Open AccessRESEARCH AND MODELING OF THE UNEMPLOYMENT IN UKRAINE: CORRELATION-REGRESSION ANALYSIS
Author(s) -
Yana Dovhenko,
Zoia Khaletska,
Lyudmila Yaremenko
Publication year - 2021
Publication title -
vìsnik odesʹkogo nacìonalʹnogo unìversitetu. ekonomìka
Language(s) - English
Resource type - Journals
eISSN - 2664-696X
pISSN - 2304-0920
DOI - 10.32782/2304-0920/3-88-13
Subject(s) - unemployment , multicollinearity , econometrics , economics , regression analysis , population , econometric model , multivariate statistics , principal component analysis , statistics , demographic economics , mathematics , demography , economic growth , sociology
The aim of the study is to conduct a statistical analysis of the modern labor market and adapt a multivariate econometric model of unemployment in Ukraine using the principal component analysis. The paper investigated the current state of unemployment in Ukraine for the last two decades. The dynamics of the unemployment rate and employment of the economically active population in Ukraine is analyzed. The analysis of the structure of the unemployed for reasons of dismissal and the tendency of changes in the size of the working age population is carried out. The gender aspect of the number of the unemployed population is investigated.Comprehensive assessments of the resources of labor potential by regions have been calculated and a rating of regions has been built. The disproportionality behind the Harrington`s desirability function was analyzed taking into account the factors of stimulants and de-stimulants. The rating assessment of unemployment for regional labor markets of Ukraine is given for the gradation of values of the desirability function.The main macroeconomic factors of influence on the level of unemployment in Ukraine have been determined. Structural and correlation-regression relationships have been analyzed. The identification of the model has been carried out. The multivariate unemployment model was adapted. The factorial database was checked for the presence of multicollinearity behind the Ferrer - Glober algorithm based on the criteria: Fisher, Spearman and Student. With the help of component analysis, the study of the relationship between factor variables was carried out. The factor loading matrix was constructed and analyzed. The matrix of the values of the principal components was calculated. The model of unemployment is constructed by the principal component analysis. The model was tested for adequacy, its economic content was analyzed. The residuals (random variables) are estimated to establish the quality of the constructed multivariate model. Dynamic models of factorial variables were built and their values for the next year were estimated. Through the normalized values of estimates of factor variables for dynamic models, the unemployment rate in Ukraine for the future (2021) was calculated.