ANALYSIS OF THE DYNAMICS OF INCOME AND COSTS OF THE POPULATION IN THE REPUBLIC OF AZERBAIJAN Текст научной статьи по специальности «Строительство и архитектура»

ECONOMIC SCIENCES

ANALYSIS OF THE DYNAMICS OF INCOME AND COSTS OF THE POPULATION IN THE

REPUBLIC OF AZERBAIJAN

Ayyubova N.,

Ph.D in Economics. Associate Professor of "Mathematical Economy" department of Baku State University, Republic of Azerbaijan https://orcid.org/0000-0003-3225-389X Mammadli A.

Master student of "Mathematical Economy" department of Baku State University, Republic of Azerbaijan

DOI: 10.5281/zenodo.7408529

ABSTRACT

In the presented work, a statistical analysis of the distribution of income indicators and costs of the population was carried out, the distribution of the population's income and the degree of stratification of the society were investigated based on the Gini and Lorenz indices. An extensive empirical analysis of the time series according to the income and costs indicators of the population was carried out, descriptive statistics were determined, the normal distribution of the series based on the Jarque-Bera test, stationarity was investigated based on the autocorrelation function and the Dickey-Fuller test, White, Cusum tests were checked. AR models with first and second order differences were established, it was determined that the AR(1) model for income is not stationary, the AR(2) model satisfies the stationarity conditions, and the AR(1) time series for costs is stationary.

The conducted research can be evaluated as the basis of studies on extensive empirical analysis, modeling and forecasting of the income and cost of the population in the Republic of Azerbaijan and create wide opportunities for prospective studies. The statistical information used in the study was obtained from the official website of the State Statistical Institute and covers the years 1995 -2021. The obtained data were processed in Excel and Eviews application software packages.

Keywords: empirical analysis of income and cost indicators of the population, Jarque-Bera test, AR models, Cusum test, autocorrelation, stationarity.

Introduction. In modern periods, the statistical study of the income and costs of the population across countries is of great importance. Accordingly, to analyze the standard of living of the population, to develop a socio-economic policy and, most importantly, to organize the social protection of individual population groups, it is necessary to collect and statistical analyze objective data about income. Income is an important economic indicator reflecting social development. If the distribution of income is fair, the social welfare in the country will increase, the poverty level will decrease and there will be optimistic expectations about the future. Currently, the change in the economic situation in a number of countries has a significant impact on the living standards of the population and its separate strata, as well as the level and structure of their income and costs to one degree or another. The population's standard of living has seriously decreased, the number of unemployed and those living in poverty has increased, the process of stratification has intensified, such cases increase the importance of the statistical study of income and expenses of the population based on the MHS concept, justifies the relevance of the topic of the research.

It is important to collect objective information about the income and costs of the population, to analyze the general state of the economy and the standard of living of the population, to develop the social policy of the state and to implement concrete measures to organize. Systematized information on the income of the

population can be used to assess the possibilities of expanding investment processes through the mobilization of internal resources. Despite the statistical analysis, modeling of macroeconomic indicators [5,14] that shape and determine the standard of living of the population in domestic and foreign literature [3,4,7,9,16,17], as well as work designed to investigate and solve other problems, the relevance of the issue considered, the application of these economic for research studies is important [1,2,6,12,13].

A number of scientific articles are devoted to these problems, taking into account regional characteristics in the process of transformation of national economies, and are devoted to the analysis of integration processes between individual countries and groups of countries of the post-Soviet space.

The main part of research. The statistical data required to conduct an econometric analysis of the uneven distribution of income and cost of the population in the Republic of Azerbaijan were obtained from the official website[15] of the State Statistical Committee of the Republic of Azerbaijan and presented in Table 1. The statistical information in Table 1 indicates In-come(Y1) - the income of the population, Absolute rate of change (Y2) - the rate of absolute change of income, Costs (Y3) - the cost of the population. The absolute rate of change of income was determined by calculating the ratio of the difference between the current level and the previous level to the previous level according to the income in the Excel software package.

Table 1.

Income and cost indicators of the population for the years 1995-2021 (in current prices, ^ in million manats)

Years Income(Y1) Absolute rate of change (Y2) Costs (Y3)

1995 1,340.5 - 1,275.8

1996 1,905.1 0.42 1,853.1

1997 2,473.4 0.30 2,411.3

1998 2,884.8 0.17 2,932.6

1999 3,687.7 0.28 3031.4

2000 4,047.3 0.10 3,272.2

2001 4,301.6 0.06 3,498.4

2002 5,018.6 0.17 4,171.2

2003 5,738.1 0.14 4,793.8

2004 6,595.1 0.15 5,549.9

2005 8,063.6 0.22 6,508.7

2006 10,198.5 0.26 8,208.1

2007 14,558.2 0.43 11,249.7

2008 20,735.4 0.42 15,891.9

2009 22,601.1 0.09 17,417.6

2010 25,607.0 0.13 19,251.5

2011 30,524.6 0.19 22,184.0

2012 34,769.5 0.14 24,564.0

2013 37,562.0 0.08 28,021.2

2014 39,472.2 0.05 30,799.6

2015 41,744.8 0.06 34,963.4

2016 45,395.1 0.09 39,775.0

2017 49,187.9 0.08 44,498.4

2018 53,103.7 0.08 47,557.2

2019 56,769.0 0.07 51,927.4

2020 55,754.1 -0.02 49,744.0

2021 57,181.5 0.03 55201.5

Source: Prepared by the authors based on the data obtained from the State Statistical Committee.

The results of descriptive statistics are important in time series analysis.

Table 2.

Results of descriptive statistics on incomes and costs of the population, absolute rate of change of income

Income (Y1) Absolute rate of change(Y2) Costs (Y3)

(23748.90) (0.161280) (20020.48)

Median 20735.40 0.136031 15891.90

Maximum 57181.50 0.427484 55201.50

Minimum 1340.500 -0.017878 1275.800

Std.Dev 20260.34 0.123748 18141.79

Skewness 0.395810 0.970919 0.644498

Kurtosis 1.619909 3.022570 1.981182

Jarque-Bera statistic 2.847730 4.085511 3.036940

Probability 0.240782 0.129671 0.219047

Sum 641220.4 4.193279 540552.9

Sum Sq.Dev 1.07E+10 0.382837 8.56E+09

Observations 27 26 27

In Table 2 represents average indicators, mean square deviations, excess, asymmetry and other characteristics of time series on income, absolute rate of change of income and costs. Satisfactory results were obtained for all three series for both excess and asymmetry. In this way, for these mentioned characteristics, the calculated results in all cases are small and very close to 0, they satisfy the conditions for the significance of asymmetry and excess. The distribution of the time series follows a normal distribution in all three cases. The results of the Jarque-Bera test also confirm this. JBy1=2.847730, prob.=0.240>0.05; since

JBy2=4.085511,prob. =0.12>0.05 and JBys=3. 036940, prob.=0.21>0.05, the hypotheses of normal distribution are accepted. Note that the average values for Y1, Y2 and Y3 are listed in parentheses in Table 2.

Results according to the autocorrelation function (ACF) and specific autocorrelation function (PACF), their graphs and the results of the ADF test with both primary and first and second order differences can characterize the stationarity of time series [8,10,11,14]. In the next step of the research, the autocorrelation functions for the incomes and costs of the population were investigated.

The presence of autocorrelation among the residuals in the autoregression model indicates that there is correlation between the levels of the time series. This dependence causes cyclical fluctuations in the levels of the series, which leads to the low quality and inefficiency of the forecasts formed on the basis of the autoregression model, because deviations of a cyclical nature, in general, are not random and can create a trend. The analysis of dynamic results and the Sample: 19952021 Included! observations: 26

Autocorrelation Partial Correlation AC PAC Q-Stat Prob

0.762 0.762 16.906 0.000

1 1 1 1 1

1 1 1 1 2 0.533 0.005 27.202 0.000

1 1 1 □ 1 3 0.513 0.160 35.520 0.000

1 11 1 Z 1 4 0.360 -0.201 39.012 0.000

1 =1 1 1 1 5 0.237 -0.010 41.764 0.000

1 : 1 1 E 1 6 0.103 -0.100 42.151 0.000

1 [ 1 1 I 1 7 -0.026 -0.064 42.177 0.000

1 1 1 Zl 1 S -0.006 0.216 42.179 0.000

1 1 1 1 9 -0.016 -0.023 42.190 0.000

1 [ 1 1 [ 1 10 -0.005 -0.069 42.510 0.000

1 [ 1 1 ] 1 11 -0.046 0.090 42.610 0.000

1 [ 1 1 IZ 1 12 -0.006 -0.217 43.005 0.000

constructed autocorrelation and special autocorrelation functions of the considered time series can form an opinion whether the series Y1 and Y3 are stationary or non-stationary according to the initial data.

Thus, in the autocorrelation analysis conducted on population incomes, probabilities less than 0.05 for all levels determine that the series is non-stationary, and the H0 hypothesis is rejected (see Picturel).

Picture 1. ACF and PACF for the order Y1 according to the income of the population.

In picture 2 presents the results of the autocorrelation analysis on population costs. In this case, the probabilities are equal to 0.00 for all levels of the series, and the hypothesis Hi is accepted as an alternative hypothesis to Ho about the non-stationarity of the considered series.

Sample: 1995 2021 Included observations: 27

Autocorrelation Partial Correlation AC PAC Q-Stat Prob

0.766 0.766 17.662 0.000

i I i I 1

i I i □ I 2 0.650 0.154 30.905 0.000

i I I 3 0.402 -0.339 36.190 0.000

i Zl I i C I 4 0.212 -0.148 37.724 0.000

i I i C I 5 -0.010 -0.155 37.727 0.000

i C I i C I 6 -0.179 -0.114 30.924 0.000

i|= I i : I 7 -0.329 -0.099 43.157 0.000

I I i I S -0.393 0.009 49.529 0.000

l I i I 9 -0.42S -0.024 57.400 0.000

I I i ] I 10 -0.371 0.065 63.809 0.000

I i I 11 -0.310 -0.010 60.509 0.000

I IZ I I IZ I 12 -0.269 -020B 72.296 0.000

Picture 2. ACF and PACF for the Y3 row according to population cost.

According to the results in Picture-1 and Picture-2, we can conclude that ACF decreases for series Y1 and Y3, and PACF has the highest autocorrelation coefficient for series Y1 for first order and for series Y3 for first and third order. Functions for other levels do not have significant autocorrelation coefficients.

In the conducted research, the capabilities of the Dickey-Fuller test were used to eliminate the non-stationarity of the studied time series based on the

primary data. Let's examine the ADF test results using the Eviews application software package. According to the test, the hypothesis that the time series has a single root is accepted if the probability of the t-statistic is less than 5% (significance level of 0.05). For the time series to be stationary, the value of the Dickey-Fuller test should be smaller than the critical value at the i%, 5%, 10% significance levels. The test results are presented in Table 3:

Table 3.

Dickey-Fuller test results

Variable T-statistic Critical values: 1% Critical values: 5% Critical values: 10% Prob.

First difference, trend and constant

Income (Y1) Absolute rate of change (Y2) Costs (Y3) -2.666340 -4.374307 -3.603202 -3.238054 0.2573

-5.152159 -5.217134 -4.416345 -4.374307 -3.622033 -3.603202 -3.248592 -3.238054 0.0021 0.0015

Second difference, trend and constant

Income (Y1) Absolute rate of change (Y2) Costs (Y3) -5.170003 -4.416345 -3.622033 -3.248592 0.0020

-7.392002 -9.236511 -4.440739 -4.394309 -3.632896 -3.612199 -3.254671 -3.243079 0.0000 0.0000

In Table 3, because the probability level for Income (Y1) is greater than 0.05, the time series is trending and stationary with the first difference, and the absolute rate of change of income (Y2) is less than 0.05, and the t-statistic value is 1% , 5%, is smaller than the significance level value of 10%, so the ^-hypothesis is rejected and the time series is stationary in the case of trend constant with the first difference. When we look at cost (Y3), since the probability level is less than 0.05, and also the value of t-statistic is smaller than the value of 1%, 5%, 10% significance level, the time series is assumed to be stationary in the case of trend constant with first order difference. The probability level of Income (Y1) with the second design difference is less than 0.05 and the value of the t-statistic is smaller than the value of the significance level of 1%, 5%, 10%, so the ^-hypothesis is rejected and the trend with the second design differences is stationary. Since the absolute rate of change of income (Y2) is smaller than

the 5% probability level and the 1%, 5%, 10% significance level, the second formulation is considered stationary in the case of a trend constant with the difference, and also the cost (Y3) is less than the 5% probability level and the 1%, 5%, 1%, 5%, since it is smaller than the 10% significance level, the second formulation is stationary with the trend constant.

White's test was used to check the heteroskedasticity and the result is given in Table-4 for income, Table-5 for absolute rate of change of income, and Table-6 for cost. nR2=Obz*R2, the number of observations was taken n=26 in both cases. It received value R2=3.506367 for income, R2=1.264556 for absolute rate of change of income, and R2=4.729267 for cost. As the probability level for all three values is greater than 0.05, heteroscedasticity is not detected in the model, and the hypothesis H0 about homoscedasticity is accepted.

Table 4.

White test results (Income Y1)

F-statistic 1.792650 Prob (2,23) 0.1890

Obs*R- squared 3.506367 Prob. Chi-Square (2) 0.1732

Table 5.

White test results (Absolute rate of change Y2)__

F-statistic 0.587917 Prob(2,23) 0.5636

Obs*R- squared 1.264556 Prob. Chi-Square (2) 0.5314

Table 6.

White test results (Costs Y3)__

F-statistic 2.548241 Prob(2,24) 0.0992

Obs*R- squared 4.729267 Prob. Chi-Square(2) 0.0940

It is used to evaluate the stability of the parameters of the model. These tests are based on the calculation of the cumulative sum of the recursive residuals and the cumulative sum of the squares of the recursive residuals and the evaluation of the corresponding equations. Test results are analyzed according to 95% confidence intervals. If the recursive estimates of the residuals deviate from the critical limits, then this indicates instability of the model parameters. Graphically, if the

blue line is located between the red lines and does not intersect with them, it confirms the H0 hypothesis about the stability of the parameters, otherwise, if the blue line intersects with the red lines, then the H1 hypothesis about the instability of the parameters relative to the length of the time interval is accepted. The results of the CUSUM test are shown in Picture-3 for income, Picture-4 for the absolute rate of change of income, and Picture-5 for cost.

98 00 02 04 06 08 10 12 14 16 18 20

96 98 00 02 04 06 08 10 12 14 16 18 20

CUSUM -----5% Significance I |_Standardized Residuals |

Picture 3. Cusum test and standardized residuals (income Y1)

98 00 02 04 06 08 10 12 14 16 18 20

CUSUM -----5% Significance

96 98 00 02 04 06 08 10 12 14 16 18 20

Standardized Residuals

Picture 4. Cusum test and standardized residuals (Absolute rate of change Y2)

98 00 02 04 06 08 10 12 14 16 18 20

96 98 00 02 04 06 08 10 12 14 16 18 20

I CUSUM 5% Significance | | Standardized Residuais~|

Picture 5. Cusum test and standardized residuals (costs Y3)

According to the results of the CUSUM test for the income of the population (Y1), the instability of its parameter is observed because it does not meet the required conditions. As the level indicators of the results of the CUSUM test on the absolute rate of change of income (Y2) are close to each other, they do not change and give the impression of stable dynamics on a straight line, and since the blue line is located between the red lines, it is assumed that the Y2-parameter is stable or steady. Also presented in Picture

4 is a representation of the standardized residuals, and in this graphical representation, the recursive values of residuals (CUSUM) and the recursive values of squares of residuals (CUSUM of Squares) do not deviate from the 95% confidence interval. According to the results of the CUSUM test on cost (Y3) in figure 5, the cost parameters can be considered stable and steady.

Let's consider autoregression models with first and second order differences to characterize the dependencies between economic indicators.

Table 7.

Autoregression model with first-order differences for population income

Summary Output

Regression Statistics

Multiple R 0.997215

R Square 0.994437

Adjusted R Square 0.993603

Standard Error 1587.557

Observations 24

ANOVA

df SS MS F SignificanceF

Regression 3 9.01E+09 3E+09 1191.807 1.05E-22

Residual 20 50406745 2520337

Total 23 9.06E+09

Coefficients Stan.Error t Statistic P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%

Intercept 1066.53 598.5724 1.7817895 0.089975 -182.07 2315.13 -182.07 2315.13

Lag1 1.570632 0.22444 6.998013 8.63E-07 1.10245 2.03880 1.1024 2.03880

Lag2 -0.670945 0.444032 -1.51103 0.14642 -1.5971 0.25528 -1.597 0.25528

Lag3 0.108678 0.299961 0.362305 0.720924 -0.5170 0.73438 -0.517 0.73438

Based on the results presented in Table 7, the formal autoregression model of the population income dynamics with first order differences is as follows: Y1 (t)=1066.53+1.5 70632 *Y1 (t-1) For the studied process to be stationary, the roots of the corresponding characteristic equation must be

outside the unit circle. The characteristic equation for the AR(1) model is: 1-1.5706z = 0. The root of the equation is equal to z ~ 0.637, that is,

I z I < 1. For the AR(1) model, the process is non-stationary. This is confirmed by the results of our previous research for Y1.

Table 8.

Autoregression model of population income with second-order differences

Summary Output

Regression Statistics

Multiple R 0.997196

R Square 0.994401

Adjusted R Square 0.993868

Standard Error 1554.373

Observations 24

ANOVA

df SS MS F SignificanceF

Regression 2 9.01E+09 4.51E+09 1864.787 2.27E-24

Residual 21 50737577 2416075

Total 23 9.06E+09

Coefficients Stan.Error t Statistic P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%

Intercept 1022.096 573.6271 1.781812 0.089975 -170.82 2215.01 -170.82 2215.01

Lag1 1.530951 0.191807 7.981716 8.54E-08 1.132065 1.92983 1.1320 1.92983

Lag2 -0.52794 0.199139 -2.65112 0.01494 -0.94207 -0.1138 -0.942 -0.1138

Based on the results in Table 8, the autoregression model of the population income dynamics with second-order differences is described by the following equation:

Y1 (t)=1022.096+1.530951 Y1 (t-1)-0.52 794Y1 (t-2)

Characteristic equation for the AR(2) model 1-1.53z-0.527z2 =0. The roots of the equation are Z1 ~ 2.34, Z2 ~ -5.24. From this we can conclude that the AR(2) model is stationary with second order differences.

Table 9.

Autoregression model of population costs with first-order differences

Summary Output

Regression Statistics

Multiple R 0.99597

R Square 0.991957

Adjusted R Square 0.990687

Standard Error 1727.656

Observations 23

ANOVA

df SS MS F Significance F

Regression 3 6.99E+09 2.33E+09 781.0782 4.55E-20

Residual 19 56711135 2984796

Total 22 7.05E+09

Coefficients Stan.Error t Statistic P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%

Intercept 926.7874 640.409022 1.44718 0.164145 -413.604 2267.179 -413.604 2267.179

Lag1 0.899451 0.23039143 3.904013 0.000954 0.417236 1.381666 0.417236 1.381666

Lag2 0.867129 0.5375847 1.613009 0.123229 -0.25805 1.992307 -0.25805 1.992307

Lag3 -0.77357 0.44574376 -1.73547 0.098849 -1.70653 0.15938 -1.70653 0.15938

Based on the results of Table 9, the autoregression model with differences of order 1 for population expenditure is described as follows:

Y3(t)=926.7874+0.899451 *Y3(t-1) The characteristic equation for the cost AR(1) model is: 1-0.8994z=0. The root of the equation: z ~ 1.118, that is I z I >1, so the process is stationary with first-order differences.

The most popular methods for analyzing the dynamics of population income are the application of the Lorenz curve and the Gini coefficient. The Gini coefficient takes a value between 0 and 1, if the result is close to 1, the inequality in the distribution of income is high, and when the index approaches 0, it shows that there is an equal distribution in the distribution of income.

Table 10.

Results according to the Gini coefficient (1995-2021)

Individual Income Cumulative Cumulative Area Under Lorenz Curve

%of population %of incom %of income

1 1340.5 3.7037037 0.20905448 0.20905448 0.003871 0.00387138

2 1905.1 7.4074074 0.29710533 0.50615982 0.013244 0.01324471

3 2473.4 11.111111 0.3857332 0.89189302 0.025889 0.02588987

4 2884.8 14.8148148 0.44989211 1.34178513 0.04136441 0.04136441

5 3687.7 18.5185185 0.57510647 1.9168916 0.06034587 0.06034587

6 4047.3 22.2222222 0.63118703 2.54807863 0.08268463 0.08268463

7 4301.6 25.9259259 0.67084578 3.21892441 0.10679635 0.10679635

8 5018.6 29.6296296 0.78266381 4.00158822 0.1337132 0.1337132

9 5738.1 33.3333333 0.89487172 4.89645994 0.16477867 0.16477867

10 6595.1 37.037037 1.02852311 5.92498305 0.20039709 0.20039709

11 8063.6 40.7407407 1.25753953 7.18252258 0.24273159 0.24273159

12 10198.5 44.4444444 1.59048277 8.77300535 0.29547274 0.29547274

13 14558.2 48.1481481 2.2703894 11.0433948 0.36697037 0.36697037

14 20735.4 51.8518519 3.23373991 14.2771347 0.46889869 0.46889869

15 22601.1 55.5555556 3.52470071 17.8018354 0.594055 0.594055

16 25607 59.2592593 3.99347869 21.7953141 0.73328055 0.73328055

17 30524.6 62.962963 4.76039128 26.5557053 0.89538925 0.89538925

18 34769.5 66.6666667 5.42239455 31.9780999 1.08395936 1.08395936

19 37562 70.3703704 5.85789223 37.8359921 1.29285356 1.29285356

20 39472.2 74.0740741 6.15579292 43.991785 1.51532921 1.51532921

21 41744.8 77.7777778 6.51021084 50.5019959 1.74988483 1.74988483

22 45395.1 81.4814815 7.07948468 57.5814806 2.00154586 2.00154586

23 49187.9 85.1851852 7.67098177 65.2524623 2.27470265 2.27470265

24 53103.7 88.8888889 8.28166103 73.5341234 2.57012196 2.57012196

25 55754.1 92.5925926 8.69499785 82.2291212 2.88450453 2.88450453

26 56769 96.2962963 8.85327416 91.0823954 3.20947253 3.20947253

27 57181.5 100 8.91760462 100 3.53856288 3.53856288

Sum 641220.4 Sum 26.5508217

Area A 23.4491783 „ Gini index 0.46898357 umi/ui/ui/

Source: Calculated in the Excel software package based on the data obtainedfrom the State Statistical Committee of the Republic of Azerbaijan

According to the calculations made in the Excel software package, the Gini index is equal to 0.46898, and the population income distribution and the degree of social stratification based on the Lorenz indices are graphically presented in Picture 6.

100 80 60 40 20 0

a>

s

O o Ö • ^H

0

1)

>

' I

ra

3

1

o

jr / / t

X X A / / ✓ / /

/ / /

''' ^ N &

_ __ ^ — " v° \ B

20 40 60 80

Cumulative % of population

100

Picture 6. Lorenz curve

The dependence between the incomes of the population and the accumulated frequencies for the population is presented in Picture 6. The Lorenz curve(dotted curve) is slightly away from the line of equal distribution of incomes(straight line), showing that there is no high level of concentration in the distribution of incomes, and that a trend is followed at an average level. This shows the partial adequacy of the ratio between the income received by the population and the share of income by population groups. Thus, 26.5% of the population of the region receives an income below the subsistence minimum. Also, considering that the Gini coefficient in Azerbaijan is 0.46 for the years 1995-2021, we can say that the income distribution is at an average level.

Main results:

1. Based on the information obtained from the State Statistical Committee, a statistical analysis of the distribution of income and expenses of the population in the Republic of Azerbaijan for the years 1995-2021 was carried out.

2. Empirical analysis of the time series according to the income and cost indicators of the population was carried out, descriptive statistics, ACF and PACF were determined, Jarque-Bera, Dickey-Fuller, White, Cusum tests were checked. Normal distribution in time series Y1, Y2, Y3 according to Jarque-Bera test, non-stationarity of income and cost time series according to primary data according to ACF and PACF, second-order differences in income according to ADF test, first-order differences of absolute rate of change in income and expenses stationarity was determined in the case of trend constant with first and second order differences. The Cusum test demonstrated the stability of Y2 and Y3 parameters.

3. Autoregression models with first and second order differences were formed for the incomes and costs of the population, the AR model for incomes with

second order differences, and for expenses with first order differences fulfilled the stationarity conditions and demonstrated adequacy.

4. The degree of social stratification of the society was characterized by the Gini index and the Lorenz equilibrium curve. The Gini coefficient was calculated and the results were depicted in a graph, and based on the available data, an absolute equilibrium curve was constructed, which visually shows the degree of stratification of the society, using the Lorenz curve with the appropriate approximation function, and the formation of a trend at the average level of income distribution was determined.

References

1. Ayyubova N.S. Econometric analysis and modeling of the dynamics of the balance of payments' development in Azerbaijan // Statistics and Economics. 2022; 19(2):14-22. https://doi.org/10.21686/2500-3925-2022-2-14-22

2. Ayyubova N.S. On the Measurement of Coin-tegration Relations Between the Indicators of the Time Series of the Current Account of the Balance of Payments and GDP (On the Example of the Republic of Azerbaijan). Voprosy statistiki. 2022;29(5):35-45. (In Russ.). https://doi.org/10.34023/2313-6383-2022-29-5-35-45

3. Ayyubova Natavan Soltan, & Mammadli Ay-taj Vidadi. (2022). Development issues of the autoregression model for the monetary income of the population in the Republic of Azerbaijan. Scientific Collection «InterConf+», (26(129), 59-68. https://doi.org/10.51582/interconf.19-20.10.2022.006

4. Ayyubova, N., & Mammadli, A. (2022). Statistical analysis of inequality in the distribution of income and expenses of the population in the Republic of Azerbaijan. Scientific Collection «InterConf+»,

(27(133), 86-100. https://doi.org/10.51582/inter-conf.19-20.11.2022.007

5. Айюбова Н.С., Исазаде А.Э. Вопросы моделирования и анализа взаимосвязей валютного курса маната и детерминантов национальной экономики Азербайджана // Евразийский Союз Ученых (ЕСУ), Ежемесячный научный журнал N° 1 (82) / 2021 Том 6, Серия: Экономические науки. ISSN 2411-6467. https://doi.org/10.31618/ESU.2413-9335.2021.6.82 стр.4-11

6. Ayyubova N.S., Gasimova M.Q. Empirical analysis of dependencies between the current account of the balance of payments in Azerbaijan and some macroeconomic indicators 3rd ISPC «Concepts for the Development of Society's Scientific Potential» (November 19-20, 2022; Prague, Czech Republic). // №27(133) 17 pp.69-85

7. Cheremisina N.V., Veiss E.V. Assessment of the level of living of the population in Tambov region. Statistics and Economics. 2019;16(1):76-84. (In Russ.) https://doi.org/10.21686/2500-3925-2019-1-76-84

8. Dickey, D.A., Fuller W.A, Distribution of Estimators for Autoregressive Time Series with a Unit Root // Journal of the American Statistical Association, 74, 427-431, 1979 https://doi.org/10.2307/2286348; https://www.jstor.org/stable/2286348

9. Кабашова Е.В. Методология исследования уровня жизни населения в статистической науке // International scientific review. 2016. №20 (30). URL: https://cyberleninka.ru/article/n/metodologiya-issledovaniya-urovnya-zhizni-naseleniya-v-statisticheskoy-nauke

10. Kontorovich G.G. Time series analysis // Economic Journal of the Higher School of Economics, Publishing House of the Higher School of Economics. v.7, 2003 No. 1, pp. 79-103 https://ej.hse.ru/en/2003-7-1/26547295.html

11. Кравцов М.К., Пашкевич А.В., Бурдыко Н.М., Гаспадарец О.И. Система эконометрических моделей для анализа и краткосрочного прогнозирования основных макроэкономических показателей

Республики Беларусь // Экономика и управление. 2007. №3. С.69-80

12. Orudzhev E. G., Ayyubova N.S. Empirical analysis of the factors affecting the balance of payments in Azerbaijan //Актуальш проблеми економши. - 2016.-Т. 181. - №. 7. - c.400-410. https://www.proquest.com/scholarly-journals/empiri-cal-analysis-factors-affecting-bal-ance/docview/1812274952/se-2

13. Orudzhev E.G., Ayyubova N.S., Empirical analysis of balance of payments dynamics // Caspian Journal of Applied Mathematics, Ecology and Economics V. 8, No 1, 2020, July, pp.3-9

14. Orudzhev E.G., Ayyubova N.S. Some systematic and statistical aspects of the decision of the Central Bank of Azerbaijan to devalue the manat (dated February 21, 2015) // Caspian Journal of Applied Mathematics, Ecology and Economics; V. 3, No 1, 2015, July, pp. 59-67.

15. Макроэкономические показатели // Официальный сайт Государственного Комитета Статистики Азербайджанской Республики [Электронный ресурс] https://www.stat.gov.az/; URL: https://www.cbar.az/page-41/macroeconomic-indica-tors

16. Orudzhev E., isazadeh A. Comparative analysis of relationship of the course of the Azerbaijani manat and its major macroeconomic determinants // Actual Problems of Economics. Scientific Economic Journal, №3(201), 2018, pp.94-104 https://www.proquest.com/scholarly-jour-nals/сравнительный-анализ-взаимосвязей-курса/docview/2405311456/se-2

17. Orudzhev E., isazadeh A. Currency course of the manat and its impact on selected macroeconomic indicators of the national economy of Azerbaijan // Actual Problems of Economics. Scientific Economic Journal, №12(198), 2017, pp.54-62 https://www.proquest.com/scholarly-journals/валютный-курс-маната-и-еуо-влияние-на-отдельные/docview/2405311764/se-2