ABSTRACT
The objective of this study consists of analysing the role of protection incidents on migration aspirations and destination preferences of refugees and migrants in Libya. We use a unique dataset that combines the 4Mi micro-level data from the Mixed Migration Centre with traditional macro-level variables. The micro-founded gravity equation allows using McFadden’s choice model that explains whether the experience of persecution in the origin country and protection incident in Libya shape onward migration aspirations and destination preferences of people on the move. Main results show first that refugees and migrants who reported having experienced persecution in their origin country that made them leave are more likely to aspire to migrate onward to Europe. Second, having experienced protection abuses in Libya increases onward migration aspirations to migrate to France, Germany, Sweden or the UK, but not to Italy. The findings call for comprehensive, contextualised and tailored approaches to programme and policy interventions that aim to provide long-term solutions, assistance and protection to people on the move. They also call for early warning and information systems in forced migration which would relate to the situation in countries of origin and transit.
Acknowledgements
I would like to thank Ayla Erin Bonfiglio and Ana-Maria Murphy-Teixidor for their guidance and support from the very early stages of this project, as well as the editor and the two anonymous referees for extremely detailed and helpful feedback that has substantially improved the paper. Comments and suggestions from Bram Frouws and Jane Linekar on another version of the research study were also extremely useful. I would like to thank Jim Van Moorsel and Sarfaraz Ali Shah Syed for their support. The views expressed in this paper are those of the author and do not necessarily represent those of the Mixed Migration Centre (MMC). All errors and omissions are mine.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Appendix A displays the list of variables included in the model.
2 Other results related to control variables at the individual and structural level are reported in Appendix C.