Determinants of the state of poverty using logistic regression

The aim of the paper was to identify the determinants of the state of poverty using logistic regression. The analysis focused on economic poverty considered through the prism of income. Three states of poverty were considered: poverty, near poverty (household’s income from 100% to 125% of the adopted poverty threshold) and above near poverty (income higher than 125% of poverty threshold), using the ordinal logit model and – after the rejection of the proportional odds assumption – the multinomial logit model. The analysis was preceded by a presentation of the basic facts concerning three states of poverty. Based on the conducted analysis it can be stated that the education of the household's head, place of residence, labourforce status and socio-economic group were very important factors of the state of poverty, and they change the odds of being in above near poverty relative to poverty and the odds of being in near poverty relative to poverty.