Crime and birth rate in Ukraine: modelling and statistical analysis

Abstract

The article is devoted to empirical testing of the hypothesis that there is a correlation between the birth rate and the overall level of crime in Ukraine. A detailed analysis of the relevant correlation field confirms the hypothesis, but such a relationship, as shown by the calculation of determination coefficients, depends on the level of lag between the indicators and reaches its maximum value at a time lag of two years between them. The study has also identified an atypical area of direct linear relationship between birth rate and crime (2018–2021), which does not correspond to the general reverse trend. The combined two-period birth rate model built as a result of the correlation and regression analysis proved to be adequate, with a fairly high interpolation accuracy (R = 0.865; MAPE = 5.38 %). In addition, for the first period (1992–2018), an overall decrease in the crime rate by 1 % was accompanied by an increase in the birth rate by 0.71 % with a time shift of two years. In the second period (2018–2021), on the contrary, there was a simultaneous decrease in both indicators. Such atypical behaviour (compared to the previously identified general pattern) is explained by the systemic influence of a number of factors, among which the key ones are the hybrid military actions of the Russian Federation on the territory of Ukraine and the corresponding significant increase in migration processes due to the introduction of a visa-free regime with the EU countries. An attempt has been made to combine separate time periods with one regression equation by introducing additional indicator variables. The generalised model of the change in the birth rate depending on the level of criminality of a country obtained as a result of the proposed approach proved to be adequate, and all its parameters were statistically significant. In particular, according to the estimates made on the basis of the model, it was found that with an overall reduction in the level of criminality of the country by 10 thousand crimes in two years, the birth rate is expected to increase by 12.73 persons per 100 thousand population, provided that other indicators of the model remain unchanged. The confidence interval of this value is (9.61; 15.84) for a five per cent level of significance. The proposed approach made it possible to conduct an approximate assessment of the effect of the increase in the birth rate in 2012–2013 due to the increase in financial payments at childbirth.