HARVARD UNIVERSITY
*
AGENCY FOR ECONOMIC ANALYSIS AND FORECASTING
*
INSTITUTE FOR MARKET ECONOMICS
THE SHADOW ECONOMY IN BULGARIA
Cornell University
Steven Kyle Ph.D.
Harvard University
Andrew Warner Ph.D.
Agency for Economic Analysis and Forecasting
Lubomir Dimitrov
Radoslav Krustev
Institute for Market Economics
Svetlana Alexandrova Ph.D.
Krassen Stanchev Ph.D.
Gergi Stoev
SOFIA
APRIL 2001
ii
This work was supported by USAID’s consulting assistance on economic
reform (CAER II Project). The objectives of the project are to contribute to
broad-based and sustainable economic growth and to improve the policy
reform content of USAID assistance activities that aim to strengthen markets in
recipient countries. Services were provided by the Harvard Institute for
International Development (HIID) and its subcontractors. It was funded by the
U.S. Agency for International Development, Bureau for Global Programs,
Field Support and Research, Center for Economic Growth and Agricultural
Development, Office of Emerging Markets, contract PCE-C-00-95-00015-00,
Task Order #39. Work subsequent to the conclusion of this contract was
performed at Cornell University.
iii
Executive Summary
“There cannot be any question that the underground economy is a real phenomenon with important
implications that deserve attention and study."
Vito Tanzi
Measuring the Shadow Economy in Bulgaria1
GDP accounts are customarily compiled in several alternative ways, each focusing on
aggregating transactions in different ways, but all (at least in theory) adding to the same total.
Two of the most common aggregations are that focused on expenditures (based on the
standard national income accounting identity of C + I + G + X-M) and that based on revenues,
or incomes. The two methods should, of course, add to the same number since they measure
different sides of the same activity: what money people receive on the one side, and what they
do with it on the other.
However, Bulgarian GDP statistics using revenue as the approach give growth
rates 2 percentage points lower than the expenditure approach for 1998 and 1999. In other
words, data based on what people actually spend show growth rates of 5.4% (1998) and 4.4%
(1999), while official figures based on revenues are 3.5% and 2.4%, respectively. This is
evidence that there are underreported incomes. It is of interest not only for statistical but also
for economic policy purposes to have more detailed information about the discrepancies
between official statistics and activities not covered by the official statistical system. It is
particularly interesting to know the size and structure of unreported, hidden economic
activities, or what has come to be called the “shadow economy.” Currently published
estimates of the size of the shadow economy vary from 20 to 25% of officially measured
GDP, implying that there is a far larger issue than that implied by the differential growth rates
cited above.
The objective of this study is to estimate the size of the informal sector, its structure,
the incentives for its development and its effect on the economic growth and the
competitiveness of the Bulgarian economy. Two different methods were used to get results
that are compatible for international comparisons; also, alternative calculations allow a range
of estimates which can help to balance the methodological weaknesses of the individual
approaches: the modified (indirect) Physical Input Approach based on energy/electricity
consumption data, and the (direct) Microeconomic Approach based on firm level survey data.
1
The team would like to extend special gratitude to the experts from the Agency for Socio-Economic Analysis,
especially Docho Mihaylov, Director for carrying out the field work and initial analysis, Ms Evdokiya Nikolova a student at Harvard University, for contributing to the first chapter of the report, as well as to Dr. Friedrich
Bauersachs, Senior Economist at the Institute for Market Economics for his comprehensive consultations during
preparation of the final report.
iv
Five hundred thirty firms are covered by the survey taken from the following sectors:
wholesale and retail trade, transport, construction, mechanical and engineering, food and
drinks production, chemical industry, textile and knitwear production, tourism and agriculture.
These sectors account for 56% of total GDP, and while limited resources necessitated a
smaller sample than would be desired under conditions of unlimited funding, some results are
quite interesting and indicate the need for further investigation.
The study not only provided estimates of the size of the informal economy but also
allowed a basis for analyzing some of the most important aspects of its underlying structure
and the incentives for its growth. Of particular interest are effects of the tax and social
insurance system, effects of labor contracting and wage level as well as the effects of
administrative costs.
Results of the Modified Physical Input Approach
The basic rationale of Physical Input Approaches to measuring the size of the shadow
economy is that energy consumption (electricity, plus other sources) in a given country is
proportional to total economic activity and any change in energy consumption which does not
correspond to changes in the measured total activity level of the country indicates a change in
the size of the shadow economy. These results provide useful indicators of changes in the
shadow economy over time, but cannot be used to quantify the absolute size of the shadow
economy since this depends on an initial estimate of the size of the shadow economy in the
base year. This estimate is necessarily arbitrary to some degree in the absence of specific
micro-level data allowing definition of an explicit relationship between energy use and
economic activity. Results show that the Bulgarian shadow economy in 1998 declined below
the estimated base year (1989) share of 30%. According to our calculations the share of the
shadow economy in 1998 GDP in Bulgaria was 22%. The largest shares were observed in
1990 (32.2%) and 1996 (34.4%), declining thereafter.
Results of the Microeconomic Approach
The survey performed provided a description of some of the factors important in promoting
shadow economy activities as well as two different direct indicators of its size: one based on
tax evasion and one on unreported wages.
The Microeconomic Approach takes into consideration the business environment insofar as it
is the result of legislation, government action and institutional gaps and the response of the
individual firms to these factors. The general economic environment has a serious impact on
economic activities, particularly the tax and social insurance system and employment and
wage conditions.
The survey shows that most companies do not use bank credit for initial investments as is the
practice in developed market economies. Personal savings comprise 65% of initial
investments, bank credit accounts for only 18%, and financial resources available through
international programs account for less than 3%.
v
Almost 70% of the enterprises prefer paying for inputs and other costs in cash. Nineteen
percent cite the greater freedom that they have with this form of payment, and 14% of the
sample cite low quality of banking services and the higher costs of payments through banks.
However it is clear that the high percentage of cash payments facilitates non -reporting of
economic activities. This is supported by the observation that the companies (66%) work without
invoices. The costs saved by following this practice reach approximately 24% of turnover.
One of the key issues cited by respondents for insufficient business growth in Bulgaria is
license and permission procedure. Business attitudes towards licensing and permission
requirements are extremely negative. However, only 1.5% of the sample answer that they
operate without legal licenses, showing that the risk of sanctions is high. The average cost
(state fees plus consultants’ and lawyers’ pay) of obtaining a license is estimated at 14.5% of
companies’ monthly turnovers.
In summary, the survey shows a marked preference of firms for irregular practices, which
facilitate tax evasion, while the high proportion using licenses indicates that, though many
companies operate in the officially reported economy at least to some extent, it is obvious that
they avoid reporting some proportion of their activities. The following two sections describe
preliminary estimates of the extent to which this actually occurs.
Evidence from tax and social insurance payments
The taxes that are most frequently evaded are the value added tax (VAT) and social security
payments. Payroll taxes and the personal income taxes are also near at the top of the list of
the most frequently evaded taxes. The survey show that almost 17% of corporate tax is
evaded by purchasing fictitious invoices. Total tax evasion can be estimated at 33% of GDP
according to sample-based calculations.
Evidence from labor contracting and wages
Between 13% and 15% of those sampled reported hiring people without any contract during
the 1997-99 period, as well as during their first accounting year. This practice allows firms to
avoid the cost of pension and health care taxes on their officially contracted employees. There
is a tendency for growth in the total number of the employed without any contract over the
1997-99 period. In 1999 the total number of employed fell by over 14%, while at the same
time the number of employed without any contract increased by 22%. This demonstrates a
clear tendency for substitution in employment to avoid contracts. The results show that
approximately 3% of employed people (ca. 80 000) are not legally registered. Therefore the
actual unemployment rate for 1999 seems to be lower than the National Statistical Institute
figures indicate.
The agriculture and the trade sectors show the largest share of workers without a contract,
while the service sector shows the smallest (0.5%). The survey shows that the salaries
actually paid were higher than reported wages (average 230 BGN) by 10% to 50%. Most
firms hide around 34 -35% of their labor costs. The high level of avoidance is a testament to
the very high total tax burden in reported wages. Both employers and workers have an
interest in avoiding these payments.
vi
Conclusions
This study has shown that though the size of the shadow economy has declined from its peaks
in the mid 1990’s, it remains a sizable portion of the Bulgarian economy. While in many
ways shadow activities have the potential to be dynamic growth sectors, bringing them into
official economy would help spread the burden of social programs more broadly. However, it
is clear that the current level of taxation and administrative costs is regarded as so high that an
attempt to impose these taxes on all would result in the elimination of many shadow activities
rather than bringing them into the official economy. Accordingly, one clear lesson is that
improved tax collection must be coupled with reduced taxes and deregulation. The size of the
shadow economy is also of interest to policy-makers seeking to promote growth. Our results
show that a substantial portion of the response to policy initiatives is effectively hidden from
the official view. Thus, an ability to correctly estimate the size and structure of the shadow
economy will not only provide more accurate statistics but can help improve growth policies as
well.
vii
I. INTRODUCTION - DEFINITION OF SHADOW ECONOMY (SE)
AND SIGNIFICANCE OF THE STUDY
ESTIMATION OF THE SHADOW ECONOMY IN BULGARIA
It has been reported in the popular press that statisticians assume in their analyses
approximately 9 trillion USD of world-wide output is not reported,2 largely due to the
existence of the shadow economy3. Friedrich Schneider4 concludes that shadow activity is
nearly 15% of the officially reported GDP. His assumptions are the result of research carried
out in 76 developed and emerging economies. This shows that the informal sector can be as
important as the official economy, especially when it accompanies the economic development
of the transition countries. According to an estimate made by Johnson, Kaufman and ZodiaLobaton5 the shadow economy in the transition countries varies between 7 - 43% for the
period 1989-1993.
There are many causes for the existence of the shadow economy, but some of the most
important can be readily identified. These are high tax burdens, weak banking systems,
business regulations and legislation, in-efficiency of government institutions and high
unemployment rates. The shadow economy tends to be greater in the developing and
transition countries due to more corruption and low incomes.
To date there is no precise definition of the shadow economy. Friedrich Schneider and
Dominic Enste define it as a multitude of activities that are not reported by the official
statistics. According to Feige6 the development of the shadow economy is due to regulations
and rules imposed on business by the state. De Soto7 has also contributed to the explanation
of the shadow economy phenomenon. He holds that the quality of regulations as well as their
enforcement are of great importance for the development of the shadow economy and
emphasizes the change in the attitude of the economic agents towards the institutions and the
legislation, especially in the transition countries.
For the last three years the development of the Bulgarian economy has been
characterized by macroeconomic stability and financial soundness. At the same time the state
has extended control over business activities by increasing the number of the legal regulations
concerning licensing, permissions and registration. Current licensing and registration
2
The Shadow Economy, The Economist, August 28, 1999, Vol. 352.
IMF estimated global GDP in 1998 was 39 trillion USD, almost 9 trillion USD (9*1012) of activity equivalent
approximately to the American output went undetected, ibid.
4
Schneider, Friederich and Dominic Enste, Shadow Economies Around the World – Size, Causes, and
Consequences, Jena, 1999.
5
Studies of Illegal and Unreported Activity (Michigan: WE Institute for Employment Research, 1996) and
Journal of International Affairs, Vol. 53.
6
See: Philip Smith, Assessing the Size of the Underground Economy. The Canadian Statistical Perspectives,
Canadian Economic Observer, Catalogue number 11- 010, 18 March 1994.
7
See: de Soto, Herhando The Other Path, NY: Harper & Row, 1989. De Soto's formulation is based on the
cases formulated in the context of Peru's economic development (that Peru is governed by a set of laws that are
relatively less efficient that those that guide the informal sector).
3
procedures impede business activity and create favorable conditions for corruption in state and
local administration. Surveys of the Institute for Market Economy show that business
regulation and the constantly changing number of regulations are important reasons for firms
to prefer the informal sector of the economy8.
In theory and practice the most common methods for measuring the shadow economy
are the following:
The direct approach9 is based on a direct inquiry with the firm managers, state and
local administration representatives by means of interviews and questionnaires. An advantage
of this method is the variety of the information collected about the structure of the shadow
economy, and the incentives that lead to it. The outcome of such research depends on the way
the questionnaire is formulated and the willingness of the businessmen to give truthful
answers. The disadvantage of the approach is the degree of reliability of the information
given the illegal nature of many shadow activities. This presents difficulties for estimation of
the actual size of the shadow economy. For this reason, the direct approach is not much used
in practice.
Commonly used indirect methods for estimation are based on differences between
national expenditures and revenues, an assessment of the labor market analyzing the
differences among the officially registered employment, the unemployment rate and the
number of people who are actually employed within the economy.10
Another indirect method is the currency demand approach11 used in the estimation of
the informal economy in OECD countries by Schneider, Johnson, and Kaufman, 1998. In the
past few years Kaufman and Kaliberda have applied an assessment of the shadow economy
through energy consumption costs.12 This method is appropriate for comparative analyses.
The physical (electricity) approach has been applied by Johnson and Lacko to the transition
countries for the period 1989-1995. According to this method of estimation, the size of the
shadow economy in GDP for Bulgaria was 26.1% (1989-1990), 32.7% (1990-1993) and 35%
(1994-1995). Bulgaria is one of the countries in transition where the size of the shadow
economy shows an upward trend.
The objective of the analysis reported in this paper is to estimate the proportion of the
informal sector in the economy, its structure, the incentives for its development and the effect
on the economic growth and the competitiveness of the economy. In this research the shadow
economy is measured both by the energy consumption and the direct approach by sectors.
Firms from the following sectors of the economy are included: wholesale and retail trade,
transport, construction, mechanical and engineering, food and drinks production, chemical
industry, textile and knitwear production, tourism and agriculture.
8
See: In Search for Growth: Policies and Lessons from Bulgarian Transition, IME Newsletter, Vol. 5, No 11-12,
1999.
9
The direct approach is used by Isachsen Krovland and Storm (1982) for the estimation of the SE in Norway and
Denmark.
10
See Friedrich Schneider and Dominic Este, "Shadow Economies Around the World- Size, Causes, and
Consequences, Max-Planck-Institute for Research into Economic Systems, 1999.
11
The currency demand approach has been used by Cagan (1958) and further developed by Vito Tanzi (19801983).
12
See: Kaufman and Kaliberda, The Underground Economy in Poland.
2
II. GENERAL METHODS FOR EVALUATING SIZE AND EFFECTS
OF THE SHADOW ECONOMY
The same reasons that justify our interest in the GDP and the whole economy of a
country make necessary the estimation of a country’s shadow economy. The latter term, while
intuitively clear, has been difficult to define. Philip Smith13 has referred to the unofficial
economy as "market-based production of goods and services, whether legal or illegal that
escapes detection in the official estimates of GDP." However, this description is almost as
broad as the term shadow economy. An alternative possibility is to define the concept in
terms of its causes or indicators - a more detailed classification has therefore been given by
Schneider and Enste14 who divide underground economic activities in several categories by
three factors: monetary and non-monetary transactions; illegal and legal activities. Legal
activities are further divided into those associated with tax evasion or tax avoidance.15. A
good way to generalize these definitions would be to think of shadow economic activities as
those, which provide a way to avoid taxes.
SIGNIFICANCE OF THE STUDY
On its way to joining the EU Bulgaria has to show strong and sustainable GDP growth
to catch up with the lowest income economies in the Union. If research on shadow activities
proves a high relative share of shadow economy to official GDP, this finding can reveal a
serious resource for economic growth and the number of years needed for convergence to the
EU average would be drastically reduced. Government measures can be targeted toward
improving the business environment and removing administrative barriers to small and
medium enterprises (SMEs). Reducing the tax burden, which is usually associated with the
main motive for being in the shadow economy, can decrease the costs for firms to stay in the
shadow sector. This will increase the tax base significantly and improve the tax collection
ratio. Though these are benefits and costs to informal operations, on the macro level the total
effect of the SE in Bulgaria should be negative. This is because companies cannot use various
public and private services. The most obvious example is the tendency to avoid the bank
services. This results in reduced volume of sales, especially for exports. While the shadow
sector may be quite flexible on local markets, it is less competitive on external markets. The
negative impact of the SE on national competitiveness is the main disadvantage for a small
economy like Bulgaria, which should have an export-oriented policy. Monitoring key
economic statistics, we can point out the following indicators that suggest a high level of the
shadow economy in Bulgaria and justify our survey on the subject:
1. The transition from planned to market economy. Following the collapse of the
planned economy in the late 80’s, the newly established private sector is concentrated in the
service sector in the form of SME’s. These are characterized by frequent changes of main
activities, non-bank sources of financing, short business history and institutional gaps. This
makes the work of the NSI (National Statistical Institute) extremely complicated as it
practically impossible to cover the fast moving private sector, even if it is official and legally
13
Philip Smith (1994): "Assessing the size of the Underground Economy: the Canadian Statistical Perspective",
Canadian Economic Observer, Catalogue No.: 11-010, 3.16-33, at 3.18.
14
Friedrich Schneider and Dominic Enste (1999): "Shadow Economies Around the World—Size, Causes, and
Consequences", Lectiones Jenenses, Heft 20, 9.
15
Schneider and Enste, 9.
3
operating. In addition, with the transition to a market economy, the NSI has had to introduce
a new system of national accounts and operate in a totally different environment from that
which existed under the old regime.
2. In kind production. This is very important in rural areas, but significant in kind
production is found also in the big cities, especially in the service sector. It is a traditional
phenomenon for Bulgaria and quite typical for planned economies when the quality and
variety of goods and services was poor. In depth estimates of the share of in kind production
are presented in the annual report of the AEAF16 for 1997 and estimate that about 10-20% of
household incomes come from in-kind production (20-40% for rural areas and 5-10% for
large towns). These activities are difficult to calculate in the official GDP figures, and serious
underestimation is likely to exist.
3. Different outcomes from expenditure and income approaches for GDP. One of
the indicators for the existence of a shadow economy is the different result that comes from
the expenditure and income approaches for measuring the GDP. While in accounting terms
these methods should produce equal results, in the case of Bulgaria the approach that sums
sectional value added gives growth rates about 2 percentage points lower than the expenditure
approach for 1998 and 1999. In other words, the expenditure approach shows 5.4% and 4.4%
GDP growth for 1998 and 1999 respectively, while official numbers are 3.5% and 2.4%. This
is probably the consequence of the firms’ tax avoidance tendencies to underreport incomes.
4. Aging population and high social insurance burden. The dependency ratio for
Bulgaria (the ratio of the pensioners to employees) is very high  the highest of all applicant
and member countries of the EU. In combination with the pay-as-you-go pension system the
incentives to avoid the social insurance burden are enormous. Typical practice for SMEs is to
pay taxes on the legally required minimum wage and deliver the rest of the remuneration
unofficially. Though the share of the private sector in gross value added is 65.3% for 1999
and 63.3% in the total number of employees, the revenues to the National Social Insurance
Institute amount to much less than 50% (official numbers not published).
5. Currency substitution of the economy. During the financial crises in the early
1990’s and 1996-1997 a large share of national savings was transformed into foreign currency.
Some of the transactions were made in foreign currencies and the process seems to have a
long memory. With the introduction of the currency board just a small part of the
de-dollarization took place in the short run. Around 50% of total bank deposits are held in
foreign currency, chiefly United States dollars (USD). The prices of real estate are still quoted
in USD. USD are still offered in the exchange offices as their rates (adjusted for a variety of
commissions and fees) for buying and selling USD is have been continuously below the
official rate of the central bank.
6. Geographical situation and cross-border trade. Bulgaria is situated on the
junction of many international roads associated with legal or illegal traffic of goods and
people. Many unregistered transactions and incomes are not adequately covered in the official
statistics, which can be also a factor for the immense supply of foreign currencies.
16
See "Bulgaria 1997 Economic Survey", Agency for Economic Analysis and Forecasting, Sofia, 1998, part 3:
"The crowding out effect of the economy in kind over the market economy", p. 28.
4
Various methods have been created to estimate the size of the shadow economy of a
country.17 Among the most widely used have been the currency demand approach, the
physical input (electricity) approach and different survey (direct) approaches. In Bulgaria only
the indirect approach based on currency demand has been used so far. Under the assumption
that underground transactions are made with cash payments, an increase in the demand for
currency would be translated into an increase in the shadow economy. A currency demand
function is built with all possible factors affecting demand, and estimated econometrically
over time. Then an estimate of the SE is given by the difference in the money demand when
the tax burden and government regulations are at their lowest and highest levels. Note that we
have no way of assessing the unofficial activities unless we assume a constant velocity of
money over time and within the official and unofficial sectors.
In Bulgaria, Nenovsky and Hristov of the Bulgarian Central Bank conducted a study of
the transactions demand for money.18 In the concluding section of their paper they estimate
the size of the Bulgarian Shadow Economy using the following procedure:
1. Tax burden is approximated by the ratio of Taxes to Total Consumer Expenditure.
2. Money used in the Shadow Economy equals the difference between money demand
with and without taxes.
3. The velocity of money is assumed to be the same in the shadow and the official
economy.
4. After computing the portion of money in transaction used in the shadow economy
(C) as a percentage of the one used in the official economy, consumer expenditures
in the unofficial sector are the product of this quantity of money and velocity:
Equation 1
E  C *V
Consumer expenditures are 80% of GDP, so underground income is calculated as
Equation 2
Y shadow 
E
.
0 .8
The shadow economy is thus calculated on a monthly basis starting from the middle of
1997, yielding averages of 15.2% for 1997, 35.3% for 1998 and 24.1% for 1999.
There are a number of problems with the money demand approach in general - a
comprehensive list has been given in Schneider and Enste,19 a part of which is reflected in the
points below:
17
For a detailed description of each method along with its shortcomings and applications in different countries,
see Schneider and Enste, 1999.
18
Nikolai Nenovsky and Kalin Hristov (1999): "A Study of the Money in Transaction after Establishing the
Currency Board in Bulgaria," working paper, Bulgarian National Bank.
19
Schneider and Enste (1999), 48-50.
5
1) The approach does not capture unofficial activities in which money does not exchange
hands, eg., barters.
2) Tax burden is often assumed to be the main cause of shadow economy, thus if there
are other significant factors, the result would be an underestimate.
3) Changes in money demand might not necessarily translate to changes in the size of the
shadow economy: the former might be due to a slowdown in demand deposits.
4) As mentioned above, the velocity of money may not be constant, and moreover it may
not be the same in the official and unofficial sectors. While this is a major weakness
of the model, as Nenovsky and Hristov acknowledge, it is difficult to measure the
velocity of money in the official economy, and impossible to measure it in the
informal economy. However, without an assumption for this variable, the analysis
cannot proceed. In practice, the authors suggest that money changes hands more
quickly in the shadow than in the official sector, which results in an underestimate of
the shadow economy size.
5) In the particular case of Bulgaria, the instability of the Bulgarian lev (BGN) has
prompted the use of foreign currency, mainly US dollars and German Marks, yet the
portion of it to the amount of Bulgarian currency circulating in the shadow economy is
impossible to estimate. If we suppose that the value of dollars or marks is at least as
much as the one of the BGN, then the black market estimates would double, becoming
30%, 70% and 48% respectively for 1997-99.
The general objections to the money demand approach are quite important in the case
of Bulgaria. The bank crisis in 1996 led to a dramatic fall in bank assets as well as in the
money multiplier and money velocity. This practically makes the money demand approach
useless for Bulgaria, as the key assumption of constant velocity does not hold. Table 4
included in the appendix shows the velocity of money, calculated for M3 has increased from
1.54 in 1992 to 4 in 1997. It is only possible to analyze the period after the introduction of the
currency board (July 1, 1997) as done in the above survey.
Compared with GDP, the drop in total assets was followed by a drop in domestic
credits. The level of bank intermediation was reduced, and it is not possible to distinguish
between the reasons for the increased cash transactions and the dollarization of the economy.
It would definitely be mistaken to attribute it only to the shadow economy, since the normal
reaction of economic agents is to avoid the unstable bank system and to prefer cash
transactions. During the three year period of macroeconomic stabilization, confidence in the
banking system is in a process of slow recovery. It is inevitably accompanied by a decrease in
money velocity and the assumption of constant velocity can provide inaccurate estimates of
the shadow economy even for the last three years of financial stabilization. For these reasons,
we believe that the physical input approach is a much better tool for evaluation of shadow
activities in Bulgaria. It also provides a longer time series for comparison of the current
situation with the pre-transition period.
6
III. ALTERNATIVE ESTIMATION: THE PHYSICAL INPUT APPROACH
Given the problems with the Money Demand approach, we turn to an alternative
macro- estimation of the shadow economy in Bulgaria using energy consumption. Among
those who have previously used this approach are Kaufmann and Kaliberda.20. We also try to
measure the total economic activity level (TA) in Bulgaria by assuming that the
electricity/energy consumption in the country is proportional to the total activity TA. Thus,
any change in energy consumption, which is not matched by a corresponding change in the
total activity level in the country, should reflect a change in its shadow economy (SE) level.
In other words, the growth in the ratio of energy to total activity is an indicator of the growth
in the parallel markets of a country. We can then establish a formula for the level of the
shadow economy as a percentage of the GDP which we denote by SE.
By our assumption of constant proportion of energy to total activity,
Equation 3
TAn  En *
TAbase
Ebase
where subscripts n and base stand for the base year and the year in question, and E
denotes Energy (or Electricity) consumption. We also have
Equation 4
SE 
TA  Y TA

1
Y
Y
where Y is the country’s GDP, let SEbase = x be the fraction of shadow economy of
GDP in the base year. Then
Equation 5
TAbase  Ybase (1  x )
TAn  E n *
Ybase (1  x )
Y * En
 (1  x ) * base
E base
E base
So for the shadow economy in year n we have:
20 Daniel Kaufmann and Aleksander Kaliberda (1996): "Integrating the unofficial economy into the dynamics of
post socialist economies: A framework of analyses and evidence," Washington, D.C., The World Bank, Policy
research working paper, 1691.
7
Equation 6
SEn 
TAn
1
Yn
Y
Y 
SEn  (1  x) *  base / n   1
 Ebase En 
From the formula we see that, ceteris paribus, the shadow economy is determined both
by the level of the shadow economy in the base year and the ratios of GDP to energy
consumption in the base and given years. Note that here the term base year simply refers to
the starting year of our calculations, and once we know the shadow economy in a given year,
and the relevant statistics for energy consumption and GDP, we can recursively find the
shadow economy levels for the subsequent years. Unfortunately, this convenience of easy
calculations also underlies the shortcoming of the method. We can never find an absolute
value for the shadow economy size without using some exogenous estimate for this size in the
base year.
In case we are only interested in how a change in the shadow economy in the base year
alters the shadow economy in the current year (say we want to see how robust our calculations
are in case the level of shadow economy in the base year is allowed to vary within a range of
error), then the formula for changes follows from above by subtraction:
Equation 7
Y
USE  Ix * 
E

base
n
base
/
Y
E
n
n



We now see that, holding everything else constant; the change in the SE in any given
year is proportional to the change in SE in the base year. The coefficient of proportionality,
however, is not necessarily 1 -- it depends on the ratios of GDP to energy consumption in the
base as well as the current year.
As we mentioned above, E in the formulae could either refer to energy or electricity
consumption. The standard approach considers electricity consumption in the economy as a
whole. In the current paper we try to make the measures of shadow activities more precise by
looking at total energy consumption as well; also we deviate from the standard approach in
that we compute the shadow economy size separately in the different sectors. We believe that
the latter is the best way to capture the structural changes that have been taking place
extensively in all transition economies since the change of regimes.
We thus compare four modifications of the Physical Input approach: by looking at
electricity consumption separately for sectors, and as a whole to compare with the standard
approach and existing results; we then compute equivalent values by replacing electricity with
total energy consumption. The latter should reflect the substitution effect that takes place
among the energy sources when their prices change relative to the corresponding CPI and
inflation rates.
8
ELECTRICITY - TOTAL CONSUMPTION AND CONSUMPTION BY SECTORS
The advantage of the electricity approach is the accuracy of the data for electricity
consumption. The standard method considers only total electricity consumption in the
economy. However, when significant structural changes take place, economic production can
shift from energy intensive industrial sectors to services or agriculture, which are
characterized by low or zero electricity consumption. In the case of Bulgaria this process
cannot be ignored.
Assuming a constant GDP/electricity ratio with 1989 as a base year, we test the
assessment of the shadow economy with the standard approach and structural adjustments.
We use real GDP data that we compute based on the Bulgarian GDP in 1989 and indexes of
GDP growth for the years 1989-1998. Assuming a "base" level x of the shadow economy in
1989, we use the available data for GDP and Electricity consumption to express in terms of
the levels of the shadow economy in the years 1990-1998. For example, since the GDP levels
for 1989 and 1990 are respectively 39,579 and 35,977 while the Electricity consumption
levels are 38,816 and 47,528 respectively, by formula (1) the shadow economy in 1990 is
Equation 8
Y
Y 
SE1990  (1  x) *  base / 1990   1
 Ebase E1990 
 39,579 35,977 
SE1990  (1  x) * 
/
  1  (1  x) * 1.347  1
 38,816 47,528 
Similarly, we proceed to find the levels for the consequent years. Finally, we need an
estimate for the shadow economy in the base year, x. Since it is difficult to give a precise
value for that, we consider a range of possible values and show that no matter what the
starting value is, the trends in the shadow economy growth are the same.21
To adjust the estimation for the structural changes, we divide the GDP into two
sectors: industry and other. We take the ratio of value added to electricity consumption for
each sector and calculate the shadow economy by sectors. In Table 1 the data for the sectoral
approach is calculated for a uniform distribution of the shadow economy by sectors in the base
year. Figure 1 shows how these methods differ for a given base year level of the SE. Below
we test how asymmetric distribution by sectors can affect the final results.
Table 1
Electricity Approach, Share of Shadow Economics in GDP
under Alternative Base Year Estimates of SE/GDP
21
See appendix.
9
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Standard
Sectoral
Standard
Sectoral
Standard
Sectoral
20.0
61.9
56.2
56.1
61.4
57.3
70.1
31.8
30.4
16.6
20.0
24.5
16.6
23.5
30.9
28.1
75.1
49.7
48.5
36.0
25.0
68.7
62.7
62.6
68.1
63.8
77.2
37.3
35.8
21.5
25.0
29.6
21.2
28.1
35.6
32.7
81.5
55.1
53.9
40.8
30.0
75.4
69.2
69.1
74.8
70.4
84.3
42.8
41.2
26.4
30.0
34.8
25.9
32.7
40.3
37.4
88.0
60.5
59.2
45.6
When calculating the relative shares of shadow economy the final results are strongly
dependent on the assessment for the shadow economy in the base year. Such an assessment
can be quite misleading, but regardless of the initial share the dynamic trends are not affected.
The graph below shows estimates for different scenarios for the base year.
Figure 1
Ratio of Shadow Economy to GDP
Standard Electricity vs. Sectoral Approach
100.0
80.0
% 60.0
40.0
20.0
0.0
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Standard electricity
Sectoral approach
The advantages of the sectoral approach are that we have additional figures for the
shadow economy by sectors, and we can distribute the concentration of the informal activity
by sectors. However, we decided to break the GDP figures into only two sectors: industry and
others (including services and agriculture). The data for agriculture and services is reliable on
an aggregated level, but on a disaggregated level is strongly biased by the form of ownership.
In 1989 most agricultural production was concentrated in the state owned co-operative farms,
and the consumption of electricity was adequately counted for the sector. With the land
restitution, these state farms were liquidated and production was transformed to small private
farms. In most cases these are individual households and their electricity consumption is
10
counted as household consumption. This is also true for most of the micro firms operating in
the service sector. The relative shares of the shadow economy to GDP can also vary with
different scenarios for the distribution of shadow economy by sectors in the base year. In
Table 2 and Figure 2 we show calculations for different sectoral distribution of the relative
shares. However intuition suggests that in the case of Bulgaria, the share of shadow activities
in the industrial sector is much lower than it is in the service sector. The reasons are both
methodological failures that do not allow the statistical institutions to capture the total activity
in the service sector and the flexibility of the service sector to underreport revenues for
avoidance purposes. It is also important to note that in the base year 1989 the methods used
for calculating GDP did not accurately cover many of the services, as the government supplied
most of them at zero cost to the consumer.
Table 2
Sectoral Approach, SE/GVA (%), Sectoral Approach
with Different Assumptions for the Distribution of SE by Sectors in Base Year
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Total
(uniform)
Industry
Others
Total
Industry
Others
Total
25.0
29.6
21.2
28.1
35.6
32.7
81.5
55.1
53.9
40.8
10.0
15.8
13.5
6.2
6.0
2.9
19.8
27.0
23.7
15.1
46.9
49.2
35.6
73.5
106.2
102.5
236.0
134.5
134.1
107.6
25.0
30.2
22.7
30.5
39.3
36.0
91.6
59.8
59.6
45.2
30.0
36.8
34.2
25.5
25.3
21.6
41.6
50.1
46.2
36.1
17.7
19.6
8.7
39.0
65.3
62.4
169.3
88.0
87.7
66.4
25.0
29.5
20.8
27.3
34.4
31.7
78.2
53.6
52.0
39.3
11
Figure 2
Ratio of Shadow Economy to GDP
Sectoral Electricity Approach
140%
120%
100%
80%
60%
40%
20%
0%
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
ENERGY APPROACH
The electricity approach is based on the empirical assumption that the ratio of GDP to
electricity consumption is constant. In the long-run changes in relative prices of energy
resources or other supply side effects can force the consumers to substitute different sources
of energy. (See Figure 3) The substitution effect can seriously affect the assessment of the
shadow economy.
Figure 3
Changes in the Energy Consumption
(Base Year = 1989)
140
120
100
80
60
40
20
0
1989
1990
1991
1992
1993
1994
electricity
12
1995
energy
1996
1997
1998
To calculate a better estimate of the shadow economy, we compare the consumption of
total energy sources, including electricity, coals, fuel, gas, heating, calculated terajoules and
changes in electricity consumption. The decrease in energy consumption is even greater than
the drop in electricity. This means that some of the consumers have shifted from other
sources of energy to electricity leading to changes in the ratio of GDP to electricity and
overestimation of the shadow activity. Accordingly, to improve the quality of the physical
input approach, we substituted total energy consumption for electricity in the calculations
outlined above. This method is a reliable tool to remove any substitution effects that might
distort the estimates. A slight concern in using the total energy approach is the probability of
statistical errors in the energy consumption data and the chance that hidden consumption may
exist for some sources such as kerosene, diesel or petrol.
Figure 4
Changes in Real GDP and Total Economic Activity
and Changes in the Share of SE
120
100
80
60
40
20
0
1989
1990
1991
1992
1993
GDP
1994
TA
1995
1996
1997
1998
SE/GDP
In order to calculate the actual sizes of the Bulgarian shadow economy in the years
1990-1998, we need to know the level for the base year 1989 (see formula (1) above). There
exist estimates in the economic literature, yet given the inaccurate and contradictory results
those yield, we allow for a range of possible values of the shadow economy level in 1989.
Thus, as in the Electricity approach above, we can get a sense of what ranges the shadow
economy levels move in and see that the trends in these levels are unaffected.
13
Figure 5
Shadow Economy Levels for the Years 1989-1990 in Bulgaria,
Based on Five Different Values for the Shadow Economy in 1989
(10, 20, 30, 40 and 50% respectively)
Ratio of Shadow Economy to GDP
Standard Energy Approach
0.6
Series 1
0.4
Series 2
0.2
Series 3
0
-0.2 1
2
3
4
5
6
7
8
9
Series 4
Series 5
-0.4
Figure 5 shows that regardless of the initial level of the shadow economy in 1989, the
pattern of change across the years of transition is, of course, identical. However, it is also
clear that the lowest three estimates for initial values generate negative results in the middle
years. This gives some basis for accepting higher initial estimates for SE/GDP.
However, before accepting these results we must consider that just as with the
standard electricity approach, the above results suffer from omission of the sectoral changes in
the economy. If we believe that the black markets flourish most in the service (and
agricultural) sector and least in industry, and consider the fact that the participation of industry
in the Bulgarian GDP has almost halved since 1989 while the agricultural sector has stayed
almost the same, we can see how such distortion can significantly modify the overall results.
We refine the method of looking at total energy consumption by considering energy
consumption in the different sectors. For correspondence to the electricity results, we break
the economy into the same two sectors, industry and everything else (service and agriculture).
The difference in results between the standard and sectoral approach is shown in Figure 6 for
a shadow economy of 25% in the base year.
14
Figure 6
Estimation of Shadow Economy to GDP Ratio
by Energy Consumption
40,0
20,0
%
0,0
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
-20,0
standard
sectoral
Table 3
Energy Approach, SE/GDP (%)
Standard
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
20.0
21.9
-2.9
-10.0
-7.5
-9.0
-6.7
4.3
-1.5
-8.0
Sectoral
Standard
Sectoral
Standard
Sectoral
20.0
22.0
2.8
3.7
11.8
8.6
13.3
24.8
19.5
12.8
25.0
26.9
1.1
-6.2
-3.7
-5.2
-2.8
8.6
2.6
-4.1
25.0
27.1
6.8
8.0
16.5
13.0
17.5
29.6
23.8
17.4
30.0
32.0
5.2
-2.5
0.2
-1.4
1.1
13.0
6.7
-0.3
30.0
32.2
10.8
12.4
21.3
17.3
21.6
34.4
28.2
21.9
With the above four methods we capture the structural change and substitution effects
that Eilat and Zinnes22 proposed as a way to reflect changes in efficiency (in use of electricity)
by considering the changing fraction of the private sector over time. We already include a
partial estimate of efficiency in the sectoral division of the shadow economy. Private
ownership has the largest share in the service sector in the economy, so to an extent its growth
is accounted for in the growth of the service sector in the Bulgarian GDP, and the latter we
capture in our new sectoral approach.
22 Eilat, Yair and Clifford Zinnes, The Evolution of the Shadow Economy in Transition Countries:
Consequences for Economic Growth and Donor Assistance, CAER Discussion Paper No. 83, September 2000.
15
CONCLUSION
The physical input approach provides easy estimates of SE trends, but it is not a
precise tool. The energy consumption can be biased by different factors that affect final
results. As we see in the case of Bulgaria, different assumptions and different energy
resources yield drastic changes in final results. The energy approach can be more reliable for
economies with stable growth paths, which do not witness dramatic structural shifts.
However, for transition economies like Bulgaria, the energy consumption approach as well as
the currency demand approach should be used only as an indicative measure of SE dynamics.
We believe that some structural factors of the traditional electricity approach are eliminated
with our sectoral approach when we use total energy consumption instead of electricity
consumption.
We believe the most reliable method that can be used as an indicator for the shadow
economy is the total energy approach by sectors. Calculations from this method show that the
shadow economy in 1998 has declined below the level of the 1989 base-year level. If we take
what we believe as a realistic assumption of 30% level of SE in 1989, the SE economy in
1998 as a share of official GDP should be 22%. The absolute peaks of the shadow activity are
calculated for 1990 (32.2%) and 1996 (34.4%). Since 1996 we observe a declining trend in
the relative share of the shadow economy. This is not surprising given the fact that the
introduction of the currency board in mid 1997 marked a steady rise of tax revenues to GDP.
We also believe that the energy approach is basically a good tool for the assessment of the
trends of shadow activities rather than relative shares. The base year relative share must
always be given as an external value for the model and the only way to assess this value is the
micro approach.23
23 It should be noted that it is also possible to work the other way around, i.e. to assess the last year shadow
economy and work out the trends backwards from that point.
16
APPENDIX (CHAPTER 1)
Table 4
Selected Indicators of the Bank Sector in Bulgaria
(From 1991 until 1995, SSB is not included)
1991
Non-performing
credits (%)
Assets/GDP*
Velocity of M3*
Money multiplier
3.77
Credits/GDP (%)
1992
1993
1994
1995
1996
1997
1998
1999
2.23
1.53
4.75
55.50
1.80
1.58
4.92
74.10
1.13
1.76
4.55
74.00
1.30
2.25
5.31
21.20
0.32
4.00
2.77
13.40
0.35
3.55
2.76
13.40
0.35
3.50
2.86
100.38 107.10
89.58
66.93
75.63
20.89
21.60
20.0
2.28
1.54
-
* Note: The bank assets cover both local and foreign currencies, but M3 does not include foreign currencies in
cash outside the banks, as the BNB does not assess the amount of foreign currencies used for cash transactions or
cash saving.
Table 5
Electricity Consumption (in kWh)
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Total
Agriculture +
Forestry
Industry
Construction
Households
Others
38,816
47,528
42,000
38,899
39,628
39,306
43,750
30,479
28,031
25,960
1,069
994
866
686
556
563
539
600
355
233
20,773
19,149
14,925
13,173
12,353
12,793
14,075
13,236
11,355
11,049
994
896
606
421
382
304
356
251
303
265
10,183
10,475
10,405
9,685
10,021
9,806
10,956
11,486
9,882
10,540
5,797
5,116
3,226
3,856
4,722
4,672
15,185
4,906
6,136
3,873
Source: Bulgarian Statistical Abstract, various years
17
Table 6
Electricity Consumption (in kWh)
Total
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
38,816
47,528
42,000
38,899
39,628
39,306
43,750
30,479
28,031
25,960
Agriculture
1,069
994
866
686
556
563
539
600
355
233
Industry
Service/Others
21,767
20,045
15,531
13,594
12,735
13,097
14,431
13,487
11,658
11,314
15,980
15,591
13,631
13,541
14,743
14,478
26,141
16,392
16,018
14,413
Source: Bulgarian Statistical Abstract, various years
Table 7
Electricity Consumption (in kWh) - Two-Sector Division
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Total
Industry
Others
38,816
47,528
42,000
38,899
39,628
39,306
43,750
30,479
28,031
25,960
21,767
20,045
15,531
13,594
12,735
13,097
14,431
13,487
11,658
11,314
17,049
16,585
14,497
14,227
15,299
15,041
26,680
16,992
16,373
14,646
Source: Bulgarian Statistical Abstract, various years
18
Table 8
Real GDP - Total and by Sectors
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Agriculture
Industry
Services
GVA
GDP
4,394.000
4,231.422
4,413.373
3,760.194
2,624.615
2,873.954
3,290.677
3,047.167
4,049.685
4,106.381
23,507.000
20,568.630
16,249.210
15,209.260
14,266.290
15,122.270
14,305.660
12,617.600
11,191.810
11,673.060
11,742.000
11,213.610
10,451.080
7,639.743
7,685.581
7,447.328
7,745.221
7,024.916
5,669.107
5,697.453
39,643.000
36,013.657
31,113.671
26,609.201
24,576.486
25,443.549
25,341.563
22,689.679
20,910.600
21,476.889
39,579.000
35,977.310
32,955.220
30,549.490
30,091.240
30,632.890
31,521.240
28,337.590
26,353.960
27,276.350
Source: Bulgarian Statistical Abstract, various years
Note: The above figures were calculated from indexes for GDP growth from 1989 to 1998
(1989=100) and actual levels of GDP for 1989.
Table 9
Two-Sector Division of GDP
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Industry
Other
GVA
GDP
23,507.0
20,568.6
16,249.2
15,209.3
14,266.3
15,122.3
14,305.7
12,617.6
11,191.8
11,673.1
16,136.0
15,445.0
14,864.5
11,399.9
10,310.2
10,321.3
11,035.9
10,072.1
9,718.8
9,803.8
39,643.0
36,013.7
31,113.7
26,609.2
24,576.5
25,443.5
25,341.6
22,689.7
20,910.6
21,476.9
39,579.0
35,977.3
32,955.2
30,549.5
30,091.2
30,632.9
31,521.2
28,337.6
26,354.0
27,276.4
Source: Bulgarian Statistical Abstract, various years
19
Table 10
Energy Consumption in Bulgaria (in terajoules)
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Total
Agriculture+
Forestry
Industry
773,551
714,096
521,006
448,004
453,259
454,094
478,974
481,382
422,798
408,846
39,955
31,929
22,775
17,844
18,852
16,433
18,121
14,799
13,674
437,640
379,848
283,541
212,529
204,229
220,836
250,251
244,879
223,793
186,196
Construction Households
23,170
12,818
9,767
7,166
7,077
6,676
6,134
5,678
4,395
156,986
121,422
140,612
155,797
139,423
142,587
148,323
123,628
136,609
Others
114,137
71,296
62,321
68,223
67,906
63,027
63,925
54,900
61,661
Source: Bulgarian Statistical Abstract, various years
Table 11
Energy Consumption in Bulgaria (in terajouls) - Two Sector Division
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Total
Industry
Other
773,551
714,096
521,006
448,004
453,259
454,094
478,974
481,382
422,798
408,846
437,640
403,018
296,359
222,296
211,395
227,913
256,927
251,013
229,471
190,591
335,911
311,078
224,647
225,708
241,864
226,181
222,047
230,369
193,327
211,945
Source: Bulgarian Statistical Abstract, various years
20
Figure 7
Bulgarian Shadow Economy Levels Based on a Range of Possible Levels in 1989
Ratio of Shadow Economy to GDP
Standard Electricity Approach
1.2
1
x = 0.20
x = 0.25
x = 0.30
x = 0.40
0.8
0.6
0.4
0.2
0
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Source: Figures for electricity consumption and real GDP from the tables above, and calculations by formula
(1) in the text; x = level of the shadow economy in 1989.
Figure 8
Shadow Economy Levels Based on the Same Range
of Possible Levels of the Shadow Economy in 1989 as Above
Ratio of Shadow Economy to GDP
Electricity Sectoral Approach
140%
120%
100%
80%
60%
40%
20%
0%
1989
1990
1991
1992
1993
21
1994
1995
1996
1997
1998
Figure 9
Shadow Economy Levels Based on a Range of Possible
Values for the Shadow Economy X in the Base Year 1989
Ratio of Shadow Economy to GDP
Standard Energy Approach
0.5
0.4
0.3
x = 0.2
0.2
x = 0.25
x = 0.3
0.1
x = 0.4
0
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
-0.1
-0.2
Source: Figures for energy consumption and real GDP from the tables above, and calculations by formula (1)
in the text; x = level of the shadow economy in 1989.
22
Figure 10
Shadow Economy Levels, Computed by Estimating the Shadow Economy
in the Industry and Non-Industry Sector of the Economy
Ratio of Shadow Economy to GDP
Sectoral Energy Approach
0.4
0.35
0.3
0.25
x = 0.2
0.2
x = 0.25
x = 0.3
0.15
0.1
0.05
0
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Source: Figures for energy consumption and real GDP from the tables above, and calculations by formula
(1) in the text; x = level of the shadow economy in 1989.
23
THE MICRO APPROACH -- SURVEY
OF SMALL AND MEDIUM ENTERPRISES
This section describes the results of a survey carried out under the auspices of the
Agency for Economic Analysis and Forecasting, Harvard University, and the Institute for
Market Economics. Data collection took place during the summer of 2000.
FIRM PROFILE24
The sample consists of 509 Bulgarian companies, operating in ten economic sectors.
These sectors were selected based on their importance for the economy. According to data for
1999 the relative share of these sectors in gross value added is about 56%. The data collection
method was face-to-face interviews with company managers using standardized
questionnaires.25
Included in the survey are 87% private enterprises, 3.5% state enterprises, and 9.4%
co-operatives. Forty-seven percent of the firms are registered as sole proprietorships in
accordance with the Law on Commerce. Most of the enterprises were registered between
1991 and 1993. In small towns the firms emerged between 1991 and 1992, and the legal
status of most firms has not changed since that time, i.e. 82% of the firms have preserved their
status. However, 41.2% of the firms in the chemical industry, mechanical and engineering,
transport and textile production have undergone a change due to privatization of enterprises
between 1997 and 1999.
ECONOMIC AND FINANCIAL DEVELOPMENT OF THE FIRMS
As seen in Table 12, the firms’ profit varies by years (1997-1999). Fifty-two percent
of the enterprises declared a change in their profits for 1999. Only 3% of the firms have high
profits. According to the survey 30% of the companies reported losses in 1999  more than
the number of companies in 1997 (19%). Most companies with a turnover of 50-75 000 BGN
have registered losses.
Table 12
Percent of Companies Declaring Losses and Profits in 1999
Company Size
- Turnover
Less than 50 000 BGN
50 - 75 000 BGN
75 000 -100 000 BGN
100 - 200 000 BGN
200 - 500 000 BGN
More than 500 000 BGN
Losses
11.76
4.44
14.89
13.33
12.50
15.25
Moderate Without Moderate
Losses
Losses
Profit
24.84
8.89
12.77
13.33
25.00
10.17
24
22.88
24.44
14.89
10.00
12.50
10.17
39.87
55.56
55.32
56.67
50.00
57.63
Significant
Profit
Total
0.65
6.67
2.13
6.67
0.00
6.78
100.00
100.00
100.00
100.00
100.00
100.00
This research analyses the financial performance and development of the firms studied.
The questionnaire was amended after a pre-test with ten companies involving three settlements. American
Statistical Association collected the data during June-July 2000.
25
24
Total
12.02
18.03
18.03
48.91
3.01
100.00
Source: IME calculations on the basis of the survey
The data from the survey show that small and medium enterprises remain active in
spite of uncertainty about the size of their profits. We can draw the conclusion that the
financial state of enterprises is relatively stable (firms with a high turnover have reported
significant profits), but it is also possible that consistently reported negative profits are simply
an indication of reliance on unreported shadow activities for continued solvency.
The survey indicates a decrease in profit margin during 1999, as compared with 1997,
indicating that the firms are operating in a more competitive market. The decrease is also due
to higher imports as a consequence of foreign trade liberalization following introduction of the
currency board, as well as to the development of competition in the domestic market.
Table 13 shows investments in wholesale and retail trade firms are under 15 000 BGN
annually, while the value of the investments in food and drinks production and in agriculture
amounts to 60 000 BGN. Transport, mechanical and engineering enterprises invest more.
Tables 13 and 14 below show investment in the period 1997-1999. Growth of investment was
evident in firms with an annual turnover of more than 200 000 BGN.
Table 13
Assessment of Annual Change in Total Value of Investments in 1999
Company Size- Turnover
Less than 50 000 BGN
50 - 75 000 BGN
75 000 -100 000 BGN
100 - 200 000 BGN
200 - 500 000 BGN
More than 500 000
BGN
Total
Strong
Decline
Small
Decline
No
Change
Small
Strong
Growth Growth
36.43
21.62
32.50
26.09
16.67
22.22
27.13
13.51
17.50
13.04
10.00
9.26
19.38
27.03
17.50
13.04
26.67
18.52
12.40
29.73
30.00
30.43
40.00
38.89
29.07
18.53
20.13
25.24
Source: IME calculation on the survey – cross-tabulation
25
4.65
8.11
2.50
17.39
6.67
11.11
Total
100.00
100.00
100.00
100.00
100.00
100.00
7.03 100.00
Assessment of Annual Change in Total Value of Investments in 1997
Company Size- Turnover
Less than 50 000 BGN
50 - 75 000 BGN
75 000 -100 000 BGN
100 - 200 000 BGN
200 - 500 000 BGN
More than 500 000BGN
Total
Strong
Decline
Small
Decline
No
Change
Small
Growth
Strong
Growth
Total
13.04
0.00
5.88
4.00
8.57
4.26
10.87
9.38
17.65
24.00
8.57
14.89
31.88
34.38
23.53
20.00
34.29
36.17
36.96
53.13
50.00
48.00
37.14
25.53
7.25
3.13
2.94
4.00
11.43
19.15
100.00
100.00
100.00
100.00
100.00
100.00
8.36
12.86
31.19
39.23
8.36
100.00
Source: IME calculation on the survey – cross-tabulation
The dynamics of firm investments can be characterized as stable, with a definite
tendency to increase for the large-scale firms. In 1999 about 6% of the small firms with a
turnover below 50 000 BGN declared a decrease in investments. At the same time 50% of the
larger companies reported investment growth in 1999. According to the managers, sources of
investment funds are profits (66.0%), personal savings (53%), and bank credits (24.2%);
42.9% of the managers claim that investment is insufficient, 40.3% - medium, only 16.8%
consider it high.
The value of sales has grown in comparison with the first accounting year. All firms
generated greater profit than experienced at the start of business, i.e. the first accounting year.
For construction, chemical, mechanical and engineering firms, the value of sales ranges
between 100 000 and 200 000 BGN. In other sectors the value of sales is generally between
75 000 and 100 000 BGN. Wholesale and retail trade firms register low sales - between 50
000 - 75 000 BGN.
There is a deviation between the firms’data for sales trends in the survey and the data
of the National Statistic Institute (NSI) for sales in the overall economy during 1998-1999.
The reason could be that the survey estimate is based on an unweighted average of the
respondents’ answers. At the same time the distribution of the companies in the survey
according to size shows that in 1999 the large firms (with a turnover of more than 200 000
BGN), which have a very small share in the total number of enterprises in the country,26
contribute most to the growth of sales.
26
According to the survey of the Agency for Economic Analysis and Forecasting of the small and medium
enterprises, in 1999 the share of the small -medium enterprises was 98.9%.
26
Table 14
Annual Change in Total Value of Sales in 1997
Company Size- Turnover
Less than 50 000 BGN
50 - 75 000 BGN
75 000 -100 000 BGN
100 - 200 000 BGN
200 - 500 000 BGN
More than 500 000BGN
Total
Strong
Decline
Small
Decline
No
Change
Small
Growth
8.33
0.00
0.00
0.00
0.00
8.33
25.00
0.00
5.88
0.00
8.33
8.33
23.61
41.67
41.18
40.00
16.67
8.33
36.11
41.67
35.29
50.00
75.00
66.67
5.19
15.56
26.67
43.70
Strong
Growth
6.94
16.67
17.65
10.00
0.00
8.33
Total
100.00
100.00
100.00
100.00
100.00
100.00
8.89 100.00
Source: IME calculation on the survey – cross-tabulation
Table 15
Annual Change Total Value of Sales in 1999
Company Size- Turnover
Less 50 000 BGN
50 -75 000 BGN
75 000 -100 000 BGN
100 - 200 000 BGN
200 -500 000 BGN
More 500 000BGN
Total
Strong
Decline
Small
Decline
No
Change
Small
Growth
46.48
28.57
29.41
25.00
0.00
15.79
21.13
21.43
35.29
16.67
14.29
15.79
18.31
7.14
17.65
25.00
28.57
5.26
12.68
42.86
11.76
16.67
57.14
47.37
34.29
21.43
16.43
22.86
Strong
Growth
1.41
0.00
5.88
16.67
0.00
15.79
Total
100.00
100.00
100.00
100.00
100.00
100.00
5.00 100.00
Source: IME calculation on the survey – cross-tabulation
The following conclusion can be made: the growth of investment and sales in 1999 is
due to high investment of companies having a turnover of more than 200 000 BGN. In 1997
the investment trend was almost equal for all companies, regardless of their size. It is
interesting to point out that for most firms, regardless of the economic sector, starting capital
is typically acquired from personal and family savings (65% on average). Only 18% of the
enterprises have used bank credit (17% in the chemical industry, where the share of the
personal savings is low).
27
Figure11
The Major Sources of Starting Capital
(%)
total
agriculture
tourism
chemical
mechanical and engineering
food and drinks production
textile and knitware production
construction
transport
wholesale
retail
0
10
20
30
40
50
60
70
80
90
%
other
borrowing from the unofficial financial sources
financing from international program s forbusiness prom otion
firm credit
bank credit
loans from friends
loans from fam ily
personal saving
To provide capital in subsequent years, firms use personal savings (58.1%), profits
(51.5%) and bank loans (29.1%). The managers point out that serious problems for obtaining
credit are: collateral requirements, complicated procedures when applying, need to present
bank guarantees, interest rates, and need for the firm’s historical accounting record.
28
Figure 12
The Most Significant Obstacles for Getting Bank Credit
1 - not a problem at all; 5 - very serious
problems
total
agriculture
tourism
chemical
mechanical and engineering
food and drinks production
textile and knitware production
construction
transport
wholesale
retail
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
mean
credit interest rate
collateral requirements
necessity of firm history
relations with bank managers
complicated procedure of applying for bank credit
nk guarantee requirements
long term from application for getting credit and decisions
Of the managers interviewed, 47.3% point out that there is no problem obtaining
information and co-operation from the bank employees for “facilitating the procedure” to
receive credit. This means that bank employees are willing to grant favors (accept bribes),
thus breaking the rules of bank servicing -- equal credit access for firms and adherence to
business ethics. This percentage is quite high, revealing a disposition of bank employees
toward receiving additional payments. Such activities are indications of corruption, and they
help to promote the shadow economy.
Business development is determined by various factors - household incomes and other
demand factors, possibilities for financing, interventions from the state and mechanisms for
business regulation (licenses, permissions and other similar measures aiming at regulating
29
businesses and restraining economic freedom). Figure 13 lists the barriers impeding
businesses, as ranked by managers, according to firm turnovers.
The firms’ managers point out tax burden, social insurance and insufficient financing
as major obstacles to development and expansion of businesses.
Figure 13
What Are the Obstacles to Business Expansion?
1 - not a problem at all; 5 - very serious problem
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
4.19
4.22
4.16
4.21
High taxes/social security
3.88
3.64
3.61
3.78
Insufficient financing
3.31
3.15
2.97
3.22
Legislative basis
2.86
Insufficient demand
1.62
1.76
1.75
1.66
Lack of qualified employees
1.42
1.46
1.44
1.4
Lack of efficient management
3.29
3.15
3.17
Up to 75 000 lv
Between 75 000 and 200 000 lv
Over 200 000 lv
Total
The above graph reveals the following about business expansions by company size:

The tax burden is evaluated as the most significant barrier for all types of
companies.

Insufficient financing is a higher barrier for the small than for the big
companies, though all firms, especially those with a small turnover, indicate
insufficient financing as a problem.

Legislative basis restricts development of smaller more than bigger companies.

The lack of qualified employees and management does not significantly restrict
either small or big companies.

Construction, wholesale and retail trade companies indicate insufficient
demand as a serious problem.
30
Figure 14
Obstacles to Business Expansion in Sectors
1 - not a problem at all; 5 - very serious problems
insufficient dem and
lack of effecient m anagem ent
legislative basis
tourism
mechanical
and
engineering
textile and
knitware
production
transport
retail
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
lack of qualified em ployess
high tax es/ social security
insufficient financing
The current tax and social insurance system does not contribute to the development of
legal business. Therefore it is necessary to improve the respective regulations so more
companies can operate in the formal sector.
FACTORS FOR THE DEVELOPMENT OF THE SHADOW ECONOMY
Paying in cash is an indicator of informal economy since authorities cannot tax money
flows among firms. In this way transactions are not registered officially, and information
concerning money circulation is distorted. The survey data shows that 54% of the respondents
make and receive payments in cash and 32% make and receive payments through the banking
system.
Most firms prefer cash transactions. This can be seen in wholesale and retail trade,
transport, tourism, agriculture, textile and knitwear production sectors. Analyzing the
responses we can conclude that 69.4% of the enterprises prefer paying in cash, and 19.2% cite
the greater freedom they have with this method of payment. According to 14%, the cause lies
in the low quality bank services and higher costs of payments through banks. It indicates the
managers’ and owners’ propensity to avoid transparency of transactions and to operate in the
shadow economy.
31
32
Table 16
Mean % Saved Cost if Delivery is Without Invoice
by % of Deliveries Without Invoiced
17. What is the share of delivery from
unofficial market
without invoice?
18. What is the share of the cost
you could save if the delivery
is without invoice?
In mean %
0%
Up to 33%
34% - 66%
Over 66%
8
18
25
23
Total
13
The greatest cost saving results from trading the large percentage of deliveries without
invoice and, in 66% of the cases, reaches a 23% savings for companies that are supplied
without invoices. Figure 15 shows that enterprises construction, agriculture, tourism,
wholesale and retail trade are quite inclined to under report incomes by not issuing invoices.
Figure 15
Shares of the Cost Saved by the Firms
When Delivery Is Without Invoice (%)
20.00%
18.00%
16.00%
14.00%
12.00%
10.00%
8.00%
6.00%
4.00%
2.00%
0.00%
Retail
Wholesale Transport Construction Textile and Food and Mechanical Chemical
knitware
drinks
and
production production engineering
Tourism Agriculture
Foreign trade protection in Bulgaria is moderate compared to other countries. For
1999 the weighted average of customs duties on industrial and agricultural products is 11.4%.
The liberalization of the foreign trade relations with the European countries, countries from
CEFTA and the bilateral contracts for preferential customs duties signed with most of the
countries should generate willingness by firms pay the resulting import duties.
33
According to the survey, the reasons to avoid tariffs are over-regulation and red tape.
Along with the newly adopted Customs Act and the regulations on the enforcement of that act,
an automatic licensing procedure was introduced in 1998 for many firms engaged in foreign
trade. The procedure for releasing goods from customs, however, is still complicated and
document processing is quite slow. The level of automatization and computerization is low
for the customs authorities as a whole.27 Administrative requirements and bureaucratic
customs procedures, combined with numerous documents that must be attached to the export
and import transactions, create motivation for firms to operate in the shadow economy, thus
saving time and money. A good indicator of shadow activities is the existence of two types of
bookkeeping, one for official authorities and for tax purposes and one for real incomes and
expenses. Only 9% of firms admit to using the two types of bookkeeping in order to avoid
payment of real taxes. The data collected from this question, however, is not sufficient for
estimation of the SE, and there is always a natural tendency to avoid truthful answers to direct
questions like this. Nevertheless, it is interesting that as many as 9% of respondents were
willing to admit to keeping two sets of books. This figure can be considered to be a lower
bound on the true proportion keeping such books.
LICENSING AND PERMISSION REQUIREMENTS
Licenses and permissions (permits, certificates) are needed for continuous execution of
a concrete activity, e.g. tour operator license or permission from the Sanitary Inspection Office
for a café. In certain cases permits or certificates are also needed for separate transactions,
e.g. importation of a commodity under the import-permission regime. For some of the
activities, registration is also needed, which is different from general company registration,
e.g. registration with the central bank for a transfer of money abroad. All three - licenses,
permissions and registrations - are considered here without any specific distinction of the
businesses’ attitude toward them.
The licensing and permission requirements usually constitute a substantial part of the
cost of operation "out of the shadow." These costs include direct expenses made for obtaining
and maintaining the license or permission, as well as the indirect costs in terms of man-days
for the firm spent dealing with the issue.
27
For more details on the export and import regulations in the Customs Act and the regulations for enforcing the
act, see the FIAS report: <www.bfia.org>.
34
Table 17
Direct and Indirect Costs (in BGN) of Compliance
and Non-Compliance with Selected Legal Requirements
NON-COMPLIANCE
COMPLIANCE
Regulation
Registration at court28
Registration at BULSTAT
Registration at the National
Social Security Institute (NSSI)
Licensing/permits (total)
Indirect
Costs
Direct
Costs
Total Indirect Direct
Costs Costs Costs
Total
Costs
74
30
80
0
154
0
30
-
100
-
130
-
30
0
0
45
30
75
149
50
199
74
130
204
Source: IME data
Table 17 notes:
1. In some answers the respondents have not taken into account state fees; this is valid for company
registration.
2. Indirect costs are estimated on the basis of the man days spent in compliance or non-compliance. The
number of man days is converted into BGN through the average value added per day per person in the
private sector in 1999. (Source of macro-data: National Statistical Institute and National Employment
Agency)
3. In the costs of non-compliance, the companies probably do not consider the sanction and the probability
of being sanctioned. Therefore the indirect costs of non-compliance should be considered undervalued.
4. The last row in Table 17 shows the totals. However, the total costs of compliance (or non-compliance)
could not be applicable to a concrete firm since such a firm’s activity is most likely not to fall under all of
the regulations above. Also, it is not correct to sum the total costs of compliance and non-compliance,
since most of the firms have answered both for the costs of compliance and the costs of non-compliance,
although facing only one of the two types of costs. The last row thus gives an idea of the relationship
between the costs of staying "in the shadow" and the costs of operation "out of the shadow," as well as
for the shares of the indirect and direct costs within the total costs.
In Table 17 the direct and indirect costs of compliance and non-compliance with some
legal requirements are given.29 In comparison with the other three general requirements of
starting a legal business operation, the licensing and permission requirements are, on average,
more expensive in terms of direct costs and time spent, than the court, statistical and
mandatory insurance registrations. At the same time, the cost of non-compliance with such
regulations is higher than the cost of compliance, which means that such requirements cannot
motivate companies to move "into the shadow." However, these regulations cannot by
themselves motivate companies to legalize their activities either. One should not forget that
most businesses in Bulgaria comply with at least some (but maybe not all) regulations.
Therefore there could be companies that comply with the licensing requirements, but still keep
value added hidden.
28
In some answers the respondents have not taken into account state fees; this is valid for company registration.
Based on the primary data from a joint survey of IME and Bannock Consulting on the administrative barriers to
business in Bulgaria.
29
35
Another interesting point is that, when complying with the licensing and permission
regulations, indirect costs are three times higher than direct costs. This is probably due to the
administration’s clumsy procedures and low-quality service.
Although there is an explicit political will to reduce the number of licensing regimes,
which is consistent with the overall policy of economic liberalization, no real signs of
improvement exist. Table 18 gives the development of the licensing legislation during the
past ten years.
Table 18
Number of Activities Requiring License or Permit
Explicitly Mentioned in Laws
Year
Newly Enforced
Licensing Requirements
Until 1989
1989
1991
1992
1993
1994
1995
1996
1997
1998
1999
2
4
3
1
5
6
21
13
10
21
20
Total Licensing
Requirements
2
6
9
10
15
21
42
55
65
86
106
Source: IME
Based on data obtained in the survey, 67.8% of companies claim that their activity
requires some kind of license and/or permission. The percentage is highest in the tourism
sector (100%), where every tour operator and tourist agency must have a license and every
hotel and restaurant must hold a category certificate. The figure is slightly lower in the
transportation sector (93.3%), followed by retail trade (87.4%) and foodstuff production
(83%).
Businesses’ attitudes toward licensing and permission regimes are extremely negative.
Almost 28% of the companies in the survey think that the licensing regimes should be
relaxed, and 33% indicate that permission regimes should be removed. Only attitudes towards
taxes and social security payments are more negative than the attitudes towards licenses and
permissions.
The attitudes toward licenses are most apparent in transport, construction, retail and
wholesale trade and tourism sectors, where companies must pay relatively high license fees in
order to legalize their activities. In tourism the license for a tour operator costs BGN 5 000,
and the license for a tourist agency – BGN 2 500. Although no renewal is needed (the license
is valid for the whole life of the company), these sums are high for the average small firm.
36
Only 1.5% of the companies in the survey answered that they operate without a license
when it is legally required. At the same time, it is likely that a large percent of the licensed
companies still remain "in the shadow," hiding some substantial portion of their activities.
Again the rate of total non-compliance is highest in the tourism sector, where as much
as 5.6% of the companies operate without a license. The rate of total compliance is lowest in
the wholesale business (88.9%), probably because of the huge number of licenses, permits and
certificates needed and large indirect costs and informal direct expenses.
The average costs (state fees plus consultants’ and lawyers’ pay) of obtaining a license
are estimated at 14.5% of a company’s monthly sales. Half of the firms included in the survey
paid 200 BGN per year for renewal and maintenance of their licenses.30 The share of the
licensing and permission costs in highest in the chemical sector (33.7%) and lowest in the
wholesale trade (7%). However, other expenses (such as traveling to the licensing agency),
indirect costs (in terms of time spent dealing with the administration), and informal expenses
(e.g. bribes) are not included in the estimation.
EFFECT OF THE TAX AND INSURANCE BURDEN ON THE SHADOW ECONOMY
Significance of the tax burden for Bulgaria
Tax evasion is one of the main incentives for the existence of the shadow economy. In
Bulgaria the larger part of the social security system is based on the pay-as-you-go principle,
which does not establish a definite relationship between the payments which are made and
obtaining the respective benefits. As social security payments are obligatory, they must be
regarded as de facto taxation. With healthcare insurance the amount of payments made is not
related in any way to services offered for these payments. With pension insurance, there is
only a minimal presumed relationship between the payments and the pension, which is
received later on. As a result of the constantly increasing age structure of the population, all
revenues are allocated to pay out pensions, the average amount of a pension giving no more
than 20% of the market capitalization of the average pension insurance. In case of death the
payments are 100% lost. Therefore pension insurance is clearly a tax which is paid by the
active population and consumed by the population at retirement age. That is why in Bulgaria
the tax burden is defined as: "All direct and indirect taxes, pension and healthcare insurance."
Comparing the personal income tax rate and the insurance rate we can conclude that the
incentives to evade insurance payments are much greater.
30
The analysis of the respondents' answers to "Can you estimate the annual cost to a company like yours for
renewal and maintenance of the license?" shows high dispersion in terms of declared costs. (answer could be yes
or no!)
37
Table 19
Income Tax Rates on the Average Monthly Salary in the Public Sector
Average for the Year, Based on Monthly Data
Year
Income tax rate (%)
Payroll tax (%)
1992
1993
1994
1995
1996
1997
1998
1999
2000
15.0
14.5
15.0
14.7
18.4
17.4
17.6
16.5
16.4
42.2
42.2
42.2
42.2
41.2
44.0
44.0
44.4
45.7
Payroll/Income Tax
(as for the budget revenues)
2.1
2.1
2.1
2.0
1.9
2.1
2.0
2.2
To fulfill one of the main survey objectives, direct and indirect questions were
employed to estimate the effect of the tax burden. Managers of small and medium-size
enterprises (SME) were asked to assess this burden from two aspects:
1.
The extent to which the tax burden give incentives to shadow economy development.
2.
A quantitative evaluation of the proportion of evaded taxes.
Survey Results and the Tax Burden on Businesses
According to the businessmen interviewed, the tax burden is the most serious obstacle
to the development of the business in Bulgaria. On a scale of 1-5, with 5 representing the
most serious problem, high taxes scored an average of 4.2, i.e. between a serious and a very
serious problem. The rest of the factors receive an average assessment of less than 4 and the
question regarding lack of qualified employees comes close to an average of only 1. This
result confirms those from the research about Bulgaria in the Global Competitiveness Report
1999, where the level of employee qualification is pointed out as one of the three factors
which, in M. Porter’s opinion, positively influence Bulgaria’s competitiveness.
38
Figure 16
What Are the Obstacles to Business Expansion?
1 - not a problem at all; 5 - very serious
M
e
a
n
5.0
V
a
l
u
e
3.0
4.0
4.2
3.8
3.2
3.2
2.0
1.7
1.4
1.0
high
taxes/social
security
insufficient
financing
legislative
basis
insufficient
demand
lack of qualified lack of efficient
management
employees
The high tax burden is also considered a major reason for tax evasion. This opinion is
fully supported by 77.8% of the interviewed, partially by 18.4% and only 3.8% do not agree
with it.
Efficiency of the tax administration
In the analysis of the incentives for tax and insurance evasion we have to take into
consideration the tax burden as well as the quality of services in tax administration.
Ineffective functioning of the tax authorities and unpredictable legislation are factors, which
can further increase the cost of reporting tax and insurance payments. These expenses are
both direct (hiring more people, influencing by giving bribes, etc.) and indirect (opportunity
costs of the time which, though both are important, is spent to create contacts). As an indirect
measure of the administrative impediments, we have used the time spent by businessmen for
contacts with state authorities, measured in days per month. The answers obtained show that
businesses lose the most time in obtaining licenses. Second in terms of waiting, is time
needed for working with tax authorities. Consequently the tax rate, and not the administrative
barriers, is a primary incentive for considering the tax burden a major problem, though both
are important. At the same time, summing the total number of days spent for tax or insurance
purposes, we obtain 7.5 days per month (which shows that there are considerable opportunity
costs). The time spent reporting for healthcare insurance is more or less equal to the time
spent for the pension insurance. If we take into consideration the fact that the burden of
healthcare insurance is 5-6 times less than that for pension insurance, we must regard the
administrative procedures related to the newly established healthcare insurance system as
extremely ineffective.
39
Figure 17
How Much Time Do You Waste in Your Work
with Public Administration
days per month
5.00
4.00
3.00
2.00
1.00
0.00
Permissions Tax and
Local Fees
Social Security Health
Insurance
Statistics
Licenses
Quantitative evaluation of the shadow economy aiming at tax evasion
A higher tax burden would imply an increase in incentives for tax avoidance and
evasion and is a primary cause for transition from the formal to the shadow economy. We
take tax avoidance to mean the formal restructuring of the business in order to avoid certain
taxes, as opposed to illegal evasion of taxes. Rather, it involves taking advantage of certain
"loopholes" which permit legal escape from objectives of tax legislation. In both cases the
evaluation of unpaid taxes does not necessarily imply that a firm is 100% in the shadow
economy.
In the formulation of the questions, an explicit distinction between the two ways of
saving taxes was made in order to assess the overall effect and avoid unnecessary tension
among those interviewed.
40
Figure 18
Do You Have an Impression About the Kind of Taxes
or Other Liabilities that are Most Frequently Avoided?
response number
250
200
221
221
190
178
174
150
100
101
50
6
0
VAT
Payroll tax
(pension,
health and
other
insurance)
Personal
Duty and
income tax other import
surcharges
Corporate
tax /tax on
profit/
Excise
Other
The taxes that are most frequently evaded are value added tax (VAT) and social
security payments (Figure 18). However the assessment of taxes that are most often evaded
and the quantitative evaluation of the proportion of the evaded taxes are different. Thus, for
example, VAT, payroll taxes and the personal income taxes are at the top of the list of the
most frequently evaded obligations, but they are pushed down to the last places in the ranking
within total evaded taxes. This is so because of the different tax regimes that allow only
partial reduction of declared incomes in order to evade personal income and payroll taxes.
This practice is confirmed in the analysis of response to the inquiry about employment (see
section below on the labor market).
Table 20
Do You Have an Impression about the Percentage of Taxes
or Other Liabilities that Are Most Frequently Avoided?
Liability
Number of
Responses
Mean (%)
Duty and other import surcharges
Corporate tax /tax on profit
Excise
Payroll tax (pension, health and other insurance)
VAT
Personal income tax
Other
110
118
75
165
163
127
5
39.4
38.5
36.8
36.0
32.4
31.4
48.0
41
Weighted average
35.5
Concerning the evasion of VAT, there is no denying the practice of simultaneously
selling goods where VAT is and is not included depending on the client and his or her desire
to receive an invoice. This practice is reflected in the question: "Do you know any firms
where two different prices are set for one and the same good or service depending on whether
an invoice is requested or not?" Approximately 45.7% of the interviewed consider it a
common practice, 28.1% think that it is practiced only in a few firms, 11% are of the opinion
that there are no such companies and 15.2% do not want to answer the question.
Relationship between evaded taxes and GDP
When there are tax evasion and tax avoidance of indirect taxes (VAT, duties, excise),
as well as cases of underreporting personal income and payroll tax earnings, the undeclared
amount is equivalent to the shadow economy size. With the purchase of documents certifying
expenses in order to evade taxes on profit, there is only a transformation of the value added
from one firm to another without influencing the value of GDP. For example, company X
could use forged documents certifying expenses which would lessen its actual profit. Thus
the tax on profit is evaded and the calculations for the GDP would be lowered by the amount
of the fictitious expenses. At the same time a company Y which incurs actual losses could
issue a fictitious invoice without incurring any tax on profit payments for itself. It would
increase its revenues through the issue of the invoice and those revenues would be recognized
as value added. The result of the above example is that no tax on profit has been paid, but this
has not led to a decrease of GDP. In order to evaluate the shadow economy from the point of
view of the tax avoidance, the assessment of tax on profit has to be corrected. According to
the average value of the answers given, 25.3% of the corporate tax is evaded by purchasing
fictitious documents certifying expenses. In order to calculate the exact effect the weighted
average of those who think that there is no such practice (91 of those interviewed) must be
deducted from the above result. In this way we obtain the weighted average size of the tax on
profit evaded by purchasing fictitious invoices - 16.4%. The relative size of the evaded
corporate tax (Table 20) is 38.5% and therefore the effect on GDP is an underestimate of the
value added of 22.1% of tax due on profit. If we make the necessary calculations to obtain the
overall average size of the evaded taxes for the various types of taxes, we can see that shadow
activity for tax avoidance or evasion is 33% of GDP. This result is probably an overestimate
as it is possible to avoid excise taxes by declaring certain sales as non-excise goods. In this
case the effect on GDP is the saved excise tax from which we deduct the increase in gross
consumption due to the lowered price that results. The latter depends on the elasticity of
demand of the respective good and is impossible to evaluate based on currently available
information. On the other hand, the total value of the shadow economy must not be reduced
to that caused by tax evasion. It also includes omissions of the statistical institutions when
they give an account of the total value added and of some illegal activities such as arms, drug
and people traffic, unlicensed software, unregulated use of copyrights, etc.
42
Analysis of the tax burden by sectors
The survey covered ten sectors of the economy (selected according to the proportion of
the sector in the gross value added). For each sector separate indicators were calculated to
reveal the possible existence of divergent conclusions among sectors. In Figure19 the
integrated index of the tax burden quantitative evaluation is calculated by sectors. The index
varies between 0 and 1 and is a weighted average value of the three possible answers: "totally
agree" takes a value of 1, "agree to a certain extent" takes 0.5 and "do not agree" takes 0. The
weights that are used to obtain the mean value are the number of the answers that have been
given for each sector separately. The graph does not show major deviations from the mean
value and the coefficient of variation is low (0.0014). Only the chemical industry has a
deviation of over 5 per cent from the mean value.
Figure 19
Sectoral Distribution of the Tax Burden
as a Barrier to Business Expansion
0- no barrier to 1- most serious barrier
Mean = 0.89
Chemical
Construction
Textile and knitwear production
Retail
Agriculture
Transport
Food and drinks production
Wholesale
Mechanical and engineering
Tourism
0.7
0.75
0.8
0.85
0.9
0.95
A greater difference in the judgments of the businessmen by sectors was observed with
the specification of the tax or insurance payments. Five industries pointed to the personal
income tax as the most serious barrier. Three of the sectors indicated pension insurance as the
most serious obstacle, and two the VAT. Only in the "textile and knitwear production" was
43
the tax on profit regarded as the most substantial impediment. We have to make allowance
for the fact that healthcare insurance is distinguished from pension insurance, and this skewed
the assessment of the total insurance burden.
Table 21
Which of the Following Taxes, Fees and Social Insurance Payments
are the Most Serious Obstacles for Complete Legalization of the Businesses?
Sectors
Chemical
Construction
Textile and knitwear production
Retail
Agriculture
Transport
Food and drinks production
Wholesale
Mechanical and engineering
Tourism
VAT
(%)
Pension
Insurance (%)
Income Tax
(%)
Corporate Tax
(%)
69
61
52
49
52.0
66
59
67
54
63
66
On the whole the perceived significance of each tax is determined by the specific
character of the business and not by the existence of a discriminatory sectoral approach in the
adoption and enforcement of the tax legislation. For example, in the transport and tourism
sectors, competitiveness depends directly on the VAT rate and its application with respect to
foreign clients. It is these sectors which identify the VAT as a major barrier. No statistically
significant difference can be observed in the effects of the tax burdens calculated according to
the size of the enterprise.
Conclusions about the tax burden
1. The high tax burden is considered to be a major barrier to business development, the
main problem being the income tax and insurance rates; whereas in the opinion of
those interviewed tax administration inefficiency has a lesser effect. In this context
any economic measures aiming at decreasing the tax burden would positively
influence business environment and economic growth. Special attention must be paid
to reduction of social security payments, which are some of the most frequently
avoided payments.
2. According to those interviewed, on average 35.5% of the taxes due are evaded. This
amounts to about 33% of the GDP, the percentage being a little overestimated because
of the peculiarities of the methods for evading excise taxes. This form of the shadow
economy does not take into consideration the statistical omissions concerning full
coverage of the small businesses as well as totally illegal ("underground") business. A
quantitative evaluation of taxes avoided is not equivalent to the proportion of shadow
44
economy in the gross value added. These figures are clearly suggestive of the large
share of the shadow economy in the overall economy.
3. The considerable size of the shadow economy resulting from the tax burden is quite
evenly distributed across sectors.
4. In macroeconomic analyses and forecasts of potential GDP growth, as well as in
international comparisons of living standards, the shadow economy must not be
disregarded as a factor. It is also a considerable resource which can be used for
lowering the tax burden without causing a large reduction of fiscal revenues if the
taxable base is extended, and the risk and size of sanctions for tax violations are
increased.
The shadow economy and the labor market
One aspect of the informal economy is underreporting a part of earned incomes by
economic agents. A major reason for this is evasion of payments due to the state budget, the
National Social Security Institute, the Health Insurance Fund, etc. either by the individuals
themselves or by employers. Some economists claim that a decline in the economic activity
of the population is an indicator of intensification of underground occupations, ceteris
paribus.
Figure 20
Level of Economic Activity of the Population
56.00
54.00
52.00
% 50.00
48.00
46.00
44.00
IX
93
XII
93
III
94
VI
94
IX
94
XII
94
III
95
VI
95
IX
95
XII
95
III
96
VI
96
IX
96
XII
96
III
97
VI
97
IX
97
XII
97
III
98
VI
98
IX
98
XII
98
III
99
VI
99
IX
99
XII
99
III
00
Figure 20 shows a steady decline in the economic activity of the population. In 1996
and 1997, the labor force was 52% of the population aged over 16; whereas since 1999 it has
fallen to under 50%. This tendency should not be interpreted only as an indication of possible
growth of the informal sector, since the unemployment rate has increased in the second half of
1999.
45
A more detailed study of individuals leaving the labor force shows that we cannot
discard the hypothesis of declining economic activity as an indicator of a stream of people
towards the shadow economy. According to some estimates about 40% of the unemployed
between the ages 15 and 24 leave the labor force, and this percentage is nearly two times the
contingent of unemployed leaving the labor force in the older groups. Conceivable reasons
for this greater percentage in the youngest group are possibilities for continuing one’s
education, raising children, etc. If we analyze the streams of unemployed people leaving the
labor force in terms of sex and age, we can see that there is no significant difference between
departures of men and women. Instead there is a pattern more consistent with passage to the
informal sector than with departure for child rearing.
The research was intended to monitor the share of the informal sector in the labor
market. This phenomenon has two aspects:
1. Some individuals are hired without any labor or non-labor contracts. Their income is
not declared as an expense of the employing company, or as personal income of the
individuals;
2. Officially companies might declare lower wages than actually paid. In this way a part
of the enterprises’ newly created income is not officially declared. A modification of
the latter is hiring people to perform a particular job using a non-labor contract rather
than a labor contract. Thus until the beginning of 2000 the individuals had to make
their own social security payments.
Between 13% and 15% of the companies in the survey declared hiring people without
any contract in each of the years under examination (1997-1999), as well as during their first
accounting year. Nevertheless among those companies in the survey, there is a general growth
tendency in the total number of employees without any contract.
46
Table 22
Structure of Employment within the Firms Sampled (%)
First
accounting
year
Hired on a labor contract
Hired on a non-labor contract
Hired without any contract
94.99
2.80
2.22
1997
1998
1999
93.26
4.52
2.22
93.50
4.41
2.09
92.16
4.85
2.99
Until the beginning of 1999 individuals hired without any contract were 2% of the
total number of the employed by the firms, whereas in 1999 this figure grew to 3%. In this
same year the proportion of those employed on a non-labor contract increased by 0.4%.
Table 23
Change in the Number of Employed with Respect to the Previous Year (%)
1998
1999
Hired on a labor contract
Hired on a non-labor contract
Hired without any contract
10.86
7.79
3.90
-15.68
-5.94
22.34
Total number of employed
10.57
-14.46
The trend in the number of employed according to type of contract with employers is
particularly indicative. In 1998 the total number of employees in the companies included in
the sample grew by over 10%, whereas the number employed by a labor contract increased
faster than the number of those employed by other agreements. Those employed without any
contract grew in number too, but only by about 4%. In 1999 the total number of employees in
the companies included in the survey fell by over 14%. At the same time the number of
employees without any contract was the only category to grow - by an impressive 22%.
Regardless of their small proportion in employment as a whole, this change is indicative of a
general tendency in the preferences of enterprises to engage more and more people without
any contract.
47
Table 24
Allocation of Firms According to the Proportion of Hired Workers
Without Any Contract in the Total Number of Employed (%)
0%
Between 0% and 50%
Between 51% And 100%
100%
First
accounting
year
1997
1998
1999
85.84
7.53
3.01
87.16
6.56
3.01
87.12
5.81
2.78
85.07
8.29
3.32
3.61
3.28
4.29
3.32
The results reveal that approximately 2.5% - 3% of the employed are not registered,
i.e. between 70,000 and 90,000 work without contracts. Since the second half of 1999, the
statistics of the National Employment Service have been showing unemployment higher by
50,000 to 100,000 people than shown by the National Statistical Institute. Therefore it is
possible to regard these individuals as people who have an occupation during the year, or at
least during given months of the year, but do not declare it.
48
Figure 21
Share of the Employed Without Any Contract
to Total Employment in the Firms by Sectors (%)
Other services
Trade
Agriculture
Construction
Industry
0.00
2.00
4.00
6.00
8.00
First accounting year
10.00
1997
12.00
1998
14.00
16.00
18.00
1999
A more detailed analysis shows that agriculture and trade are the sectors where noncontract employment is most common. Over 16% of the workers in agriculture had no labor
contract during the first accounting year. In 1997 and 1998 they accounted for 12.4%, but in
1999 those who were employed without any contract accounted for 13.8% of the total of those
employed in agriculture. Among trade firms there is a continuously rising trend in the number
of the employed without any contract - from 3.8% in 1997 to 5.4% in 1999. In the remaining
three sectors, the proportion of employed without any contract is very small, and even among
industrial enterprises and service companies (with the exception of trade) it is as low as 0.5%.
It should be pointed out that the number of individuals employed without any contract among
industrial firms diminished during the period 1997-1999 in contrast to almost all other sectors.
Comparing the proportion of people hired without any contract to the total number of
employed in different sectors of the economy, it can be concluded that, for the most part, these
employees are seasonal or temporary, given their concentration in agriculture and trade.
When an individual is hired for a relatively long-term or even permanent job, it is almost
impossible to engage him or her without any contract.
49
Figure 20
Share of Employed Without Contracts
to Total Employment in Firms, by Type of Location (%)
Village
Small town
Regional center
Sofia
0.00
5.00
10.00
15.00
First accounting year
1997
20.00
1998
25.00
1999
Figure 21
Share of Employed Working Without Contracts
to Total Employment in Firms, by Type of Ownership (%)
Cooperatives
Private firms
State owned firms
0.00
0.50
1.00
1.50
2.00
2.50
First accounting year
1997
3.00
1998
3.50
4.00
4.50
1999
The data shown in Figure 21 is not surprising as far as it concerns the state enterprises.
No employment without contract is registered during any of the years reported. It is clear that
there is less scope and less incentive for state enterprise managers to engage in such tax
avoiding practices. We can also see from Figure 21 that the cooperatives are more willing to
50
hire workers without any contract than are private enterprises (it may be due to the greater
weight of the agricultural firms in the totality). However the differences are not as large as
with the other groups of firms in the totality of companies surveyed.
Figure 22
Reasons for Employed to Use Non-employment
in Place of Employment Contracts
60
59.13
51.19
50
40
33.5
29.46
%30
20
11.66
10
0
High taxes
and social
insurances
on wages
People with
the relevant
qualification
prefer these
Higher
flexibility
for human
resource
The Labor
Code creates
difficulties
for firing
Other
Approximately 60% of the employers interviewed answered the questions about why
they prefer to engage workers for particular activities on non-labor rather than on labor
contracts. In about 60% of the cases, the firms have not hired people using non-labor
contracts, and at least 20% of the firms answered the question although they have engaged
people only on a labor contract. The most important reasons for preferring non-labor to labor
contracts are:
First, the high salary-related costs to the firms (social security payments, health care
and unemployment insurance, etc.) – cited by about 60% of those who answered the question.
Second, the labor contracts are preferred as they facilitate achieving a better flexibility
while optimizing the number of employees in the companies – cited by more than 51% of the
cases.
Presumably the disposition to hire on non-labor contracts, rather than on labor
contracts, is an indication that a given employer is more willing to engage people without any
contract. On this basis we can estimate the average number of individuals employed without
any contract within the total employed by the firms, which have specified any of the reasons
for preferring a non-labor to a labor contract.
It can be concluded that:

Employers seek to hire people using non-labor rather than labor contracts with the
purpose of reducing payments due to the state budget, the National Social Security
51
Institute, etc. In fact, they transfer the responsibility of paying to the individuals
who are to be insured;

Hiring people without any contract is used in cases of temporary or seasonal work.
In this way a greater flexibility can be achieved. Also if an individual turns out to
be inappropriate for the job, it would be easier to dismiss him or her. And it is
likely, if a person appears to be particularly valuable for the firm, after an initial
trial period during which the employed has worked without any contract, to offer
him or her a long-term labor contract.
Table 25
Average Percentage of Employed Without Contracts in Total Employment
to Firms Indicating a Specific Reason for Hiring
on Non-labor Rather than Labor Contracts (%)
1997
Because required payments related to the employees’
salaries are high (insurance, etc.)
Because individuals with appropriate job
qualifications prefer this type of contract
Because this type of contract allows a greater
flexibility for the human resource management
Because under the existing Labor Code, it is difficult
to dismiss employees who do not have
appropriate qualifications
Other
1998
1999
6.24
6.62
9.66
5.70
8.80
9.00
8.30
8.50
11.09
7.89
8.99
13.32
16.46
15.70
13.42
Another aspect of tax and insurance evasion is officially reporting salaries that are
lower than that actually paid in order to reduce the burden of the various payroll taxes. Until
the beginning of 1999, another way of achieving the same result was hiring on a non-labor
contract, especially when the individuals who were to be engaged had a labor contract with
another employer and were, therefore, already insured.
The participants in the study were consecutively asked if they knew any companies
where the officially declared salaries differed from the ones that were actually paid and if so,
what was the approximate difference between them. Nearly two-thirds of those who answered
the question (over 62.5%) believe that it is a common practice for most firms and another
17.4% think that there is such a phenomenon but few companies really practice it. Only 6.3%
of the interviewed are of the opinion there is no such thing as paying more than the officially
declared salaries or, at least, they do not know of such firms.
52
Figure 23
Do You Know of Firms Where Official Wages
Differ from Wages Paid in Reality?
13.81
Yes, this is a common
practice
6.28
Yes, I know of a small
number of such firms
No, I do not know such
of firms
17.36
62.55
I do not want to answer
It can be concluded that not declaring total employees’ salaries is much more common
than hiring people without any contract, and it is a much more significant way of expanding
the informal sector of the economy.
If we assume that the respondents have answered in reference to information about the
concealed real salaries in firms similar to theirs (and indirectly in their own companies), there
are no significant differences in the structure of the answers by sectors, type of municipality
and property. The percentage is highest in trade and construction enterprises, where it is quite
commonly considered that salaries paid are different from officially stated salaries. The
proportion of industrial firms among the participants who refused to answer this question is
relatively high. This may be due either to lack of information or can be interpreted as an
indication that in this sector the practice is fairly common.
53
Table 24
Do You Know Any Firms Where the Officially Declared Salaries
Differ from the Actually Paid? (% of Responses)
Yes, it is a
common
practice
Total
Yes, but in
few firms
There are no
such firms
I do not want
to answer the
question
62.55
17.36
6.28
13.81
62.32
66.67
54.88
69.29
56.96
15.94
20.51
26.83
12.86
16.46
3.62
2.56
6.10
7.14
11.39
18.12
10.26
12.20
10.71
15.19
According to the type of municipality:
Sofia
54.29
Regional center
67.84
Small town
57.97
Village
57.14
15.71
14.90
20.29
23.81
11.43
3.92
8.70
7.14
18.57
13.33
13.04
11.90
According to the type of property:
State enterprises
50.00
Private firms
64.34
Cooperatives
51.06
31.25
15.90
25.53
12.50
5.54
10.64
6.25
14.22
12.77
Including:
Industry
Construction
Agriculture
Trade
Other services
Note that in the bigger cities (with the exception of Sofia) the underreporting of the
salaries actually paid is much more common than it is in the other municipalities. Also the
proportion of those who have answered that there are no such firms is only 3.9% in the
regional centers, while in the other groups of enterprises it is higher, and for the companies in
Sofia it is 11.4%.
A relatively high percentage of the private enterprise firms - 64.3% - also think that
actual salaries are higher than those reported. This is also the group where the smallest
number of the firms answered that there is no such phenomenon in practice – about half as
many as in the other groups of companies. It can be assumed that the private firms, to a
greater extent, practice declaring smaller salaries than actually paid in order to evade
payments to state authorities.
54
Figure 25
Approximate Percentage of the Difference
Between the Real Wages Paid and Official Wages
50.00
44.85
45.00
40.00
35.00
30.00
% 25.00
24.01
20.00
15.00
11.61
10.00
11.61
7.92
5.00
0.00
Below 10%
No such
practice
Between 10 and
50%
Over 100%
Between 50 and
100%
The respondents who expressed the opinion that there is such a thing as declaring
lower salaries than those actually paid were asked the difference measured in minimum
salaries. Almost half answered that the salaries actually paid were higher than those declared
by 10% to 50%. In the opinion of those interviewed, the salaries actually paid were higher
than those officially declared by about 46%. In other words if the average monthly salary in
mid-2000 was around 230 BGN according to the National Statistical Institute, then, according
to the answers of those interviewed, the monthly salary actually paid was about 350 BGN.
Table 26
Average Monthly Salary in the Economy for 1999
by Sectors (preliminary data)
Industry
Construction
Agriculture
Trade
Other services
Total:
Average Salary
for 1999
Difference between Actual
and Official Salaries
Actual Salary
BGN
%
BGN
307.81
192.37
177.74
171.05
211.32
52.13
58.68
35.48
47.07
37.14
468.27
305.25
240.80
251.56
289.80
205.05
45.98
299.33
Source: NSI, Firms Survey
55
The surplus of actual over official salary varies by groups of firms. This difference is
the greatest for construction and the industrial sectors, where the actual salaries are higher
than declared by 59% and 52% respectively, and the difference is the smallest in services (not
including the trade) and agricultural sectors where the spread between actual and official
salaries is 37% and 35%, respectively.
Salaries registered by the NSI for the industrial sector are 73% higher than those for
the agricultural sector. At the same time the surplus of salary actually paid according to the
answers of the industrial firms is 47% higher than is the estimate of the agricultural
enterprises. If we assume that all those interviewed have answered mostly on the basis of
their own experiences and the practices of firms similar to theirs, then it appears that the
actual salary in the industry is approximately two times higher (about 94%) than that for the
agricultural sector.
According to data reported, the average salary in construction is 6% lower than the
average for the country, and 9% lower than other services (except for trade), whereas the
salary actually paid for construction is the second highest in the country and is 2% greater than
the average salary.
Analyzed according to municipality the difference between the two types of salaries
decreases as the settlement size decreases. It is largest for Sofia and the regional centers
where the difference reaches 50-51%; in the small towns it is around 42%; and in the villages
it is only 31%. According to data from the NSI, higher salaries are generally paid in the larger
municipalities.
Figure 26
Difference Between the Real
and the Officially Declared Wages (%)
Sofia
51.2
Regional center
50.14
42.14
Small town
Village
30.66
0
10
20
30
%
56
40
50
60
Table 27
Average Monthly Salary and the Surplus of the Actual Over Official Salary
by Sectors for 1999 (preliminary data)
State firms
Private enterprises
Cooperatives
Total
Average salary
for 1999
Difference between
actual and declared
salary
Actual salary
BGN
%
BGN
221.28
183.19
35.83
48.33
29.75
300.57
271.72
205.05
45.98
299.33
Source: NSI. - Firms Survey
As expected, private firms taking part in the survey shared the opinion that the salaries
actually paid were much higher than those in the state enterprises. Moreover, the state
enterprises estimated the difference as being larger in comparison to the cooperatives in the
survey. It can be presumed that state enterprises also under report salaries, however to a lesser
degree than private enterprises. The fact that state firms do not hire people without any
contract must be taken into consideration. Hence even managers of state enterprises are
probably trying to cut payments because of insurance funding by declaring lower than actual
salaries instead of using non-labor contracts or no agreements at all.
The data provided by the NSI shows that the proportion of salary in total employee
compensation (here we include the salary as well as all additional payments set by the Labor
Code, insurance payments, etc.) was around 71% in 1997 and 1998. Compensation for the
employed was 5,900 and 8,500 billion BGN respectively for 1997 and 1998. Let us assume
that participants in the survey have correctly evaluated the difference between the registered
and the salaries actually paid (approximately 46%). Then the compensations that were, in
reality, paid would be 7,800 and 8,500 billion BGN respectively for 1997 and 1998. If we
presume that the salary-to-compensation ratio does not change significantly if firms declare
salaries actually paid, then the enterprises’ labor expenses would be 8,600 and 12,400 billion
BGN for 1997 and 1998, respectively.
57
Table 29
Compensations of the Employed and Salaries in 1997 and 1998
(million BGN)
1997
1998
Salary
Other payments
Compensation of the employed
Salary actually paid31
Other payments
Compensations actually paid
4.22
1.67
5.89
6.15
1.67
7.82
6.00
2.47
8.47
8.77
2.47
11.24
Outstanding compensations due to the employed32
8.59
12.37
Source: NSI, Basic macroeconomic variables ‘98 - NSI, p. 81-82.
Therefore the underreporting of approximately one-third of real salaries brings about
10% savings in labor costs for employers, which is a considerable incentive for
misrepresentation of the salaries actually paid to the employees.
If the assumptions made in Table 29 are correct, then the firms do, in fact, not declare
33% of enterprise income allocated for labor costs. Moreover we have made the conclusion
that about 2.5% - 3% of the employees work on no contract at all, and therefore the costs of
their labor for the enterprises is either not declared or not reported as a newly created revenue.
It can, therefore, be concluded that most probably firms hide around 34%-35% of income
which is actually used to pay their employees.
TAX AND SOCIAL INSURANCE BURDEN ON THE LABOR COSTS
For the purposes of investigating the informal economy, we use the following concept
of tax and insurance burden on employee’s income: all payments to the state (including
personal income tax and mandatory insurance) as a percent of total labor costs per employee
(including gross wage and insurance installments "paid by employer".)33 In this manner we
can easily track the shifts of demand for labor and effects on business environment
In Table 30 the tax and insurance burden in the second half of 1999 is presented.
Estimated on the basis of officially reported average gross wages in the private sector, the
burden is 42.4%. This means that when the gross wage (i) is BGN 177.68, which translates to
a net wage (i - iii - iv) of BGN 144.48, the total labor costs (i + ii) are BGN 250.88. If a given
firm, having average labor costs of BGN 250.88, chooses to operate legally, its employees will
earn BGN 144.48 on average. However if the firm chooses to operate "in the shadow," it
probably will be able to hire the same employees for the same net pay without any additional
31
The NSI data for salaries is increased by 46%.
Insurance and other payments obtained from NSI data have been indexed to estimate actual salaries.
33
According to the Bulgarian Social Security Code, both the employee and the employer formally pay the
insurance installments; this, however, means that in order to obtain the net wage, only part of all insurance
payments should be subtracted from the gross wage.
32
58
labor costs, i.e. if the company has "gray" employment, it will have to increase its labor costs
by 74% in order to leave the shadow economy.
Table 30
Tax and Insurance Burden in 1999
(personal income tax plus mandatory insurance as a percent of the total labor costs)
1999
Gross
Wage
Total
Labor
Cost
Insurance Paid
by Employee
Income
Tax
Tax-Insurance
Burden (%)
ii
i + ii
iii
iv
(ii+iii+iv) /
(i+ii)
177.68
166.10
166.76
73.20
68.43
68.71
250.88
234.53
235.47
8.00
7.47
7.50
25.20
22.19
22.36
42.41
41.83
41.86
141.95
177.74
193.26
58.48
73.23
79.62
200.43
250.97
272.88
6.39
8.00
8.70
15.91
25.21
29.25
40.30
42.41
43.08
i
Total private
Agriculture
Construction
Retail and
wholesale
Transport
Manufacturing
Insurance
Paid by
Employer
Source: NSI and IME calculations
Note: Average gross wage according to the NSI in November 1999.
The differences among the tax and insurance burdens among sectors results from
different average wages, which means a different personal income tax rate (because personal
income taxation is progressive). Still the differences are negligible and probably do not
contribute significantly to an incentive to operate "in the shadow". Also it is important to
stress the fact that the officially reported average wage in the private sector is probably
considerably lower than the actual worker’s pay. Therefore, because of the higher rate, the
real burden should be considered higher, although the majority of companies escape personal
income tax.
Table 31 presents the burdens in the year 2000. The changes result from new social
security legislation and changes in the personal income taxation. The methodology of
estimation is the same as in Table 30, and the result is a slight increase in the average burden
for the private sector - from 42.4% to 44.2% - as the differences between the branches stay
stable. In this way we find that, in order to operate legally, a company has to spend, on
average, BGN 277.5 per employee, otherwise the company would have to spend BGN 154.93.
This means that, in order to get out of the shadow, the company with gray employment should
increase its labor costs by 79%. Thus, compared to 1999, the incentives to stay "in the
shadow" have increased.
59
Table 31
Tax and Insurance Burden in 2000
(personal income tax plus mandatory insurance as a percent of total labor costs)
2000
Total private
Agriculture
Construction
Retail and wholesale
Transport
Manufacturing
Gross
Wage
Insurance
Paid by
Employer
Total Labor
Cost
Insurance
Paid by
Employee
Income Tax
i
ii
i + ii
iii
iv
(ii+iii+iv)/
(i+ii)
203
177
178
172
221
213
74.5
64.96
65.33
63.12
81.11
78.17
277.5
241.96
243.33
235.12
302.11
291.17
29.88
23.12
23.38
21.82
34.56
32.48
44.2
42.99
43.04
42.71
44.87
44.59
Note: average gross wage according to the NSI in May 2000
Sources: NSI and IME calculations
60
18.27
15.93
16.02
15.48
19.89
19.17
TaxInsurance
Burden (%)
V. CONCLUSIONS
The research on the shadow economy using the physical input method indicates that
the relative share of shadow economy has fluctuated between 25% and 37% of GDP over the
1989-1998 period. If we use an assumption of 30% SE in 1989, the SE economy in 1998 as a
share of official GDP is estimated to be 22%. The absolute peaks of the shadow activity are
calculated for 1990 (32.2%) and 1996 (34.4%). We observed a declining trend since 1996 in
the relative share of the shadow economy. This is not surprising given the fact that the
currency board’s introduction in mid-1997 marked a steady rise in tax revenues to GDP. The
high inflation period of 1990-1995, the hyperinflation in 1996 and the beginning months of
1997 boosted the nominal tax revenues at a much faster rate than social compensations, thus
creating strong incentives for households and businesses to escape from the official economy.
With macroeconomic stabilization and the accompanying price and trade liberalization, the
shadow economy has shrunk. The process, however, succeeded only partially, as the
administrative and the tax burdens remain the main influence for the businesses to operate in
the informal sector.
Examining energy use, rather than relative shares, is basically a good tool for assessing
of the dynamics and trends of shadow activities. The base-year relative share must always be
given as an external value for the model, and the only way to assess this value is the micro
approach. The physical approach also measures the final results, but reveals nothing about the
factors that affect the shadow economy. A survey approach to the microenvironment
represents the only tool that can unveil the complete picture. We divide the conclusions of
our micro research into the following sections:
Firms’ performance and development

Financial statements of the enterprises are relatively stable, but firms with high
sales volume demonstrate better profitability and stronger growth. The companies
with an annual turnover above BGN 200,000 are the main contributors to growth
in the economy, while SMEs experience a serious burden that restricts their
expansion.

The bank and non-bank financial intermediation is low, and personal savings are
the main source of starting capital;

Collateral requirements, the complicated procedures for receiving a bank credit,
lack of business history, and high interest rates are the most commonly cited
obstacles to receiving bank credits;

Businesses state that major barriers to business expansions are: tax burden,
insufficient financing, complicated and unpredictable legislation and inefficient
regulation that impedes the entry of SMEs;

Businesses incur significant losses, measured by the opportunity cost of the time
spend for obtaining licenses or/and permissions (an average of 4-5 days per
month).
61
Estimation of the effect of tax and insurance burden on the shadow economy

The high tax rates and social insurance payroll taxes are the major barriers to the
development of the businesses;

Tax administration has a lesser effect in the opinion of those interviewed, but the
time spent solving problems with taxes, local fees and social contributions is still
significant - 7.5 days per month or about a third of working hours;

VAT and social insurance payments are most frequently avoided by businessmen;

The shadow economy is quite evenly distributed among the sectors studied in the
survey;

An average of 35.5% of the taxes due are avoided or evaded, which amounts to
some 33% of GDP, which is a huge resource that could potentially be used for
lowering the tax burden without causing a fatal reduction of fiscal revenues.
Labor market

The share of those employed without contract in trade, agriculture, construction
(mainly for seasonal and temporary work) is the most widespread form of shadow
activity in the labor market;

Most of the companies register only a small proportion of the wages actually paid
in order to avoid social security contributions. There is a large deviation among
actual salaries, reported in industry, agriculture, construction, trade and service
firms.

The firms hide about 34%-35% of income allocated to labor compensation.
Cash transactions and resistance to requirements for proof of money origin are
indicators that indirectly show the large spread of the shadow. Screening the respondents, we
estimated that about 70% of all transactions are conducted in cash. Most of the companies
have some experience in the shadow economy. Quantitative indicators composed by
integrating the different responses and weighed by the number received, highlighted the
following factors that motivate shadow economy activity: high tax and social insurance
(77.8%); need to prove the origin of money (56%); difference between actual and reported
salaries; avoidance of customs duties (42.5%), and licensing requirements (40.8%).
To a large extent the research confirms Feige’s concept that shadow activities are a
form of non-compliance with excessive state regulations concerning licenses and permissions,
and the tax and social security systems.
Summarizing the results of the research, the following issues should be considered by
policy makers if they want to eliminate or minimize the shadow economy and improve the
business environment:
62
-
License and permission requirements;
-
Level of tax and social insurance burden;
-
Legislation related to entry and exit of businesses;
-
Institutional capacity and enforcement of laws;
-
Lack of broader public control over the public administration and institutions;
-
Corruption and red-tape.
Finally, the shadow economy reduces the overall efficiency of the economy, as
businesses are concentrated on efforts to avoid the administrative burden rather than
increasing productivity. Companies engaging in shadow activities also have to refrain from
some public and private services (e.g. bank transactions). This worsens the economic
competitiveness that is crucial for a small export-oriented country like Bulgaria.
63
RECOMMENDATIONS
Policy makers at all levels of state administration have to solve the problems of the
shadow economy through decisions and measures designed to bring results in the long-run. In
that context, the natural political contradiction between the short-term pre-election period and
the long-term benefits should be neglected. Concrete decisions should focus on efforts to:
-
Improve the business environment - the number of licenses and permissions should
be reduced; and in cases where this is not possible, their implementation should be
removed from the discretion of the ruling bodies;
-
Reduce the tax burden through lower tax rates and speed up VAT reimbursements;
-
Upgrade the institutional and the administrative capacity that will augment
efficient and timely enforcement of laws;
-
Enhance transparency of rules and regulations as a way of diminishing the
possibility for rent seeking;
-
Create conditions for effective citizens' control over the activities of state
institutions and local authorities.
64