By Amira Hijazi
Available data regarding the flow of undocumented immigrants is subject to an extreme physical bias. To correct this bias, Dr. Edward Kaplan, William N. and Marie A. Beach Professor of Operations Research, Professor of Public Health & Professor of Engineering at Yale University, uses probability to model the migration flow of undocumented Mexican immigrants to the United States since 1980. In his model, he distinguishes between two types of immigrants—solitary and circular migrants—and shows that before 2005 there was a high level of circular migration due to lax border security. However, after 2005 and the intensive security, the circular migration level decreased but the solitary migration has increased. The results shown suggest a much larger number of undocumented immigrants in the United States than the residual method. Correcting the bias of the immigration flow model was shown to be doable by applying probability theory but what about the bias of deportation of immigrants’ policies?
Dr. Mohammad Fazel-Zarandi from MIT presents an interpretable decision tree approach to answer this question! Immigration and customs enforcement (ICE) is a secretive institution that identifies and prioritizes the deportation of immigrants who commit crimes, based on the severity of the crime. However, the area under the curve criterion does not support this claim based on the models developed by Dr. Zarandi! Moreover, it questions the effectiveness of the new policy, called “secure communities program,” over the previous policy, called “priority enforcement program.”
Moving from immigration policies and laws to refugees and their needs, Narges Ahani, a PhD student at Worcester Polytechnic Institute under the supervision of Dr. Andrew Trapp, presents a new strategy for refugee resettlement that takes into consideration the dynamic nature of their arrivals as opposite to the current resettlement policy that is done only once a year which could result in inefficient use of resettlement capacity, and lower refugee welfare.
Closing the session, Dr. Nicholas A. Arnosti, assistant professor at Columbia Business School, discusses the inefficacy and unfairness of the current lottery systems, that is, the H1B visa lottery, and shows that small changes in the system could lead to a significant improvement. An example of such a change is what he calls “group lottery,” where valid groups are placed in a uniform order and processed sequentially. As a result, each group the minimum of their request and the number of tickets remaining.