Analysis and forecasting of crime in Russia as a tool of effective public administration in the social sphere

Parshintseva Lidiya Sergeevna ORCID: 0000-0002-2256-7070 PhD in Economics Associate Professor of Information Law, Computer Science and Mathematics, Russian State University of Justice 117418, Russia, Moscow, Moscow, Novocheremushkinskaya str., 69 lsparshintseva@yandex.ru

Introduction. The full development of society is not possible without the preservation of law and order, the achievement of the rule of law and the preservation of cultural and moral values. Recently, despite the actions taken in relation to the Sustainable Development Goals, there has been a tendency of increasing aggression in the world, a decrease in the general standard of living, and population inequality, hunger and poverty remain at a fairly high level ^[1-4]. All this contributes to an increase in crime and offenses ^[1-5].

Taking into account the role of preserving the rule of law in society, in 2010 the international non-governmental organization World Justice Project developed a methodology for the rule of law index, which allows for cross-country comparisons of the level of ensuring the legal environment. According to this indicator, in 2022, the Russian Federation was on the 107th place in the world out of a possible 140 with a value of 0.45. Note that in 2021, the country ranked 101st with a value of 0.47. The deterioration of the situation according to the index values was due to increased restrictions on the powers of the government (from 129 to 130 places), an increase in the corruption component (from 88 to 90 places), a decrease in the openness of the government (from 78 to 82 places), a deterioration in the observance of fundamental rights (from 114 to 122 places), a decrease in order and security (from 84 to 91 places) and regular law enforcement (from 81 to 92 places), as well as deterioration in the field of civil (from 70 to 74 places) and criminal justice (from 121 to 123 places) ^[6].

The above conditions the analysis of the state and trends in the development of crime in Russia, taking into account the peculiarities of the country in order to reduce the criminogenic situation.

The purpose of the study is to identify the main trends and patterns of crime development in Russia in order to develop effective public administration measures aimed at improving the criminogenic situation.

Material and methods of research. In the course of the research, official statistical data were used, as well as the results of scientific research by specialists in the subject area.

The study was conducted using statistical methods such as the method of relative values, index analysis, time series analysis and forecasting methods, in particular, adaptive methods, correlation and variance analysis methods, as well as methods of visual data analysis.

The results of the study. Over the past 10 years, the number of registered crimes has decreased by 14.6%, which is 335.4 thousand cases. The largest number of registered crimes occurred in 2015 and amounted to 2388.5 thousand cases. There was also a surge in crime in 2019-2020. ^[7,8]

Despite the general trend towards a decrease in crime in the period under review, there is a deterioration in the structure of the number of registered crimes. Thus, the shares of serious and especially serious crimes increased by 5.3% or 1.1 percentage points and 30.7% or 1.5 percentage points, respectively. In general, according to the calculated values of the coefficients of structural shifts of Ryabtsev (by 3.7%) and Gatev (by 5.2%), the structure of crime has not changed significantly ^[7-9].

Figure 1 shows the structure and dynamics of crimes in Russia by category for the period 2012-2022.

Source: built by the author according to ^{[7, 8]}

Fig. 1. Structure and dynamics of crimes by category in Russia

            The lowest crime rate in all categories of crimes during the period under review was observed in the Chechen Republic: in 2022, the crime rate amounted to 159 cases per 100 thousand people of the region's population, the crime rate in the category of serious and especially serious – 44 cases and in the category of medium and minor severity – 115 cases. There was also a small level of crime in the Republics of Dagestan and Ingushetia. The highest crime rate (more than 2,000 cases per 100,000 people) was recorded in the Jewish Autonomous Region (2,130 cases), the Amur Region (2,139 cases), the Komi Republics (2,026 cases), Altai (2,176 cases), Karelia (2,235 cases) ^[7,8].

All subjects within the federal districts in 2022, with the exception of the North-Western (53.0%) and Far Eastern (34.5%), and in Russia as a whole were homogeneous in terms of crime.

As a result of the analysis of variance, a significant influence of the regional factor on the crime rate was established, both in the categories of serious and especially serious crimes (the empirical correlation ratio was 0.975) and in the categories of medium and minor severity (0.928). That is, the variation in crime rates in the categories under consideration was due to the regional factor by 95.1% and 86.1%, respectively ^[7,8].

The study of the dependence of the crime rate, including by categories of serious and especially serious crimes, as well as moderate and minor severity on socio-economic factors showed in all cases the influence of three factors: the coefficient of migration growth, the unemployment rate and the retail sale of alcoholic beverages per capita ^{[7,8, 10]}. At the same time, during the period under review, the closeness of communication remained approximately at the same level. In addition, the influence of the concentration of income of the population (Gini index) on the level of crime in the category of medium and minor severity was revealed. The conclusions made are consistent with the results of studies by other scientists analyzing the influence of individual socio-economic factors on crime ^[11-16].

Table 1

Matrix of paired correlation coefficients

Note: Y1 is the crime rate, Y2 is the crime rate by category of serious and especially serious crimes, Y3 is the crime rate by category of moderate and minor severity, X1 is the unemployment rate, X2 is the retail sale of alcoholic beverages per capita, X3 is the migration growth rate, X4 is the Gini coefficient, X5 is the differentiation coefficient. by income

Source: calculated by the author according to ^{[7,8, 10]}

Thus, it can be concluded that the variation in the crime rate, including by crime categories, in addition to the regional factor, is influenced by the standard of living of the population, the migration situation and the situation on the labor market ^[17-19]. Moreover, it should be noted that there is an inverse relationship between the crime rate and the migration increase, that is, the migration outflow, prompted, as a rule, by unfavorable living conditions, shows an inverse relationship with the crime rate.

In order to develop preventive measures to reduce crime, it is important to predict its volume, including by categories of crimes ^[5]. Based on monthly data for the period from 2012 to April 2023, modeling and forecasting of the number of registered crimes, including by category of crimes, was performed. Adaptive methods were used to model the analyzed time series, which have a number of significant advantages over other known modeling and forecasting methods: high accuracy of forecasts, consideration of the time value of information and the degree of "obsolescence" of data, and other advantages.

Autocorrelation functions of the initial series (Fig. 2 a, d), the first differences (Fig. 2 b, e) and seasonal differences (Fig. 2 c, e) were used to identify components in the time series of the number of registered crimes by categories of serious and especially serious crimes, medium and minor gravity

a) b)                                                 c)

d) e)                                                 e)

Source: built by the author according to ^[3,6,7]

Fig. 2. Autocorrelation functions of the initial series (Fig. 2 a, d), the first differences (Fig. 2 b, e) and seasonal differences (Fig. 2 c, e)

            As can be seen from Figure 2 (a, b, c), seasonal and systematic components are present in the time series of the number of registered crimes by categories of grave and especially grave crimes, since the autocorrelation function gradually decreases and has outliers at lags that are multiples of the seasonality period. The behavior of the autocorrelation functions of the first and seasonal differences more clearly demonstrate, respectively, seasonality and trend in the series. Similar conclusions can be drawn on the time series of the number of registered crimes in the categories of medium and minor severity (see Fig. 2 d, d, e).

            Figure 3 shows the seasonal wave of the analyzed time series. Since the amplitude of fluctuations in the number of registered crimes, both in the categories of serious and especially serious crimes, and in the categories of moderate and minor severity, remained constant during the period under review, seasonality has an additive form.

Source: built by the author according to ^[3,6,7]

Fig. 3. Seasonal wave of the number of registered crimes

            The peak of crime growth in all categories of crimes is observed from year to year in March and October, in addition, for crimes of medium and minor severity – in June. The decline in crime for each category of crimes took place in the winter months (December, January).

            To select the best model describing the behavior of time series levels for each category, the errors of models with additive seasonality and various types of trend (linear, exponential, damped) were calculated.

            The errors of models with additive seasonality and different types of trend are presented in Table 2.

            Table 2

Results of comparative analysis of model errors

Source: calculated by the author according to ^[3,6,7]

            As a result of a comparative analysis of model errors, it was found that the best behavior of the levels of the time series of the number of registered crimes by categories of serious and especially serious crimes describes the exponential smoothing model with a damped trend and an additive form of seasonality at ?=0.2, ?=0.1, ?=0.1, for the time series of the number of registered crimes by categories of medium and a model of exponential smoothing with a damped trend and an additive form of seasonality with the parameters ?=0.3, ?=0.1, ?= 0.3 was of low severity. The selected models were overestimated by 0.7% and 0.5%, respectively, which are acceptable values for obtaining high–quality forecasts for the calculated models. The autocorrelation functions of the model residuals also indicate the adequacy of the constructed models.

The point and interval forecast of the number of registered crimes according to the selected models until the end of 2023 is presented in Table 3.

            Table 3

Forecast of the number of registered crimes for 2023

Source: calculated by the author according to ^[3,6,7]

Based on the above calculations, we can expect an increase in the number of registered crimes in June, August and October for all categories of crimes, respectively, by 5.4% (8817 cases), 2.0% (3829 cases) and 14.1% (22452 cases) and a decrease in the indicator by the end of the year compared to the beginning of 2023 by 4.0%. It should also be noted that compared to the same period of the previous year, an increase in the number of registered crimes is projected by about 1.1%.

Conclusion. The results of the study showed that the variation in crime rates by crime categories is due to the influence of the territorial factor and a significant number of socio-economic factors, among which factors of regional differentiation and concentration of income of the population, the level of poverty ^[14,15], differences in the labor market situation ^[12], the standard of living of the population and other factors are important. In order to reduce the criminogenic situation in various subjects of the Russian Federation, first of all, targeted measures should be aimed at reducing the socio-economic inequality of the population and solving certain issues of a deprivation nature ^{[2, 20, 21]}.

The performed forecast shows that at present, despite the positive trend towards a decrease in the number of registered crimes, there has been a fairly high level of tension in society and a high differentiation of the population on various grounds.

For the Russian Federation, the issues of territorial differentiation are particularly acute and require immediate solutions, since any manifestations of unevenness in the distribution of resources and opportunities lead to a deterioration of social relations and the economic situation in the country, prompting an increase in crime, which, in turn, affects the level of socio-economic life, the pace of development and security of the country.

            The results of the conducted research can be useful to executive authorities when developing targeted measures to reduce crime, taking into account the identified factors and forecasts.