Annie™ MOORE (Matching for Outcome Optimization and Refugee Empowerment)

While tens of thousands of refugees are permanently resettled to host countries every year, governments lack the capacity to know which communities to place which refugees.
Annie™ MOORE, used by the resettlement agency HIAS, deploys advanced analytics to recommend communities that are most likely to maximize refugees’ odds of being employed.
Annie™ boosts employment chances by at least 30% over manual placement and ensures that the needs of refugees and community capacities are both respected.

Innovation Summary

Innovation Overview

There are over one million refugees around the world who, according to the UNHCR, are in urgent need of resettlement, i.e., permanent relocation to another county. In recent decades, the United States (US) has admitted the vast majority of resettled refugees. However, the US government faces a crucial task: to which communities should the refugees be placed when they arrive? The outcomes of refugees depend strongly on the first community into which they are placed. For decades, the US government has been collecting data on whether refugees were successful at obtaining employment, however, until now these data were not put to good use.
Our pioneering software Annie™ MOORE (Matching and Outcome Optimization for Refugee Empowerment) uses advanced machine learning and state-of-the-art integer optimization methods to recommend optimal placements of arriving refugees across hosting communities around the US. HIAS, one of nine US refugee resettlement agencies, has been using Annie™ since May 2018. Annie’s objective is to suggest placements of refugees that would maximize their employment chances while simultaneously ensuring that the needs of the refugees (e.g., child support or language support) are met and the service capacities of hosting communities (e.g., housing or places in training programs) are not exceeded.

The primary beneficiaries of Annie™ are refugees resettled in the United States who now have access to better employment opportunities. The secondary beneficiaries of Annie™ are the residents of communities where the refugees are hosted. Annie™ ensures that communities only welcome refugees that are most likely to find employment there and whom they can help integrate. This reduces community tensions and builds goodwill toward refugees. The tertiary beneficiary is HIAS itself which have been able to streamline and their processes and, by extension, add value to the broader American refugee resettlement program. As more refugees are able to get into employment and become self-sufficient, the return on government funding for resettlement dramatically increases.

Annie™ only recommends the allocation of refugees across communities to HIAS. Therefore, HIAS staff have complete control over the final allocation and ultimate responsibility over where the refugees are placed. Nevertheless, we have found that HIAS staff have been extremely impressed with Annie’s ability to place refugees and rely strongly on Annie’s suggestions.

Annie™ is already perfectly adapted to the refugee resettlement context of the United States. Therefore, Annie™ can be adopted almost immediately by the other eight resettlement agencies contracted by the US government. The benefits of this adoption would spill over to all the other agencies: the richer and more complete data shared across agencies would allow Annie™ to more accurately predict employment likelihood for every agency. Finally, the interface and methods developed for Annie™ can also be easily translated into international contexts. For example, Annie™ could be used for the Syrian Vulnerable Persons Resettlement Schemes in the United Kingdom or for the allocation of recently arrived refugees in Sweden.

To understand how Annie™ works, one can consider two separate steps. First, they apply machine learning methods to find patterns in a large dataset covering all refugee placements by HIAS over the past decade. Their  machine learning algorithm is able to pick out the characteristics of refugees that make them likely to get employment in particular communities. Therefore, the model is able to predict the likelihood of employment for newly arriving refugees in any of the communities where HIAS operates. In the second step, organisers decide which communities to allocate all refugee families to maximize overall employment. However, to do that, they need to keep track of different needs of refugees and the service capacities of the communities. In particular, they need to ensure that the total number of placed refugees does not exceed the government-approved annual quota. Moreover, the communities must be able to meet the needs of refugees that join them. For example, if all family members only speak Arabic, we ensure that the community has support staff who are Arabic speakers. There are many constraints of this kind and it is extremely difficult to keep track of them when allocation is done manually. The organisers' integer optimization methods, however, allow us to maximize employment which ensuring that all the needs and capacity constraints are met. Therefore, Annie™ is simultaneously able to improve employment outcomes and ensure that none of constraints are violated!

The inspiration for the research came from the National Resident Matching Program that matches tens of thousands of medical to their residency programs and from patient-organ donor matching programs around the world. These innovations were behind the Nobel Prize in Economics awarded to Al Roth and Lloyd Shapley in 2012.

Innovation Description

Innovation Development

Innovation Reflections

Project Pitch

Supporting Videos

Leave a Reply

Your email address will not be published. Required fields are marked *

Year: 2018
Level of government: National/Federal government

Status:

  • Implementation - making the innovation happen
  • Evaluation - understanding whether the innovative initiative has delivered what was needed
  • Diffusing Lessons - using what was learnt to inform other projects and understanding how the innovation can be applied in other ways

Innovation provided by:

Join our community:

It only takes a few minutes to complete the form and share your project.