C38 - Multiple or Simultaneous Equation Models: Classification Methods; Cluster Analysis; Principal Components; Factor ModelsReturn

The social policy of the European Union is focused mainly on the most vulnerable group of citizens. The social instruments that countries use to increase the living conditions and to decrease social exclusion and vulnerability can differ between countries. The submitted paper uses cluster analysis to compare 27 member states of the European Union based on their social policy instruments in 2019. The study analyses the dependence of social policy instruments and household financial vulnerability using correlation and regression analysis. It examines whether higher levels of social benefits and government expenditure result in lower financial vulnerability. However, the results suggest that the financial vulnerability of households is not associated with low benefits. Clusters reporting a low level of benefits recorded a low financial vulnerability of households and vice versa. Results of the correlation and cluster analysis prove that the increase of the first factor (benefits in material need, expenditure on active labour market policy, family benefits, disability benefits) and the third factor (unemployment benefits) are associated with a decrease in financial vulnerability of households.

Irresponsible behaviour practices of tax entities are becoming more sophisticated. In addition to repressive tools, the tax administrator may use a proactive approach to detect an entity’s illegal activity. The paper aims to identify the relevant factors and assess their impact on estimating the probability of irresponsible behaviour of companies towards the tax administrator in individual sectors of the economy. Based on data from financial statements, we monitored the characteristic features of tax-responsible and tax-irresponsible companies in the Slovak Republic in the period 2015–2018 in 15 sec-tors. We created predictive models of their irresponsible behaviour towards the tax administrator for each industry using logistic regression. The results show that an analysis of the quantitative features of companies makes it possible to identify factors that alert the tax administrator in advance to potentially illegal behaviour of tax entities.

The topic of this paper is customers' trust and satisfaction with online group buying. The model of customer satisfaction was designed as a synthesis of the Expectation - Confirmation Model (ECM) and the DeLone & McLean IS Success Model. Reputation and perceived size of the online group buying website, system quality, information quality and service quality were factors analysed in the model. Trust was perceived as a predictor of customers' satisfaction. Data from the questionnaires were analysed using structural equation modelling. It was found that reputation and perceived size of the online group buying website have the most important impact on customers' trust and satisfaction.

The research database consists of the OECD countries except Iceland, Latvia and Turkey, which were excluded because of insufficient data. The primary objective of the study is to group the countries according to their average wage, GDP per capita, minimum wage and unemployment rate. Another objective, of no less importance, is to determine which of the three remaining above variables significantly affect the average wage, while defining the type and strength of this relationship. Yet another important goal is to develop forecasts of the wage level and GDP per capita for each OECD country by 2020. In terms of clustering OECD countries by the four variables, the Czech Republic always ranks alongside Chile and three post-communist countries, Estonia, Hungary and Poland. GDP per capita is the only explanatory variable significantly affecting the average wage. The dependence of these two variables is represented by a second-order polynomial (concave parabola), the selected regression parabola explaining approximately 88 percent of the variability in the observed levels of the average annual wage. The conversion of the average wage, GDP per capita and minimum wage to purchasing power parity allows consideration of different price levels and thus comparison of purchasing power parity of the population in different countries.

The article investigates the connections between the prices of biofuels and many traded commodities and other relevant assets in Europe, USA and Brazil. The analysis uses a comprehensive dataset covering price data for 32 relevant traded titles over the period 2003-2015. Main contribution of this article is a combination of minimum spanning tree and hierarchical tree approaches with expert economic understanding of biofuels market leading to identification of price connections in a complex trading system. Our analysis of mutual price connections discovers the major defining features of world leading biofuels markest over the last decade. We provide characteristics of main bioethanol and biodiesel markets with respect to technical and local features of the production and consumption of particular biofuels.

The equivalised total net annual incomes of the Czech households (in CZK) in 2007-2010 are analysed in the text. The set of all households is very nonhomogeneous (with respect to incomes) and the aim of the analysis is to determine more homogeneous subsets (components) and to describe the distribution of incomes in these components. The components are supposed to be artificial, the membership of households in components is not known (or observable). A multivariate mixture of normal distributions (four dimensional component distributions) is used to describe a vector of logarithms of incomes, models with 2 to 9 components are fitted. Maximum likelihood estimates of unknown parameters were found with the use of EM algorithm. Akaike information criterion was used (accompanied by bootstraped test) and models with 3 or 4 components were selected to be acceptable for the description of distribution of incomes. Cluster analysis was performed in order to classify households into components and good performance of the model was found.