Modelling refugee migration under cognitive biases: Experimental evidence and policy

In this paper, we develop a model for refugee migration. As refugees' migration choices are made in a risk-laden environment, we compare two different theoretical frameworks of decision making under risk, namely Expected Utility Theory (EUT) and Cumulative Prospect Theory (CPT). This last framework accounts for a reference point, loss aversion, and probability distortion. We estimate refugees' risk and time preference parameters using field experimental data and show CPT better explains refugees' risk behaviour on average. We also investigate policy implications based on simulations. We show that, under CPT, compared to standard EUT, the value of migrating is consistently lower and the migration decision is more sensitive to policy changes. Our results suggest refugees may self-select based on their risk preferences, those exhibiting more loss aversion or less probability sensitivity being more likely to renounce migration as a reaction to migration policies.