Policy maker and researchers are alarmed with pervasive substandard working conditions and mistreatments in domestic work worldwide. Our study contributes to this literature by exploring the types of abuse committed on domestic workers. We draw on an original dataset from a sample of domestic workers in Portugal (N=684) and offer empirical evidence on types of abuses and the factors that increase (reduce) the risks of those abuses. Multiple correspondence and cluster analyses offered three segments of domestic workers according to the types of abuses reported in the survey. We labelled those segments as labour abuses, the ones associated with the employment relationship, multiple abuses, which include labour and other severe abuse, and no abuses, the segment with no occurrence of abuses. The next stage of empirical analysis consists in exploring the predictors of each cluster of abuses. The results from multinomial regression model offer evidence on factors that reduce/increase the risk of each type of abuse. While migrant workers are more vulnerable to all types of risks, trust related issues, child caring, and specific skills protect domestic workers against risk of abuses. However, general skills and elderly caring increase the risks of multiple risks. Policy makers should be endowed with information on how to reduce the risks of each type of abuses.

--

domestic workers; abuses; informality; migration; multinomial regression model