Available online at www.sciencedirect.

com

ScienceDirect
Available online at www.sciencedirect.com
Procedia Computer Science00 (2018) 000–000
www.elsevier.com/locate/procedia
ScienceDirect
Procedia Computer Science 120 (2017) 333–340

9th
9thInternational
InternationalConference
ConferenceononTheory
Theoryand
andApplication
Applicationof
ofSoft
SoftComputing,
Computing,Computing
Computingwith
with
Words and Perception, 22-23 August 2017, Budapest, Hungary
Words and Perception, ICSCCW 2017, 24-25 August 2017, Budapest, Hungary

Estimation of impact of the changes made to the tax legislation to

the tax receipts through fuzzy numbers
Akif Musayevabcd, Selin Uzelaltinbulatd, Samira Mammadovab , Latafat Gardashovab*,
Aygun Musayevac
a
Institute of Economics, Azerbaijan National Academy of Sciences, Baku, Azerbaijan
b
Azerbaijan State Oil and Industrial University, Baku, Azerbaijan
c
Azerbaijan University, Baku, Azerbaijan
d
Near East University, Nicosia, TRNC;

Abstract

The methodology for estimation of the impact of changes made in legislative base of tax system to the tax revenues is proposed
in the paper. The adjustment of time series of tax revenues and relevant computer simulation of them was conducted by the fuzzy
numbers according to the expert evaluation. Double expert estimation methodology was used in order to ensure the maximum
reliability and truthfulness of the existing information. The relevant fuzzy regression dependency of gross domestic product
(GDP) on the time series of tax revenues, the relevant fuzzy regression dependency on the time series of adjusted tax revenues
was arranged.

© 2018 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the scientific committee of the 9th International Conference on Theory and application of
Soft Computing, Computing with Words and Perception.

Keywords: tax receipts; expert evaluation; GDP; tax reform; fuzzy regression.

1. Introduction

The analysis of tax policy, forecasting of tax revenues, stability of tax system from point of sustainability of

* Corresponding author. Tel.: +994505840901.

E-mail address: latsham@yandex.ru

1877-0509 © 2018 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the scientific committee of the 9th International Conference on Theory and application of
Soft Computing, Computing with Words and Perception.

1877-0509 © 2018 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the scientific committee of the 9th International Conference on Theory and application of Soft
Computing, Computing with Words and Perception.
10.1016/j.procs.2017.11.247
334 Akif Musayev et al. / Procedia Computer Science 120 (2017) 333–340
2 Akif Musayev et al./ Procedia Computer Science00 (2018) 000–000

revenues to ensure the state expenses are issues of decisive importance for all states. That’s why, the national
institutes of tax policy implement necessary structural, administrative and legislative reforms for continuous
improvement of tax system to meet to challenges of the global economic environment.
The tax policy institutions are obliged to take into account many factors in this process at the same time.
The impact of the economic growth and tax reforms on general tax revenues and structure is analyzed by
Kanghua et al (2013). As it is noted in the period of 1950-2011 7 tax reforms were conducted in China and the
current tax system has been formed and included value added tax and profit tax from entities. This article reviewed
economic evaluation of 7 tax reforms, the essence of the impact of the tax reforms on the tax revenues and structural
changes, the impact of the mutual relationship of the economy with taxation. The impact of the tax reforms on
economic growth was studied over empiric test and empirical analyses. The authors analyzed the impact of the
economic growth and tax reforms on tax revenues and tax structure with regression model.
The impact of the taxes on economic growth was studied by Santiago et al. (2012). The authors have
investigated the connection of the long-term economic growth with change of the tax base according to studies of
the countries having high, medium and low income level at least within 20 years. The goal of the article is to detect
the restrictions by studying the connection between the tax structure and the long-term growth. The authors
reviewed the methods offered by Pesaran (1999), Arnhold (2011), Kneller (1999) in this field, conducted deeper
researches based on wide data, analyzed the impact of the main taxes on the economic growth, but they also used the
empiric method.
The impact of the tax structure on the economic growth, the empiric evidences of OECD member countries in
this field were reviewed by Jing Xing (2011). The evaluation of the impact of income-neutral tax structure on the
long-term per capita income level was conducted for the period of 1970 – 2004. It has been noted that there are a
few researches dedicated to relationship of economic growth with tax structure. The changes occurring in tax system
of OECD countries were reviewed. As we know, there are many disputed questions in taxation system and this
causes various scientific discussions. The advantages of consumer taxes in comparison with the profit tax are
disputed in theoretical and empirical studies. For example, abovementioned author did not discover the advantage of
the consumer taxes against of profit tax or the corporate tax against personal income. The only conclusion to be
noted is that change of the tax revenues is associated with long term higher income per capita in the direction of
property incomes.
Timothy and Torsten Persson (2014) have analyzed the problem of low taxes in developed countries. They note
that in order to perceive the taxation, economic growth and the mutual relations between them, it is necessary to
think about management of this growth process. This article reviewed political, social and cultural factors along with
the factors influencing the economic structure (such as weak management institutions, non-availability of
transparent information). The authors suggest the necessity of the dynamic approach for each of the above
mentioned factors and this creates bilateral relationships between political, or social, or cultural factors and
economy. Thus, it is evident that taxation is much more complicated problem than only tax collection, stable
taxation is one of the main features of developed state of the government.
Taxation of incomes and economic growth, the issues of empiric analyzes of 25 rich countries of the OECD were
reviewed by Margareta and Asa Hansson (2012). The impact of the taxation of incomes on economic growth was
analyzed based on the indicators of 1975-2010. It was determined that tax deductions from corporate and personal
incomes make negative impacts on the economic growth. The indirect methods were used in scientific researches
conducted in this field, including analysis of the connection of general state expenses and economic growth, and
studying of the impact of the tax deduction on economic growth. Only in some scientific researches (Koester and
Kormendi (1989); Plosser (1992); Slemrod (1995)) direct connection was analyzed. The tax system reform is used
in many countries of the world in order to strengthen the competitive advantage of the country. The indicators
expressing the increase of demands for social protection and strengthening of tax competition make the methods of
increasing tax efficiency more important, it expresses the necessity of studying the impact of various kinds of taxes
on economic growth.
The design of the tax structure enabling to ensure economic growth is investigated by Asa et al. (2008); new
grounds are created by indicating that the changes in tax deduction cause significant increase in some cases based on
the indicators used in fields and individual campaigns. This practical tax reform requires creation of balance
between efficiency, equality, simplicity and increase of income. The article reviews the tax structure and general tax
direction, investigate the per capita GDP of tax system structure, the impact, use and labor productivity through its
components, the ways directing the generalized tax construction of reform are described in general. The possibility
 Akif Musayev et al. / Procedia Computer Science 120 (2017) 333–340 335
Akif Musayev et al./ Procedia Computer Science00 (2018) 000–000 3

of impact of total economic activity on the economic growth is investigated by taking into consideration the features
of each country.
Desislava et al. (2012) studied the main directions in distribution of final tax load in European Union countries in
the period of 1995-2010 and their impact on the economic growth. The comparative analysis of final tax load of
various countries shows that the GDP is measured with tax norm and construction of tax structure leads to study of
the impact of the taxation on the economic growth. The used methodology is the regression analysis. The regression
analysis method is used with the purpose of testing substantiation of Davoody and Zou forecasts. The regression
method was built on conceptual structure described and the income level of per capita GDP was accepted in
regression equation as – dependent variable, and various types of income taxes and public expenses were accepted
as independent variable.
Evaluation of tax buoyancy, elasticity and stability is reviewed by Jonathan (1998). This article reviews the
methodology of calculation of trend and elasticity of increase of tax revenues in the conditions when of GDP
increases. The reports are conducted based on the data of Madagascar:
a) The increasing trend is calculated for each year and the average indicator is set;
b) The increase of tax collection for the recent years, calculation of GDP and calculation of increasing trend is
used;
c) Calculation of GDP base of increase of tax collections for recent years;
d) Analysis of the diagram of annual tax collections with the purpose of achieving the growth rate of the tax
collections. Growth rate is accepted as fuzzy variable ratio (annual).
First of all, it is clear that for business environment, at the same time, for business subjects the stability of tax
system is more important. Thus, tax system changing frequently does not enable the business subjects to plan their
future businesses correctly. At the same time, serious changes in tax system either creates relevant difficulties for
associates of tax authorities in correct performance of their duties or creates financial difficulties as this requires
serious transformation costs for state authorities. In this context, either structural, or administrative or legislative
changes provided for the tax system requires serious evaluation. If we consider that evaluation of the changes until
application require comprehensive changes of information base for evaluation of collection, then we can conclude
that the reviewed problem is so complicated, but the settlement of it is important.
One of the factors required to be stressed particularly is ability of changes to be expressed with quantitative
terms not only until application of the changes but also after application.
This article reviews the problem of estimation of the impact of the changes conducted in legislative base of the
tax system on tax revenues.
The paper’s structure is organized as follows. Section 1 is devoted to the review, analysis and comparison of
known researches. In the Section 2 we discuss proposed methodology for estimation of the impact of the changes
conducted in legislative base of the tax system on tax revenues. Results of computer simulation are described in
Section 3. The conclusion is presented in the Section 4.
.

2. Formal description of the proposed methodology

The modern scientific researches describe some approaches to the settlement of the problems mentioned above.
We can show the following studies as an examples of application of these approaches: the researches conducted in
Harvard University (Glenn et al.(2000))and our previous research( in Berkley University of the USA, 2014 and in
Vienna, Austria , 2016). We should note that the researches conducted in Harvard University are based on
adjustment of time series of tax revenues and calculation of buoyancy and elasticity ratios on the basis of adjusted
time series and GDP time series data of relevant period and acceptance of their difference as value of changes,
assessment of impact of the administrative and legislative changes on the tax receipts with the belonging functions
based on the Mamdani type logical extract on the basis of the expert values obtained in the result of double expert
evaluations. Adjustment of time series of tax revenues by fuzzy numbers based on the expert evaluations and
relevant simulations through them were conducted in the present research work. In order to ensure reliability and
maximum conformity of the database, double expert estimation methodology is used. On the first stage, it is
supposed that m number of changes is provided for the beginning of the next financial year in the legislative base of
tax system (in Azerbaijan, on the first of January of the next year). The impact of each of these changes on tax
revenues is assessed by n number of experts in 100 point scoring system (Table 1):
336 Akif Musayev et al. / Procedia Computer Science 120 (2017) 333–340
4 Akif Musayev et al./ Procedia Computer Science00 (2018) 000–000

Table 1. First group expert assessment

Changes in legislation Expert 1 Expert 2 … Expert n

1
q~11 q~12 q~1n
2
q~ 21 q~22 q~2 n
…
M
q~m1 q~m 2 q~mn

On the second stage, the second group of n number of experts estimates the truthfulness of the estimation of the
first group of experts between [0,1] (Table 2):

Table 2. Second group expert assessment

Experts (Group II) Expert 1(Group I) Expert 2(Group I) … Expert n (Group I)

1 ~
r11 ~
r12 … ~
r1n
2 ~
r ~
r … ~
r
21 22 2n
… .... .... … .....
m ~
rm1 ~
rm 2 … ~
rmn

The view of both groups of experts in respect to impact of changes made in tax legislation on tax revenues is
aggregated based on data in the first and second tables

 q~ ij ~
rij
~
yi 
j 1
, i  1, m
n

 ~rij
j 1

Here, ~
y i is expression of impact of the first changes made in tax legislation on tax revenues in percentage and
is triangular shaped fuzzy number (Fig.1.).

 x  a1
 a  a , if a 1  x  a 2
 2 1
a  x
 A ( x)   3 , if a 2  x  a3
 a 3  a 2
0, otherwise



Fig. 1. Triangular shaped fuzzy number

As a result, we obtain the line of values characterizing impact of each change made in tax legislation on total tax
revenues.
~y , ~y ,…, ~y
1 2 m

We obtain expression in percentage of the changes made in the financial year where the tax legislation is
reviewed as follows
 Akif Musayev et al. / Procedia Computer Science 120 (2017) 333–340 337
Akif Musayev et al./ Procedia Computer Science00 (2018) 000–000 5

~ 1 m
Dt   ~ yi
m i 1
Here, t is the financial year reviewed.
It is necessary to note that relevant estimation is conducted for collection of changes made in each t financial
year and is regulated as the line of Tt tax revenues
~
~ ~ T 
ATt 1  Tt 1   t ~ ~

 Tt  Dt ;

(1)

So, the time series of initial tax revenues

~ ~ ~
T1 , T2 ,..., Tk ; (2)
is substituted with the time series of adjusted tax revenues
~ ~ ~
AT1 , AT2 ,..., A Tk ; (3)
Relevant regression dependency is arranged for the time series of GDP (GDP) (2)

~ ~ ~
GDP1 , GDP2 ,..., GDPk ;
~ ~
T a 0  a1GDP (4)
Regression dependency is arranged for relevant line of adjusted (3) tax revenues

~ ~
AT b 0 b1GDP (5)

It is clear that the forecast tasks defined for the next year according to (4) and (5) equalities will be different and the
forecast task defined in accordance with (5) equation will be conforming to reality.

3. Computer simulation

Let’s look through the above mentioned methodology on the economic indicators of Azerbaijan Republic. 18
experts have evaluated the impact of each of 18 changes in legislation on tax receipts in the scoring interval [0-100],
the degree of confidence on this evaluation in the interval of [0-1] and the results were given in the Tables 3, 4.
The view of both groups of experts in respect to impact of the changes in tax legislation on tax revenues was
aggregated based on the data in the Tables 3 and 4, and the following result was obtained for 2001:

18

 q~  ~r
j 1
ij ij
~
~
y1   19. 8 862
18

 ~r
j 1
ij

Here, ~
y1 is expression of impact of the first changes made in tax legislation on tax revenues in percentage.

As a result, we obtain the line of values ( ~

y1 , …, ~
y18 ) characterizing the impact of each change made in tax
legislation on the total tax receipts:
6 Akif Musayev et al./ Procedia Computer Science00 (2018) 000–000
338 Akif Musayev et al. / Procedia Computer Science 120 (2017) 333–340

Table 3. The indicators of experts for evaluation of legislative changes applied in tax system
Changes in legislation Expert 1 Expert 2 Expert 3 … Expert 18
A About 25 About 15 About 10 About 15
B About 20 About 0 About 15 About 55
C About 24 About 12 About 40 About 23
… … … … … …
M About 40 About 50 About 30 About 10
O About 13 About 24 About 15 About 7
R About 40 About 7 About 0 About 8

Table 4. The truthfulness indicators of evaluation by second group of experts of the first group of experts
Changes in legislation Expert 1 Expert 2 Expert 3 … Expert 18
A About 0.3 About 0.1 About 1 About 0.8
B About 0.1 About 0 About 0.1 About 0.1
C About 0.3 About 0.4 About 0.4 About 0.2
… … … … … …
M About 0.7 About 0.83 About 0.03 About 0.6
O About 1 About 0.75 About 0.5 About 0.1
R About 1 About 0.8 About 0.4 About 0.62

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
(19. 8 6, 17. 5 0, 18. 5 71, 18. 7 38, 13. 1 818, 10. 6 812, 19. 3 74, 14. 1 39, 18. 7 83, 16. 9 77, 19. 3 916, 19. 8 821,
~ ~ ~ ~
17. 3 924, 15. 3 649, 17. 4 766, 0, 0, 17. 7 019)

We obtain the expression of changes made in the tax legislation in the financial year reviewed in the next step as
follow (Table 6):

18

 ~y  15.2803
~ 1 ~ ~ 1 ~
D2001  i or D2001  * (261,2931 275,0454 288,7977)  (14,51629 15.2803 16,04432)  15. 2 803
18 18
j 1
This was determined for financial year 2001. We obtain the time series for the years when changes were made in
legislation in the same manner

~ ~ ~ ~ ~ ~ ~ ~ ~
( D2001 , D2003 , D2004 , D2005 , D2006 , D2007 , D2009 , D2012 , D2013 )

~ ~ ~ ~ ~ ~ ~ ~ ~
(15. 2 803, 14. 1 647, 13. 8 932, 15. 5 17, 14. 2 57, 13. 2 92, 14. 5 912, 14. 2 5, 13. 5 17)

So, the time series of initial tax receipts

~ ~ ~
T2001, T2002 ,..., T2015 ;
~ ~ ~ ~
For example, T2001  37 4 .6132 …. T2015  17 0 6.198

~ ~
T2001  (550,62 579,60 608,58)/ (1,47 1,55 1,62 )  37 4 .6132

In this case, the tax receipts are set as the follows (Table 5).
~ ~ ~ ~
AT2001  12 7 4.05 … AT2015  17 0 6.198
The regression dependency is arranged according to the relevant line of GDP relevant to the time series of tax
receipts
~ ~ ~
GDP2001, GDP2002 ,..., GDP2015 ;
 AkifAkif Musayev
Musayev et Procedia
et al./ al. / Procedia Computer
Computer Science
Science00 120 (2017)
(2018) 333–340
000–000 3397

~ ~
T  29.85729  0.1149  GDP
and regression dependency is arranged relevant to the time series of adjusted tax receipts
~ ~
AT  352.672  0.119119  GDP

The fragment of the fuzzy numbers of GDP and tax was given in Table 5.

Table 5. The fragment of Real GDP and of tax represented fuzzy numbers
Years Real GDP(AZN) Tax(AZN)
2001 (3263,853;3435,635;3607,416) (355,8825;374,6132;393,3439)
2002 (2078,323;2187,708;2297,094) (233,0807;245,3481257,6155)
2003 (3039,206;3199,164;3359,122) (350,1689;368,5988;387,0288)

2014 (40453,24;42582,36;44711,48) (4879,331;5136,138;5392,945)

2015 (12376,54;13027,94;13679,33) (1620,888;1706,198;1791,508)

It is clear that according to the equations for T and AT, the forecast tasks defined with next will differ and the
forecast task defined according to the equality for AT will further conform to the reality.

Table 6 Adjusted tax receipts

Coeff. of   Cumm. Coeff.  Adjusted 

D, %  REAL GDP   TAX  Discretionary changes   change  of changes  Tax, AT  Log GDP 
15,28031 3435,635 374,6132 57,24206 1,180363 3,400975 1274,05 8,166877
0 2187,708 245,3481 0 1 3,400975 834,4227 7,714149
14,1647 3199,164 368,5988 52,21093 1,165022 2,919237 1076,027 8,095347
13,89318 1271,469 149,8301 20,81617 1,161348 2,513662 376,6223 7,169806
15,51719 1293,713 147,4972 22,88742 1,183673 2,123612 313,2268 7,187202
14,25737 2153,925 321,3255 45,81257 1,166281 1,820841 585,0826 7,698538
13,29224 1615,818 271,7988 36,12815 1,153299 1,57881 429,1187 7,410208
0 1930,455 273,9282 0 1 1,57881 432,4807 7,588667
14,59118 24680,45 2935,928 428,3866 1,170839 1,348443 3958,932 10,14472
0 7335,085 757,385 0 1 1,348443 1021,291 8,927666
0 6377,85 697,0173 0 1 1,348443 939,8883 8,7874
14,24966 53243,74 5941,664 846,6669 1,166176 1,156295 6870,315 10,91594
13,51685 24232,44 2798,303 378,2425 1,156295 1 2798,303 10,12635
0 42582,36 5136,138 0 1 1 5136,138 10,69182
0 13027,94 1706,198 0 1 1 1706,198 9,503851

Defuzzified results by using center of gravity method given in Table 6 were obtained based on the calculations
performed on the fuzzy numbers in Matlab environment. The following procedures were used with this purpose:

point_n = 101; % number of points of universal majority

min_x=LS; max_x=RS;[min_x, max_x];
x = linspace (min_x, max_x, point_n)';
A = trimf (x, [a b c]);
B = trimf (x, [a b c]);
C1 = fuzarith (x, A, B, 'sum');
Title ('fuzzy addition A+B')
C2 = fuzarith(x, A, B, 'sub');
340 Akif Musayev et al. / Procedia Computer Science 120 (2017) 333–340
8 Akif Musayev et al./ Procedia Computer Science00 (2018) 000–000

Title ('fuzzy subtraction A-B');

C3 = fuzarith (x, A, B, 'prod');
Title ('fuzzy production AB');
C4 = fuzarith (x, A, B, 'div');
Title ('fuzzy division ');

4. Conclusion

The methodology for estimation of the impact of the changes made in legislative base of tax system on tax
revenues is proposed in the present article. In this research work, the adjustment of series of tax receipts was carried
out by means of fuzzy numbers based on expert assessment and relevant simulations conducted with them. In order
to ensure reliability and truthfulness of the existing information to a maximum extent double expert evaluation
methodology was used. On the first stage, it is supposed that m number of changes is provided for the beginning of
the next financial year in the legislative base of tax system and the impact of each of these changes on tax revenues
is assessed by n number of experts by 100 points scoring system.
On the second stage, the second group of n number of experts estimated the truthfulness of the estimation tax of
the first group of experts between [0,1]. The expression in percentage of the impact of the first changes made in tax
legislation on tax revenues was set and the line of values characterizing the impact of each of the changes made in
tax legislation was set. In the next step, expression in percentage of the changes made in the financial year when the
tax legislation was reviewed was determined and relevant estimation was conducted for the changes made in the tax
financial year and the line of the initial tax revenues was adjusted and substituted with line. The fuzzy regression
dependency on the time series of tax revenues of the GDP, and relevant fuzzy regression dependency on the time
series of adjusted tax receipts was set.

References

Kanghua, Z., Shan, Z., Qian, Z., 2013. The Impact of Economic Growth and Tax Reform on Tax Revenue and Structure: Evidence from China
Experience, Modern Economy, 4, 839-851.
Santiago, A., Jiae, Y., 2012. Tax Composition and Growth: A Broad Gross-Country Perspective, IMF Working Paper, 257, pp. 3.
Jing, X., 2011. Does Tax Structure Affect Economic Growth? Empirical Evidence from OECD Countries, Oxford University Centre for
Business Taxation, Working Paper, 20, pp.1
Timothy, B., Torsten, P., 2014. Why Do Developing Countries Tax So Little?, Journal of Economic Perspectives, 4, pp. 99-120.
Margareta, D., Asa, H., 2012. Taxation of Income & Economic Growth; An Empirical Analysis of 25 Rich OECD Countries, Working Paper,
6, pp.1
Asa, J., Chrispher, H., Jens, A., Bert, B., Laura, V., 2008. Tax & Economic Growth, Economics Department, Working paper, 620, pp. 2.
Desislava, S., Nikolay, P., 2012. An Empirical Evidence for the Impact of Taxation on Economy Growth in European union, Book of
Proceedings-Tourism and Management Studies International Conference Algarve, Portugal, 1031-1039.
Jonathan, H., 1998. Estimating Tax Buoyancy, Elasticity and Stability, Harvard İnstitute for International Development and Suffolk University,
African Economic Policy Paper Discussion Paper, 11, pp. 1.
Aliev R.A., Jafarov S.M., Babayev M.J., Zeynalov E.R., Gardashova L.A. ,2003.The decision making and control under uncertainty. Baku:
"Nargiz", 214 p.
MusayevA., Madatova Sh.,Rustamov S.,2016. Evaluation of the impact of the tax legislation reforms on the tax potential by fuzzy inference
method., ProcediaComputer Science 102 , 507– 514
MusayevA., Madatova Sh.,Rustamov S., 2014.Mamdani-Type Fuzzy inference system for evaluation of Tax potential. 4th World conference on
Soft Computing,Berkeley, California, www.wconsc-2014-Berkeley.com