The perception of European workers on the adoption of telework during and after the Covid-19 Pandemic: The use of supervised and unsupervised learning techniques

Globally, 2020 began with unprecedented changes. With the worldwide declaration of the COVID-19 pandemic, the transition to teleworking took place instantly, becoming one of the largest historical experiments. The objectives of this study are to identify a limited number of characteristics of European workers regarding telework trends, to find typologies of groups of European countries where workers similarly perceive the forced adoption of telework imposed by the COVID-19 pandemic, and to find common patterns between Romanian and other European workers. At the same time, the research aims to predict the behaviour of European workers after the COVID-19 pandemic and how the attrition towards working from home can be improved. The Eurofund 2020 survey was used to conduct quantitative research. To achieve the first objective, unsupervised learning techniques (principal component analysis) were used to highlight the types of European workers. and how they have been affected by telecommuting. The results showed groups of European workers who were deeply affected in terms of isolation, personal life or work satisfaction, but also groups of European workers who can claim to be gainers as a result of remote work. For the second and third objectives, cluster analysis was selected as the method. The similarities and differences between the perceptions of European workers regarding the adoption of teleworking were assessed, with Romanian citizens having the same concepts as Poles and Irish. For the objective of predicting the high attrition of homework, following the results from the first objectives of the analysis, workers from European countries were divided into home workers and office workers and based on these, using supervised learning (logistic regression) one can predict which component should be improved in order to have a greater attrition regarding home workers.

Keywords: telework, European workers, principal component analysis, clustering, unsupervised learning, COVID-19

JEL Classification: A130, C150, J440