Annie MOORE: Increasing Employment of Resettled Refuges using Machine Learning and Optimisation | University of Oxford

Annie MOORE: Increasing Employment of Resettled Refuges using Machine Learning and Optimisation

Tens of thousands of refugees are permanently resettled to the United States, the United Kingdom, and other countries each year.

Karen Monken, HIAS Associate Director for Pre-arrival, uses the Annie MOORE softwareKaren Monken, HIAS Associate Director for Pre-arrival, uses the Annie MOORE software
There is ample evidence that the initial community to which refugees are resettled dramatically affects their lifetime outcomes. This project shows that if refugees are resettled to the communities that are best suited to their needs and aspirations, both refugees and communities can thrive.

The project’s pioneering software, Annie™ MOORE (Matching and Outcome Optimization for Refugee Empowerment), suggests placements of refugees in order to maximise their employment chances. Annie™ also ensures that the needs of the refugees (e.g., childcare or language support) are met and the service capacities (e.g., housing or places in training programmes) of hosting communities around the United States are not exceeded. To make placement suggestions, Annie™ uses advanced machine learning and state-of-the-art integer optimization methods.

HIAS, one of nine US refugee resettlement agencies, has been using Annie™ since 2018. Annie™ has thus far matched over 1,100 refugees resettled by HIAS. The project team estimate that Annie™ has obtained over 30 percent boost in the number of employed refugees (taking the employment rate from 30 to 40 percent) (Trapp et al., 2018). Annie™ has also reduced the fraction of refugee families who are placed in communities which cannot provide services to support them from around 20 percent to essentially zero. This has dramatically improved the quality of refugee integration in communities. 

Finally, Annie™ has empowered HIAS staff. Karen Monken, HIAS arrivals director says: “The effectiveness of my operations has increased dramatically. I now spend 80 percent less time on routine matching, and can focus my time and energy on the more difficult cases such as those with significant medical conditions, ensuring that their placement is as good as possible.”

The development of the optimisation and matching techniques was based on the research funded by Dr Teytelboym’s ESRC New Investigator Grant. The project was also supported by the National Science Foundation, Jan Wallander and Tom Hedelius Foundation, the Ragnar Söderberg Foundation, and Skoll Centre for Social Entrepreneurship Research Accelerator Grant.