I analyze the case of Turkey which is a developing country hosting largest refugee population in the world.
Firstly, on average, there is no short-run impact of refugees on crime rates (both high criminal and basic criminal court cases):
For the long-run impact, importantly, there is a staggered treatment adoption👇
So, employing the semi-parametric DiD estimation
technique of Callaway and Sant’Anna (2020), I find that there is no impact of hosting a refugee population above a threshold level.
But, it seems that treatment intensity is also different for treatment provinces:
So, I also checked the impact of treatment intensity on crime rates in the long run (using IV) and find that per capita crime rates decreased, and no effect exists on crime rates per 100,000 natives.
These results are robust to various identification threats which are explained in the paper in detail.
So, please read the paper for those discussions and for my interpretation of the results. Hope you will enjoy!
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Based on the analysis of quantitative and qualitative data, we found that Syrian refugees’ return aspirations are shaped mainly by fear, uncertainty and lack of trust regarding Syria’s situation.
(2/n)
For the majority, the end of war is a sine qua non to return to Syria but would still fall short of ensuring it alone without safety, livelihood opportunities and essential services in heath and education.