Predicting international migratory movements using Google searches
1Those aspiring to migrate typically acquire information about migration opportunities online, in their countries of origin, prior to departure. These data can be used to make short-term predictions about numbers of migrants, which may be very useful in situations such as humanitarian crises.
2Google search data, which represent searches by over 1 billion people worldwide, are free of charge for small queries. Using the data to study international migration can be helpful, especially in developing countries because, to date, information on migration and people’s migration intention is scarce or exclusively available to paying users.
3This approach tends to work better when focusing on countries where internet is more widely used and the Google search engine is more widespread, and where more people speak the languages for which search data are studied.
The solid line in the graph represents real arrivals of migrants to Spain originating from Venezuela between 2004 and 2015, according to the International Migration Database compiled by the Organisation for Economic Cooperation and Development (OECD).
The dashed line corresponds to the predictions based on the benchmark model including GDP and population dynamics at both origin and destination as explanatory variables. The dotted line is the result of the previous estimation, augmented with the data from Google searches in Venezuela. The graph shows a sharp increase in immigration since 2012, presumably due to the socio-political and economic crisis in Venezuela. The keywords searched for on Google help predicting this upward tendency more accurately.