.............................................................................................................. 3
2. Municipal Expenditure Assignment....................................................... 5
2.1 Background on the Current Situation ....................................................................... 5
2.2 Remaining Problems in Expenditure Assignment.................................................... 6
2.2.1. Have all municipal services been appropriately classified? ................................. 7
2.2.2. How will the costing standards for mandated services be updated?..................... 8
2.2.3. How should capital costs of mandated services be incorporated into the costing 
standards?........................................................................................................................ 9
2.2.4 How can stability be ensured in the expenditure and revenue assignment? ........ 10
3. Funding Local Option Services.............................................................. 12
3.1. Analysis of the Current Situation........................................................................... 12
3.2 Avoiding Deficits in 2003....................................................................................... 13
3.3 The Relationship between Operating Costs and Capital Investment...................... 16
3.4 The Problem of Capital Investment ........................................................................ 16
3.5. Funding Greatly Increased Capital Investment at the Municipal Level ............... 19
4. Options for Increasing Municipal Own Source Revenues........ 23
4.1 Overview................................................................................................................. 23
4.2. Targets for Improving Local Tax Yields............................................................... 24
4.3. Narrowness and Lack of Buoyancy in the Municipal Tax Base............................ 25
4.4. Local Tax Rates, especially the Property Tax, are Low ........................................ 26
4.5. Lack of Local Control over Own Source Taxes .................................................... 28
4.6. A Note on Macro Economic and Fiscal Policy..................................................... 29
3
1. Overview
Bulgaria’s fiscal decentralization program is well underway and substantial progress has
been made. Most notably, the Fiscal Decentralization Working Group (FDWG) appointed 
by the Council of Ministers has crafted a far-sighted agenda and legislative reform is
underway although not yet completed. The government has adopted a new system of
intergovernmental transfers that is straightforward, transparent and linked directly to
expenditure needs of assigned services. The new system is predicated on a clear division 
of service responsibility between municipal services that are “mandated” by the central
government and those that are “local option”—i.e., provided at the discretion of the local 
government. 
The FDWG has developed costing standards for the mandated services that set the level
of transfers going to each municipality. While the costing standards are a good first step,
they were developed hastily and with some serious data gaps. Furthermore, there are
serious misunderstandings about how the standards should be used in terms of budget
guidance. The FDWG is considering a proposal to further upgrade the initial standards.
One contentious issue is that the costing standards do not include provision for the capital 
investment requirements of the services, an omission that further strains the municipal
budgets and results in under-investment in facilities and infrastructure. 
The new system of municipal finance places the responsibility for funding local option
services on local own source revenues. Over the past five years, the growth in own source 
revenues has been very strong but municipalities still only have partial control over
determining those revenues. The revenue base, especially the local tax base, remains
rather narrow and the municipalities cannot set their own tax rates. Despite the trend in
rapidly increasing local revenues, there is evidence that a large number municipalities do 
not have sufficient own source revenues to fund their local option services. In this report 
we diagnose this looming deficit in local option services funding and identify different
strategies to resolve it. 
The past eighteen months has seen a concentrated effort to resolve the problem of
accumulated arrears in municipal accounts going back to 1999. The central government
and municipalities have taken aggressive action to resolve this issue in 2002. The
situation has been greatly improved by the end of 2002 although the problem has not yet
been completely cleared up.
While progress was being made on these important fronts, the municipalities have been
experiencing a downward trend in overall levels of capital investment. The level of
capital investment in municipalities has been declining for most of the past decade. The
level of municipal capital investment is quite low by international standards and there is
evidence of a substantial backlog of unmet investment needs. The central government has 
committed to increasing capital investment at all levels and the EU accession will place a 
significant burden on Bulgarian municipalities to upgrade municipal infrastructure. Even
with the planned support from the EU pre-Accession grants, municipalities will still have 
to increase their own levels of investment considerably. A large part of this increased
4
investment will have to come from local own source revenues, both to pay for increased
current outlays and to support borrowing for capital projects. We suggest ways that the
level of municipal capital investment can be doubled but that requires a combination of
(a) increased local revenue generation, (b) enactment of municipal credit legislation and
(c) a commitment to use some grant funds (both central government and EU) to leverage 
municipal borrowing. 
We conclude this paper with a series of recommendations on ways to increase local
revenue generation. These include both additions to the municipal tax base and, more
importantly, ways to exploit the tax base already assigned to municipal governments.
Further strengthening local revenue generation will be a fitting capstone to Bulgaria’s
fiscal decentralization efforts.
5
2. Municipal Expenditure Assignment
2.1 Background on the Current Situation
The starting point of the government’s Fiscal Decentralization Concept and detailed
Work Program is to clarify the assignment of expenditure responsibility at the municipal
level. In order to establish a sound basis for the new system of intergovernmental
finances, the Government of Bulgaria has accepted the concept that all municipal
expenditures are defined as either (a) “mandated by the national government” or (b)
provided at the discretion, or local option, of the municipal government1
. In addition,
virtually all municipal services have both annual operating costs and capital investment
costs in their provision. 
The FDWG has applied costing standards to determine the minimum funding levels for
mandated services in four sectors: education, health, culture and social welfare. The
estimates of the funding requirement of mandated services in 2003 is approximately BGN 
1,124 million2
. We should note that this figure is a negotiated estimate based on past
expenditure patterns and not an analysis of spending needed to reach some specified level 
of performance in those services. Nevertheless, it represents a figure which the municipal 
sector representatives and the central government representatives agree is a workable
number.
The concept of a minimum funding level is important since it connotes that municipal
governments can actually spend more of their own resources on mandated services or
shift spending among services if
they so choose. It is also important
to note that this is a minimum
funding level and not expenditure 
dictate. The minimum State funding
levels only determine the level at
which the State budget must provide
transfers to the individual municipal
budgets—the municipal government 
still has the discretion to set its own
spending levels on different services 
depending on local need.
While the estimate of mandated service costs is tied to national standards, the estimate of 
costs for municipal local option services is not. Rather, spending on local option services 
is driven by the availability of local revenue and not on an estimate of needs.
 
1
 See “Concept and Program for Fiscal Decentralization” as prepared by the Fiscal Decentralization 
Working Group and approved by the Council of Ministers in June 2002.
2
 While the central government is committed to fully funding the standards, the State Budget Act did not 
obligate the full amount but left BGN 155 million to be “funded later in the year”. The Ministry of Finance 
has agreed in May 2003 to fund the balance in August and November 2003.
SERVICE COSTS STANDARDS
The mandated service standards set a
minimum funding level and not expenditure
dictate. The minimum State funding levels
only determine the level at which the State
budget must provide transfers to the individual 
municipal budgets— municipal governments
should still have the discretion to set their own 
spending levels on the different services
depending on local need.
6
Municipalities that have historically had high levels of revenues have spent heavily on
non-mandated services and also have had higher spending on mandated services as well3
. 
Much of the differences in municipal spending were the result of differences in shared
taxes and subsidies received as well as differences in own source revenue collections.
Under the proposed new structure of municipal finances, all local option services will be 
funded by local own source revenues. This shift places much greater emphasis on the
level of own source revenues and the ability of local governments to tap the local
economic base in order to generate local revenues. 
The FDWG has recognized that not all municipalities have the local economic base to
support an adequate level of local revenue generation. It has proposed a special subsidy
program—titled the Equalization Subsidy—to provide a minimum level of revenues,
apart from the mandated services funding. The Equalization Subsidy is intended to
guarantee that all municipalities have some funds for local discretionary spending
(including capital investment). Unfortunately, in 2003 the central government has been
unable to fund this Equalization Subsidy at the level recommended by the FDWG
although the grant program has been formally established in the revised legislation of the 
Municipal Budget Act4
. The FDWG proposed about BGN 90 million be allocated for
2003; The State Budget Act provides for only BGN 15 million.
While a significant percentage of municipalities do not have good local revenue potential, 
overall the municipalities have been increasing local revenue generation at a healthy pace 
over the past five years. Across the sum of all municipalities, local revenues have risen at
an average annual rate of 20% (in current leva), rising more than 70% from 1999 to 2002. 
Receipts from local taxes and local fees have both doubled in that period.
2.2 Remaining Problems in Expenditure Assignment
The FDWG has gone a long way toward resolving the expenditure assignment issues.
However, there remain several significant problems that require further attention:
· Have all municipal services been appropriately classified?
· How will the costing standards for mandated services be updated and used?
· How should capital costs of mandated services be incorporated into the costing 
standards?
· How can stability be ensured in the expenditure and revenue assignment?
The following sections discuss each of these issues in order.
 
3
 National legislation has historically placed a low ceiling on the percentage of municipal budget revenues 
that could be devoted to capital investment, effectively forcing even rich municipalities to forego 
significant capital investment outlays even if they had the resources to invest. 
4
 The revised Municipal Budget Act is still under Parliamentary review as of May 2003 but the State 
Budget Act of 2003 has a special provision establishing the Equalization Subsidy.
7
2.2.1. Have all municipal services been appropriately classified?
There was general agreement within the FDWG on the assignment of municipal service
expenditures into the two sub-groupings but with one exception: should “municipal
administration” be a mandated or local option service. The argument for classifying
municipal administration as a mandated service rests on three points: (a) It is a required
“overhead” cost of providing municipal services, most of which are in fact mandated
services; (b) local municipal administration is the cost of having local democracy which
is national policy and (c) the costs of municipal administration cannot drop below some
minimum level no matter how small the municipality—this means that very small
municipalities will have high per capita costs at the same time that they are likely to have 
the lowest level of own source revenues. The argument for excluding municipal
administration as a mandated service is that its costs are largely a matter of local
discretion and that it basically provides services of a local nature. In the end, the FDWG
decided to make municipal administration a “shared” service with the labor costs
included in the mandated cost standards and the maintenance support costs assigned as a
local option cost.
As a practical matter, evidence from analysis of the 2002 municipal expenditure data
shows that there is a substantial “floor level” of maintenance costs required for municipal 
administration no matter how small the locality. In 2002, the average level of
maintenance expenditures for municipal administration was about BGN 10 per capita
across all municipalities. For the municipalities that fall in the bottom quartile of all
municipalities in terms of own source revenues per capita, this represents almost half of
the median value of their local revenue collections (see Exhibit 1.) In fact for very small
municipalities, the per capita cost for maintenance support to municipal administration
often rises above the BGN 10 average figure. 
This analysis suggests that supporting municipal administration is a serious burden for
the very poor municipalities and that some form of relief should be considered. This may 
take several forms. First, all municipal administration expenditures may be shifted into
the mandated services category. Second, poor municipalities may be specifically targeted 
by an increase in the Equalization Subsidy. However, the Equalization Subsidy needs to
be funded at a level adequate to provide for this as well as an overall basic level of local 
option services. 
If we examine the experience of the poorest half of all Bulgarian municipalities, we see
that they spent an average of 40% of their local own source revenues on municipal
administration maintenance in 2002. The average per capita expenditure was BGN 11
and the local revenue per capita was BGN 27. The total expenditure on municipal
administration maintenance support for this group was BGN 44.7 million in 2002. While 
there is some variation from place to place in the actual expenditures per capita, it
appears that BGN 10-11 per capita is a reasonable target level for municipal
administration maintenance. If we take BGN 10 per capita as the target and assume that
this function were shifted from the local option to mandated service category, then the
8
increase in mandated services funding for all municipalities would be about BGN 82
million in 2003 (allowing for a 4% inflation adjustment.)
EXHIBIT 1. Median Levels of Own Source Revenues per Capita 
by Quartile for All Municipalities in 2002 
0
10
20
30
40
50
60
70
80
90
1 2 3 4
QUARTILES BASED ON LOCAL REVENUE/CAPITA
MEDIAN LOCAL REV/CAPITA
2.2.2. How will the costing standards for mandated services be updated?
The Fiscal Decentralization Program created the costing standards for the mandated
services but made no specific provision for updating or refining the initial standards. The 
FDWG is no considering a revision to the Program which will update the standards. 
The main reason for updating is that the first standards constructed were done under
extreme time pressure and simply used historical data on expenditures from 2001
adjusted to 2003. In a strict sense, they are not really standards –i.e., the level of
expenditure required to reach a desired output. They are actually a reallocation of existing 
expenditure patterns to make them conform to more reasonable unit costs of services
based on characteristics of the service populations. There was an attempt to take account
of differing cost factors across different types of localities but those adjustment factors
were based on fairly crude data. 
Given the manner in which the current standards were developed, it is prudent to evaluate 
how well they are performing as 2003 progresses and to see how well they actually fit the 
circumstances in the municipalities. Given the importance of education expenditures in
the municipal budgets, it would be best to start with a review of the education costing
standards. 
9
One critical issue is the continuing confusion over the use of the costing standards to
guide budgeting decisions at the local level. The original concept of the costing standards 
was to determine what an adequate level of central government financial support should
be to the municipal level. Unfortunately, some in the government have also taken this to
mean that the standards should become spending mandates to guide municipal budgeting. 
Clearly the local governments need information on what reasonable spending levels are
but the issuing of expenditure dictates is counter-productive. If the municipal
governments have no flexibility in making local spending choices, then they cannot
achieve any of the efficiencies of fiscal decentralization. This completely negates the
rationale for fiscal decentralization and will merely perpetuate the financial dependency
of the municipalities. This virtually guarantees that the municipalities will return for
future central government bailouts since they will not be in control of two thirds of their
local budgets. 
2.2.3. How should capital costs of mandated services be incorporated into the 
costing standards?
During the past decade of economic transition the assignment of all capital expenditure
has remained the responsibility of the central government, independently of the level of
government responsible for the provision of the services associated with the capital
infrastructure. This assignment has been guided either by the capacity to finance large
projects or by the belief that only central government officials are qualified to make
capital investment decisions.
The issue of capital investment has not, in fact, been a high priority topic in the policy 
debate on local government finance. Rather, the discussion over fiscal decentralization in
Bulgaria has largely focused on the issues of expenditure assignment (mandated versus
local option costs) and overall levels of funding while largely ignoring the question of
operating versus capital costs. In fact, the debates and negotiations have centered on
determining what the annual operating costs of mandated municipal services should be as 
a basis for setting the level of central-local transfers. Implicit in these negotiations is that 
the capital costs of all municipal services (both mandated and local option) will be funded 
by local own source revenues supplemented by some level of capital grants. 
As we discuss in the subsequent sections of this report, there is evidence that overall
capital investment at the municipal level is too low and in long term decline.
Furthermore, it appears that mandated services, while accounting for two thirds of the
operating costs of local government are receiving a very small fraction of total municipal 
investment. If the capital costs of mandated services are not provided for in the mandated 
service standards, then where will the resources come from?
10
2.2.4 How can stability be ensured in the expenditure and revenue assignment?
One of the main problems tackled by the FDWG has been the frequent changing of
expenditure and revenue assignment to municipalities. These changes have been caused
either by overall sectoral reforms (healthcare in 19995
 and social welfare funding in
2003), or by ad hoc budget decisions that add or remove certain costs of individual
programs (e.g., transportation of students.) 
The table below presents the value of changes in expenditure and revenue assignments of 
municipalities in the period 1999-2003, with 1998 as a baseline. The data for each year
constitutes the changes as compared to the preceding year. FY 2002 is the only year in
which expenditure assignments of municipalities are similar to the previous year. 
The changes in the expenditure structure of municipalities are due to various reasons.
What is common between them is that:
· They have been imposed by the central government (have not been a local choice) 
and 
· They have affected only the mandated expenditure responsibilities.
 
 EXHIBIT 2. CHANGES IN EXPENDITURE AND REVENUE 
ASSIGNMENTSTO MUNICIPAL GOVERNMENTS, 1998 TO 2003
(Thous. BGN)
1999 2000 2001 2002 2003 Est. 
Total budget expenditures6
2 022 316 2 178 114 2 113 850 2 388 488 1 941 000
Additional expenditure
responsibilities 30 578 26 865 101 231
Additional revenue sources 61 874 104 483
Dropped-out expenditure 
responsibilities 3 382 72 597 260 538 335 000
Dropped-out revenue sources 7 965 77 184 47 163 335 000
Such frequent changes cause a number of negative consequences including: impeding
budget planning, inequitably treating the users of basic services, and diverting the
attention of local governments to “plugging financial leaks” rather than providing
 
5
The introduction of a health insurance system began in 1999 ?. due to which a big portion of expenditure 
responsibilities borne by the municipalities till then, were taken over by the National Health Insurance
Fund. The expected expenditures, that were supposed to drop out of the municipal budgets step-by-step, 
were therefore deducted from the municipal revenues, while no account was taken of the chronic municipal 
budget deficit and the remaining expenditure responsibilities were not clearly identified. As a result, the
financial problems of the municipalities were further aggravated and the operating budget deficit jumped
up. 
6
 Data about municipal expenditures is valid for information purposes only.
11
efficient services. Such instability also impedes the development of a municipal credit
market since potential lenders find it hard to assess financial conditions and obligations in 
a constantly changing environment. It is hoped that, with the major change in funding
social welfare services completed in 2003, the level of municipal expenditure assignment 
will attain some degree stability for the foreseeable future.
12
3. Funding Local Option Services
3.1. Analysis of the Current Situation
The new system of intergovernmental finance requires that local option expenditures be
funded by local own source revenues. One complicating factor in this new alignment is
the treatment of capital investment. The current funding of mandated services only
includes operating costs so the central government has continued the practice of
providing separate targeted capital investment subsidies. However, those subsidies are
not tied to the level of mandated services and, in fact, appear to be inadequate to meet
the basic capital costs of the mandated services. Furthermore, their allocation is based on 
a totally different allocation scheme apart from the funding of mandated services
operating costs, so there is no linkage between the two (see below for an explanation of
the allocation of the targeted capital subsidies.)
The first question is whether local own source revenues will be adequate to meet the
costs of local option services. An examination of 2002 actual expenditure data reveals
some disturbing patterns. If we compare the 2002 local own source revenues and
expenditures on local option services, we find that about half of the municipalities would 
be running a deficit that totals BGN 135 million (see Exhibit 3.) 
EXHIBIT 3. DEFICITS BETWEEN LOCAL OPTION EXPENDITURES AND 
LOCAL OWN SOURCE REVENUES IN 2002 BY SUB-GROUP QUINTILES 
(BGN millions)
Number in ( ) is number municipalities in cell 
 EXPENDITURE LEVEL
HIGH MEDIUM LOW
A B C D E DEFICIT
HIGH
1 (18) 58.8 (4) 2.9 (0) 0 (0) 0 (0) 0 61.7
LEVEL
2 (6) 12.2 (12) 10.4 (17) 4.4 (3) 0.2 (0) 0 27.2
REV
3 (9) 5.9 (8) 2.3 (12) 4.7 (11) 1.5 (5) 0.4 14.8
OWN
4 (4) 2.0 (8) 2.8 (5) 2.6 (16) 4.0 (14) 1.7 13.1
LOW
5 (8) 4.5 (6) 2.8 (14) 4.1 (13) 4.6 (11) 2.2 18.2
DEFICIT 83.4 21.2 15.8 10.3 4.3 135.0
13
Exhibit 3 shows the pattern of deficits clearly by dividing all the municipalities into
subgroups in terms of their relative performance on levels of local revenue collection per 
capita and local service expenditure per capita. Each of the cells then shows how many
municipalities fall into that cell and the total of the deficit generated between local
revenues and local service expenditures for the municipalities in each cell. The Exhibit
clearly shows that the problem has different dimensions for different groups. Overall, the 
group with the highest revenues and highest expenditures (cell A1) accounts for over
40% of the total municipal deficit. These are municipalities who have very high revenues 
per capita but their spending exceeds even those high revenues. Clearly this group needs 
to bring their expenditures in line with their (already high) local revenues.
A second problematic group is found in the cluster of cells A4, A5 and B5 in the lower
lefthand corner. These are the very poorest municipalities in terms of revenue but have
high expenditures per capita. These municipalities, because they tend to be small, do not
account for a large proportion of the total deficit but clearly their shortfalls cannot be
solved simply by providing a modest level of additional revenue. This group contains a
number of localities in the mountainous regions of the southeast where service costs per
capita are high. 
3.2 Avoiding Deficits in 2003
The figures in Exhibit 3 show the “implied” deficits in the funding of local services—i.e., 
what the deficit would have been if the municipalities had been operating under the rules 
of the new system put in place beginning with 2003. In fact, deficits were reduced in the
aggregate across all municipalities in 2002 by increased subsidies from the central
government to settle the accumulated unpaid bills. The implied deficits in local services
funding was caused by the simple fact that expenditures on local services exceeded local 
own source revenues. The question remains, can local own source revenue grow fast
enough to cover the expenditures in 2003? And what happens to the municipalities who
have limited own source revenue potential?
Exhibit 4 shows a comparison of the sum of 2002 municipal expenditures with 2003
municipal budgets—the figures have been made comparable by removing expenditures
from 2002 which are being reassigned in 2003. The main revision, as noted above, is the 
removal of social welfare payments from the municipal budgets. That, together with
several minor other adjustments, is removing BGN 335 million from the municipal
expenditure assignment in 2003. 
There are several items in Exhibit 4 which deserve comment. Overall growth in total
municipal expenditures is kept to a very modest level (one half of one percent which is
well below the rate of inflation of 4%). This is done by actually reducing the operating
expenditures in local option services while increasing the capital investment. Given the
pattern of annual increases in local operating costs over the past few years, this may
prove to be too optimistic. On the other hand, as noted in Exhibit 3 above, we see there
14
are a number of high income municipalities that appear to be spending lavishly on some
local services. Certainly there appears to be room for some spending reductions.
EXHIBIT 4. COMPARISON OF 2003 MUNICIPAL BUDGETS
WITH 2002 ACTUAL EXPENDITURES
MUNICIPAL EXPENDITURES 2002 ACTUAL 2003 BUDGET
A. MANDATED SERVICES 1,146 1,173
 A.1 OPERATING COSTS 1,080 1,148
 A.2. CAPITAL COSTS 66 25
B. LOCAL SERVICES 783 768
 B.1. OPERATING COSTS 653 595
 B.2. CAPITAL COSTS 130 173
TOTAL EXPENDITURES 1,929 1,941
Notes on Ex. 4: 2002 Actual data from Ministry of Finance adjusted by 
(a) reduction of service expenditures that are removed from 2003 
municipal expenditure assignment and (b) expenditures on unpaid bills 
from previous years. 2003 Budget data from 257 of 263 municipalities 
reporting to National Association of Municipalities, adjusted to add the 
missing 6 municipalities by their relative weight in total municipal 
expenditures in 2002. 
Exhibit 4 contains capital costs allocation between mandated and local services that are
somewhat speculative. We do know the total expenditures on mandated and local
services and the total capital costs by municipalities are but we do not have a breakdown 
of capital and operating costs between mandated and local services in the 2002 actual
data. In 2002 we estimated the breakdown of capital costs by applying the operating costs 
standards of the FDWG to determine the estimated operating costs of mandated services; 
we then subtracted this from the total costs of mandated services to estimate the residual
capital costs. For 2003, we use the reported amounts from the municipal budgets. We
should note that the budgeted amount for capital investment for mandated services in
2003 appears to be very low. 
The changing assumptions about capital costs of the mandated and local services should 
not be taken too seriously at this time. However, as we discuss below, the fact that capital 
costs have been omitted from the costing standards of the mandated services is likely to
produce a serious shortfall in both the level of capital investment in those services and a 
shortfall in municipal budgets overall. We address this issue is the next section.
15
Exhibit 5 shows a comparison of municipal revenues and expenditures in the 2003
budgets. The revenues are comprised of the transfers from the central government to
cover the operating costs of the mandated services, local own source revenues (local
taxes, local fees and other local revenue), targeted capital grants and Equalization
Subsidy for poor municipalities. 
EXHIBIT 5. COMPARISON OF REVENUE AND EXPENDITURES
 IN 2003 MUNICIPAL BUDGETS
 EXPEND/REV TYPE EXPENDITURE REVENUE
A. MANDATED SERVICES (TOTAL) 1,173 1,124
 A.1 OPERATING COSTS 1,148 1124
 A.2. CAPITAL COSTS 25 (inc. below)
B. LOCAL SERVICES (TOTAL) 768 817
 B.1. OPERATING COSTS 595
 B.2. CAPITAL COSTS 173
OWN SOURCE REVENUE 623
CAPITAL GRANTS 100
EQUALIZATION GRANTS 15
BORROWING 79
TOTAL 1,941 1,941
On the surface, it looks as if the municipal budgets are in balance. However, this
“balance” has been achieved by making two dubious assumptions: (a) the operating costs 
of local option services can be significantly reduced from 2002 levels and (b) revenues
can be augmented by loans totaling almost BGN 80 million. In addition, the
municipalities forecast a continued strong growth in own source revenues amounting to a 
12% increase over the 2002 level. While this is certainly consistent with recent trends in
local revenue growth, we should note that no new revenue sources have been assigned to 
municipalities in 2003 so that growth will have to come largely from increases in local
fees and other non-tax revenues—local tax revenues are expected to grow no more than
about 5-6% in 2003. 
If we subtract the municipal loans from the revenue side and revise the operating costs of 
local services to be the same as 2002 (with 4% inflation), then we would have a potential 
deficit in the local services of BGN 163 million. We are not predicting that this deficit
will occur but this must be considered as a possibility based on past performance. 
The potential funding gap is one of the reasons that municipalities need the flexibility to
adjust their spending across different services. Any efficiencies gained in expenditures on 
16
the mandated services can be used to offset higher expenditures on the local option
services. This potential funding gap is also a good reason to focus attention on the sub￾group of municipalities that have the highest potential for incurring substantial deficits –
the group with high local revenues and even higher local expenditures per capita as
identified in Exhibit 3 above. This group accounts for over 40% of the “implied” deficit
in 2002 and should be the target for operating cost reductions.
3.3 The Relationship between Operating Costs and Capital Investment 
All of the growth in local own source revenues over the past few years has gone into
meeting the operating costs of local option and mandated services. A review of municipal 
budget data from 1998 to 2002 shows that municipal own source revenues devoted to
capital investment declined steadily from 1998 through 2001 and began to rebound only
in 2002—see Exhibit 6 below. This means that the operating cost requirements of
municipal services have been growing faster than revenues overall, in effect decreasing
the amounts available for new capital investment. We discuss the impacts of this decline
in capital investment in the following section.
EXHIBIT 6
COMPOSITION OF MUNICIPAL CAPITAL INVESTMENT
(BGN MILLIONS)
SOURCE 1998 1999 2000 2001 2002
Muni Own
Revenue 144 98 82 48 57
Central 
Grants 68 75 76 74 139
TOTAL 212 173 158 122 196`
SOURCE: REPORTED EXPENDITURES OF ALL MUNICIPALITIES
3.4 The Problem of Capital Investment
Exhibit 6 above shows that municipalities have been able to devote less and less
resources to capital investment over the past five years. In fact, capital investment at the
municipal level has been declining for most of the past decade. Ignoring the capital
investment requirements of municipal services can be dangerous. It is the nature of
capital costs that, while they can often be deferred, they do not go away but continue to
17
accumulate. There is strong evidence that Bulgarian municipalities have been
systematically under-funding their capital infrastructure for a number of years. 
While there is no comprehensive study of municipal capital investment needs today, there 
is ample evidence that the level of municipal capital investment has been too low for a
number of years and that a sizeable backlog of investment needs has accumulated. The
World Bank has completed a study to estimate the financing requirements for Bulgaria to 
meet the environmental standards of the EU Accession, covering both capital and
operating costs—at the municipal and national levels7
. The World Bank study produced a 
range of estimates from low to high. The Bank’s low estimate for the municipalities to
upgrade environmental infrastructure to the EU standards amounted to just over BGN
1,000 million per year in capital costs—an amount almost equivalent to the total of the
operating costs of all mandated municipal services in 2003. The World Bank’s high
estimate is about BGN1,850 million in capital costs per year or close to the forecasted
total of all municipal budgets in 2003. This is a staggering amount, given that total
municipal capital investment in 2002 was not quite BGN 200 million—or about one fifth 
of the World Bank’s low estimate.
Exhibit 6 above showed that the level of municipal capital investment has been in long
term decline. That level was low to begin with by international standards and by
comparison to other eastern European countries. In general, we would expect to see about 
15-20% of total municipal budgets devoted to capital investment on average if Bulgaria
was meeting the benchmark of other countries. We should note that in some of the
countries of the CEE region, that figure has been higher as municipalities have made up 
for past investment shortfalls. Exhibit 7 shows the declining percentage of municipal
budgets devoted to capital investment over the past five years compared to the
international benchmark. Even the increase in 2002 has only brought Bulgaria up to the
10% mark, still considerably below the 15% benchmark. 
 
7
 World Bank, Bulgaria: The Dual Challenge of Transition and Accession. A World Bank Country 
Study, February 2001.
18
Exhibit 7
PERCENTAGE OF MUNICIPAL BUDGETS DEVOTED 
TO CAPITAL INVESTMENT 
IN 2003 (BGN Millions)
12.7%
9.3%
7.9%
6.5%
10.2%
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
16.0%
18.0%
20.0%
1998 1999 2000 2001 2002
INTERNATIONAL BENCHMARK
In summary, the level of capital investment in Bulgarian municipalities is low and has
been falling. It is clear, as shown by the World Bank study, that this has produced a
substantial deficit in the capital infrastructure at the local level that must be made up in
some way. The question now is whether we can estimate what an adequate level of
capital investment should be at the municipal level. 
There is no really scientific way to determine the capital investment needs across
municipalities short of a thorough analysis such as the World Bank study cited above.
However, we can derive some general estimates by simpler means in order to tell us the
general magnitude of the problem. We do this by using as a starting point the 
international benchmark of 15% of municipal budgets that would normally be devoted to 
capital investment. If we assume that the operating cost estimates for 2003 shown in
Exhibit 5 above are reasonable, then we can derive the capital investment requirements as 
a function of the operating costs (see Exhibit 8). 
19
The figures derived in Exhibit 8 are the annual investment levels that we would expect to 
see across all municipalities if the 15% benchmark were met. It does not address any
backlog in capital investment that may have accumulated as a result of prior years’ under￾investment. Since we know that investment levels have been falling for a number of
years, we can safely assume that this backlog is substantial. In an earlier report, the
authors have estimated that the backlog to be at least BGN 1,700 million just by
examining the imputed shortfall in municipal capital investment over the past decade8
. If 
we combine the results of Exhibit 8 above with an acknowledgement that the capital
investment backlog needs to
be reduced incrementally,
then we must conclude that
capital investment spending
at the municipal level should
at least be doubled—i.e., at a
level of about BGN 400
million per year. A more
precise (and probably higher) 
figure can undoubtedly be
calculated but for our
purposes now, the target of
doubling municipal capital
investment is a reasonable 
conclusion. The next
question is, where will the funding come from to make that goal possible?
3.5. Funding Greatly Increased Capital Investment at the Municipal Level
Financial resources for funding additional capital investment at the municipal level are
quite limited under current conditions. The central government was not able to fully fund 
the agreed upon mandated service costs in the State Budget Act for 2003 nor was it able
to fund the Equalization Grants for poor municipalities at a level even close to that
recommended by the FDWG. On the other hand, the IMF and Ministry of Finance
forecast a continued improvement in economic growth which should produce a robust
growth in Personal Income Tax receipts – now the primary source of transfers to
 
8
 This estimate is the accumulated capital investment backlog associated with routine 
municipal operations and does not include major upgrading of the local infrastructure 
networks. See James S. McCullough, “Municipal Capital Investment Policy Considerations” Local 
Government Initiative, Sofia. March 2003.
EXHIBIT 8
CAPITAL INVESTMENT ESTIMATE 
FOR 2003
1. OPERATING COST = BGN 1743 M
2. CAP. COST = 15/85 or 17.65% of 
OPERATING COSTS
3. CAP COSTS = BGN 308 M
20
municipalities. However, proposed cuts in tax rates are likely to keep the PIT yields from 
growing very fast although it is difficult to know just what the rate cuts will be and what 
the overall yields will actually be in the future. 
Increased funding from the central government is most likely to be devoted first to fully
funding the operating costs of mandated services and the Equalization Subsidy for the
poor municipalities. There is a good argument to be made that the costs of mandated
services should include capital costs as well as operating costs. If we apply our 15% rule
of thumb in estimating the annual capital costs that attend the mandated services costs of 
approximately BGN 1,124 million in 2003, then the transfers would have to be increased 
by about BGN 195 million to incorporate capital costs. This alone would almost double
the capital investment levels in the municipalities. However, given the shortfalls in
funding the lower, agreed-upon levels of mandated service costs in 2003, such increased
funding is probably not likely to be forthcoming.
Many municipal officials place their hopes for increased capital investment on two new
sources: borrowing and EU pre-Accession grants. As we have noted in another report9
, 
the EU grants will mainly go to (a) large scale environmental infrastructure and transport 
projects in some selected locations and (b) small scale infrastructure projects in the poor,
rural municipalities. The EU grant funding, as currently designed, is not really intended
to support general capital investment programming across the municipal spectrum. 
There is the potential for municipal borrowing but the main problem is on the demand
side: municipal governments are generally not credit worthy because of their poor
financial conditions. As we documented above, municipalities have been devoting more
and more of their own source revenues simply to cover annual operating expenses. Until
municipalities can begin to generate operating revenue surpluses for new investment and 
debt service, they will not be credit worthy and hence unable to borrow. 
Clearly, if municipal borrowing is to become an important source of capital investment
funding, then the municipalities must accelerate their own source revenue generation.
The central government now has some fairly clear-cut policy options with respect to
financing an increased level of capital investment by municipalities:
· The central government can increase transfers and/or
· The municipal governments can increase local own source revenues going into 
capital investment. 
 To make this policy choice it is important to distinguish between those municipalities
that have good revenue raising capacity and those that do not. The poorer municipalities
are not good candidates for greatly increasing local revenue generation and should not be 
expected to generate substantial own source revenues for capital investment. They should 
 
9
 See James S. McCullough, Ibid.
21
certainly not be encouraged to borrow10. Since these municipalities do need access to
capital investment funding, some provision must be made to provide those funds, 
possibly through the Equalization Grant mechanism already proposed. In the near term,
additional grant funding from the EU SAPARD program and the new World Bank Social 
Investment Fund will be directing a substantial amount of infrastructure investment into
the poor, rural municipalities. If evenly spread across the very poorest municipalities,
these two programs will provide a large measure of the capital investment “safety net” for 
those municipalities over the next 4-5 years. Beyond that, the central government does
need a national policy on how the capital investment funding needs of that group will be
met once the donor programs are finished.
The second grouping of municipalities—those with local revenue raising potential—
present a strategic opportunity for the government. This is the group where increased
capital investment can be funded by increased own source revenues on three conditions:
· The municipalities receive increased revenue raising authority covering both taxes 
and fees;
· The proposed municipal credit legislation be put in place to permit development 
of the municipal credit market; and
· The central government utilize the central grants and donor grant funds 
strategically to stimulate municipal borrowing.
The first element of the strategy is to have increased own source revenue that can be
dedicated to additional capital investment. The increased revenue will be needed for
additional “pay-as-you-go” investments plus debt service on borrowed funds. This means 
that municipalities must have increased own source revenue raising authority. 
The second element of the strategy is to facilitate municipal borrowing through adoption
of the proposed municipal credit legislation11. The main question now is whether
additional steps need to be taken beyond adopting the legal foundation – such steps can
range from a fairly modest program to facilitate the initial learning process between
borrowers and lenders to a much more interventionist approach such as creating a
municipal lending agency. 
The third element of the strategy is to use the central government and donor grants to
stimulate use of credit by municipalities with revenue raising potential. Tying grants and
loans together can have the advantage of ‘leveraging” the grant funds—that is, the grants 
may make the total project financially feasible and therefore enable the recipient
municipality to borrow the needed funds to undertake the project.
 
10 Experience with municipal loan programs elsewhere in the world shows that when non-credit worthy 
municipalities are encouraged to borrow, they rapidly undercut the whole system. 
11 See Michael DeAngelis, Ronald W. Johnson, & Christopher Z. Kaczmarski, Building the Municipal Credit 
Market for Infrastructure Finance. RTI International, September 2002.
22
The strategy can use different mixes of grant funds and own source revenues to fund
capital investment at the local level. What is important is the principle that central grant
funds should be used strategically. If grant funds are provided to municipalities with
untapped local revenue raising potential, then those grants should be used to stimulate
local revenue generation. 
23
4. Options for Increasing Municipal Own Source Revenues
4.1 Overview
Bulgarian municipalities are assigned local revenue sources that fall into three main 
groups:
- Local taxes
- Local fees, and
- Non-tax revenues.
As noted earlier, all of these revenue sources have been growing rapidly over the past
five years and now constitute almost a quarter of total municipal budgets (see Exhibit 9). 
However, if the municipalities are to fund their local option services and generate
increased capital investment, local own source revenues will have to grow even more.
EXHIBIT 9. Structure of Municipal Own Source Revenue in 2002
Revenue Type Share of total revenue Share of local revenue
Local taxes 7.0 % 30.5 %
Local fees 9.4 % 41.0 %
Non-tax revenues 6.5 % 28.5 %
Total: 22.9 % 100.0 %
The structure of own source revenue found in Bulgarian municipalities is quite different
from that of most other European countries. On average, western European municipalities 
receive over one third of their total revenues from local taxes while local fees make up
about one sixth12. Bulgaria is unusual in both the low contribution made by local taxes
and the fact that local fees outweigh local tax receipts.
The central government has recently taken steps to turn over control of local fees entirely 
to the municipal governments. An amendment to the Local Taxes and Fees Act, effective 
from January 1, 2003, gives the municipalities the authority to set the amount of fees, the 
methodology for determining the fee schedules, the definition of exemptions and the
collection procedures. 
 
12Jeanine Corvetto, Tax and Fee Authority of Local Governments in Western Europe. 
Research Triangle Institute. May 2000. 
24
Non-tax revenues are generated mainly through transactions in municipal property
(rentals, sales, concessions) and are used as a rule to finance capital expenditures. Given 
their nature, they cannot be viewed as stable revenues—either in size or future growth
potential.
Local taxes are comprised of five separate taxes as defined in the Local Taxes and Fees
Act. They include: property tax, vehicle tax, inheritance tax, gift tax and a new (in 2002) 
“Road Tax”. The Road Tax is actually a surcharge placed on the vehicle tax but its
purpose is to help fund road repairs so it is treated as a separate levy. We should also note 
that the highest yielding local “fee”—the garbage collection fee on residential
households—is actually a surcharge on the property tax so there is some mixing of the
local tax and fee base in this instance. 
The most notable aspect of the local taxes, aside from their low contribution to municipal 
revenues is the fact that local governments have no control over them. Municipalities
cannot define the tax base, have no say in setting the tax rates and play no role in tax
collection. In effect, local taxes are “local” primarily in name only and the fact that their
proceeds can be spent as the municipal council sees fit.
Local taxes are the subject of the remainder of this section since that is the area of local
revenue collection where attention needs to be focused. If local revenues are to be
substantially increased, then local taxes should play a much greater role. Control over
fees has already been decentralized to the municipal governments.
4.2. Targets for Improving Local Tax Yields
There are several issues with respect to the local taxes that deserve attention. First, the
local tax base is rather narrow since there are no taxes assigned to municipal government 
that are related specifically to local economic activity. This means that the tax base is not 
very buoyant—i.e., it grows as economic activity in the locality grows. 
A second problem with the Bulgarian municipal tax base is that the tax rates, especially
the property tax, are very low. Municipal governments around the world tend to rely on
the property tax as the major local tax. The property tax is a “good” local tax for several
important reasons. It is stable from years to year. The tax base is readily identifiable and
the computation of tax liability can be quite straightforward. In the USA and some
European countries, the property tax is the main source of debt repayment for general
obligation borrowing—the stability of property tax revenue makes it a solid base for local 
government credit worthiness. In Bulgaria, the property tax is a minor revenue source.
Because of this, Bulgarian municipalities overall receive a relatively small share of their
total revenues from local taxes.
The third main problem with the local tax system in Bulgaria is the fact that local
governments cannot adjust the tax bases or rate schedules to local conditions. The central 
government has, in fact, attempted to grant greater local control over taxation to
25
municipalities but has been blocked by the Constitutional Court. In an effort to keep tax 
rates acceptable across the entire country, the central government has set the rates too low 
for many localities. 
If Bulgarian municipalities are to continue their growth in local own source revenues,
then these three sets of problems must be resolved. In the following sections, we address
them each in turn.
4.3. Narrowness and Lack of Buoyancy in the Municipal Tax Base
Bulgarian municipalities need a broader and more buoyant tax base. This can be readily
accomplished by two means. First, the Patent Tax, which is now part of the Personal
Income Tax , can be transferred to the municipal own source revenue base. Second,
municipalities can be allowed to place a surcharge on one of the national tax bases which 
is buoyant such as the VAT or Personal Income Tax. 
Transferring the Patent Tax—which is a tax on various categories of small business
activity—into a municipal tax would increase the local revenue base by approximately
BGN 60 million. As a result, the revenues from the Patent Tax would comprise 10% of
the aggregate municipal own source revenues, while for some municipalities this share
could constitute up to about 20%. 
The Patent Tax , by its nature, is well suited to be a local rather than national tax. Indeed, 
when the Patent Tax was originally introduced in 1997, it was intended to be a locally
administered tax. However, the Constitutional Court declared this as unconstitutional13
in 1998 and it was incorporated into the national income tax code. The Patent Tax rates
are set uniformly across the nation on small businesses. The rates are not adjusted for
differences in local conditions. This lack of adaptability has resulted in some instances of 
over-taxation and, much more frequently, under-taxation in specific localities. Since the
Patent Tax is now part of the Personal Income Tax, revising the Patent Tax schedules
requires revising the income tax law in the national Parliament. 
Shifting the Patent Tax to municipal control would increase local tax revenues but would 
not resolve the rate setting dilemma until a Constitutional Amendment is approved. In the 
meantime, the National Parliament could follow the letter of the law by granting blanket
approval to locally set tax rates in the annual State Budget Act. Such a process would
have to be streamlined so that the Parliament could adopt local rate schedules en masse
rather than have to take up each submission individually. In some countries where the
latter has been tried, the process has become so cumbersome that it has ground to a halt.
A second way to broaden the municipal tax base is to grant local governments the limited 
ability to apply surcharges on the national taxes of VAT and/or personal income tax. We
recognize that such authority runs against the current stated policies of the central
 
13 Finding No 6 of the Constitutional Court of 1998.
26
government to reduce those tax rates14. However, in the interest of full consideration of
alternatives, we suggest that these options should be examined. Furthermore, we should
point out that such surcharges are common among western European countries and
should not be dismissed out of hand.
Under the surcharge approach, the central government defines the tax base and collects
both its own tax and surcharges set by the local governments. This approach ideally
avoids the problems that occur when different governments define the tax base in
conflicting ways and administer the tax in different ways. Because of their power to set
surcharge rates, the local government retains the most important attribute of fiscal
sovereignty in the tax field—i.e., the ability to set the tax rate. Of course, to effect the
surcharge system would require adopting the Constitutional Amendment to permit
municipalities to set their own tax rates. 
All discussion of tax rate setting inevitably comes back to the issue of the Constitutional
Amendment. At this point, we should note that it would be wise to construct the
Constitutional Amendment in such a way as to permit a wide range of possible types of
local taxing authority in the future. That is, the Amendment should not be so narrowly
drawn that it precludes practices in local taxation that are common today among the
European countries. 
4.4. Local Tax Rates, especially the Property Tax, are Low
The need to maintain a set of national tax rates for municipal taxes has resulted in rates
that are too low for many municipalities. Those municipalities that have stronger
economic bases and higher income citizens should be able to raise more revenues locally. 
The property tax is a good case in point. Earlier studies have shown that the tax rates for 
the property tax are extremely low by international standards15. Furthermore, because of
the high level of exemptions, these low rates are effectively reduced to an even lower
level when the actual tax liability is calculated. 
Clearly, the Constitutional Amendment is needed to allow municipal governments some
discretion in setting local tax rates. This is the highest priority. The Amendment should
cover both the setting of the rates and the ability to define the tax base, especially the
limiting of exemptions.
 
14 See “Supplementary Memorandum of Economic Policies of the Government of Bulgaria and the 
Bulgarian National Bank” submitted by the Minister of Finance to the IMF. Contained as an Appendix in 
the IMF Report Bulgaria: Second Review under the Stand-by Arrangement and Request for Waiver of 
Applicability of Performance Criteria. Washington, DC. January 22, 2003.) We note that government 
policy to reduce tax rates and shrink the government sector relative to GDP does not mean that government 
revenues will decline—see NOTE at end of this paper. 
15 See Stefan Ivanov and James S. McCullough, “The Potential and Performance of the Property Tax”, 
Plovdiv (Bulgaria) Fiscal Decentralization Study. East European Housing Sector Assistance Project 
(USAID) under subcontract to The Urban Institute, Washington, DC. October 1998.
27
The property tax deserves special attention because it is such an underutilized revenue
source at present. The property tax also should become more important as the municipal
credit market develops. As noted earlier, property tax revenues are often a prime source
for debt repayment of general obligation borrowing by municipal governments in other
countries. If Bulgaria embarks on a serious municipal credit program, then the property
tax could become a critical revenue source for guaranteeing repayment. 
Making the property tax a more important revenue source begins with increasing the tax 
rates but goes well beyond that first step. Indeed, the strategy for increasing property tax 
yields should not start with a single large tax increase but with adoption of a long term
strategy of gradually raising rates over a number of years. Our earlier study of the
property tax performance in Plovdiv has revealed that the property tax rolls are relatively 
complete and that the rate of tax collection is very high16. There is room for improvement 
in keeping the tax rolls up-to-date –i.e., recording modifications to property that increases 
its taxable value (see Box below). And, there is some indication now that in the larger
cities, property values are rising fast and that the increased value is not being captured by
the property tax system. 
The property tax is
currently being
administered by the
Regional Tax Offices as
are all of the municipal
taxes. To transfer
administration to the
municipal governments
will require the creation of
new tax records systems
(property registry,
assessment/valuation, 
billing and collection) that
will likely strain the
existing capabilities of
municipal finance offices.
The cost of such transfers
will not be small and that is 
one more reason that
property tax rates should
rise—without increased
yield from the tax, the cost
of transferring the
administration to the
municipality will likely
exceed any benefit from
 
16 Ivanov and McCullough, Ibid.
Using Spot Inspections to Improve 
Property Tax Listings
The Plovdiv Property Tax Study team developed a 
cost-effective sample survey (inspection) approach 
which could be used in other cities to identify properties 
not on the tax rolls as well as errors in the tax listings. 
The sample inspection covered a n area of the city 
which contained slightly less than 1% of the total 
households and cost BGN 1,100 to carry out.
We estimate that a set of 10 similar sample inspections 
would cost approximately BGN 6,500 and should 
produce a minimum of BGN 45,000 in additional tax 
revenues alone for Plovdiv. In addition, since the 
garbage collection fee for households is also based on 
the property tax base, any increase in property valuation 
results in a higher yield of the garbage fee. For Plovdiv, 
the impact of the proposed sample inspection program 
should achieve a payoff ratio of more than one to 
twenty - i.e., for every one leva spent on inspections, the 
increase in property tax and garbage fee revenue 
combined would be twenty leva.. 
28
increased tax collection17
. 
Because the property tax has the potential for becoming a much more important tax and
could play a critical role in development of the municipal credit system, the government
should consider a pilot program in this area. Such a program would develop prototype tax 
administration systems for municipal administration and strategies for popular
acceptance. Such a pilot program could be logically linked to a municipal credit program 
as noted above. There is a great deal of experience worldwide with the design and
implementation of property tax programs. Since the experience is quite varied in terms of 
success and failure, the design of such a program for Bulgaria should pay careful
attention to the lessons learned in the other programs.
4.5. Lack of Local Control over Own Source Taxes
The issue of local control should really be viewed as three separate issues: 
· control over defining the tax base (including the granting of exemptions);
· control over setting the tax rates; and 
· responsibility for administering the tax systems.
Much of the debate to-date has focused on the second issue—rate setting—with the
assumption that all three issues are an inseparable package. In fact, there is no reason that 
municipal governments should not be granted varying levels of control at present. For
example, municipalities could be granted immediate control over the granting of
exemptions in the property tax and deciding when revaluation of the property tax base is 
appropriate. 
Clearly, the granting of authority for tax rate setting will require the Constitutional
Amendment. However, even in this area, we have noted earlier that the Parliament could 
devise systems for allowing municipalities to set some tax rates as long as Parliament
ratified those rates in national legislation. 
The transfer of tax administration from the Regional Tax Offices to the municipalities
can also be done on a piecemeal basis. That is, there is no reason that the transfer has to
happen simultaneously in all municipalities. Indeed, for many small municipalities it may 
not be cost effective for them to take over tax administration at all. This is a function that 
can be done at local government discretion.
 
17 See Ivanov and McCullough, Ibid. for an analysis of these costs and tradeoffs.
29
4.6. A Note on Macro Economic and Fiscal Policy
One of the stated policies of the central government is to reduce the size of government
spending relative to GDP. The recent IMF report cited above contains the policy targets
of the government in reducing the percentage of government revenues to GDP from 36% 
in 2002 to 34.9% in 200618. This does not mean that government revenues will decrease
during that period. Since real GDP growth is predicted fairly strong at 5.5% over that
period, the growth in GDP will more than compensate for the decrease in share of GDP
accounted for by the government. Indeed, government revenues are forecasted to grow
between 4-5% annually in real terms during that period (or about 8-9 % in nominal
terms.) 
 
18 See previously cited IMF Report, Table 7: Bulgaria. Macro-economic Framework 2002-2007.