The Role of Migration Research in Promoting Refugee Well-Being in a Post-Pandemic Era

This paper summarizes the presentations and discussions of a virtual stakeholder meeting on Refugee Resettlement in the United States which built on the foundation of the May 2019 workshop represented in this special issue. With support from the Robert Wood Johnson Foundation and the Andrew W. Mellon Foundation and hosted by the Committee on Population (CPOP) of the US National Academies of Sciences, Engineering, and Medicine on Dec 1–2, 2020, 1 the meeting convened migration researchers, representatives of US voluntary resettlement agencies, and other practitioners to consider the role of migration research in informing programs serving refugees and migrants during the COVID-19 pandemic, continuing an emphasis on bringing global learning to those on the ground working with refugees. The goal of CPOP's work in this area has always been to build bridges between communities of research and practice and to create a dialogue for a shared agenda. We present the goals and framework for the 2020 meeting, followed by a summary of each of the four sessions and themes that emerged from these discussions. The paper ends by considering effective ways of amplifying the role of research in refugee policy and programs of refugee resettlement in the United States and how demographers and population researchers might contribute to this goal.

We present the goals and framework for the 2020 meeting, followed by a summary of each of the four sessions and themes that emerged from these discussions. The paper ends by considering effective ways of amplifying the role of research in refugee policy and programs of refugee resettlement in the United States and how demographers and population researchers might contribute to this goal.

Meeting Goals
The goals of the meeting were to create dialogue around the presentations and discussions from the May 2019 workshop that will be useful for refugee resettlement experts and organizations in the United States and elsewhere, and to continue to foster and create a cooperative venture between US VOLAGs, refugee resettlement agencies around the world, and the global community of population scientists and forced migration scholars. These goals were operationalized through four sessions: (1) concepts, definitions, and data on refugees, forced migrants and migrant populations; (2) ethical issues in migration research; (3) modelling and projecting forced migration and population displacement; and (4) research design in evaluation of migrant integration and program evaluation. With this organizational structure, continuity across the two US National Academies of Sciences, Engineering, and Medicine (NAS) meetings was ensured.
The goals of the 2020 stakeholder meeting further emphasize the role of research in practice and evidencebased programming, the value of international models and comparisons for US refugee resettlement and assistance programs, the recognition of US refugee resettlement being local and community-based, and of refugee integration and well-being. An overarching theme which guided the meeting was the context of the COVID-19 pandemic which put into relief a range of issues regarding refugee and forced migration research.
Summary of the Meeting 1. Concepts, definitions, and data on refugees, forced migrants, and migrant populations Romesh Silva of the United Nations Population Fund (UNFPA) 2 chaired the first session and began by noting that the geographic and demographic characteristics of refugees and displaced populations hold implications for concepts, measures, and data collection. Most displaced populations reside within countries where statistical systems need strengthening; displacement both within countries and across international borders is often protracted. There has been a "strategic pivot" in international statistical perspectives and practices; measures and data on forced migrants are now considered essential to support operational response within host countries. The UN Statistical Commission has promoted standardized statistical concepts and methods of data collection within national statistical practices to align with both national statistical practices as well as bi-and multi-national and regional comparisons of forced migration and migrant populations. Representing the UNHCR, Sebastian Steinmuller outlined the processinitiated by the UN Statistical Commission in 2016of developing international recommendations on national statistics on refugees and internally displaced persons (IDPs). The development of the recommendations revealed significant global and regional gaps in demographic data on refugees and IDPs, including the lack of data at subnational scales and within urban areas, paucity and inconsistency of indicators of social and economic characteristics of refugees and displaced populations, and the need for indicators of sustainable development goals disaggregated by refugee and displacement status and characteristics. A critical perspective of the work of the Expert Group on Refugee and Internally Displaced Persons Statistics has been the mainstreaming of forced migration statistics within national statistical practices and the promotion of country-led implementation of the recommended statistical definitions.
Bouchra Bouziani, of the High Commission for Planning of the Government of Morocco, described initiatives within her country to initiate a program of data collection on migration. As a sending country, transit country, and host country for refugees and migrants, Morocco embodies a diversity of migration and forced migration dynamics. Morocco has prioritized expanding the evidence base regarding migrants and refugees. A whole-of-government approach has been implemented to build capacity throughout government administration for analysis of data on migrants and refugees. Traditional sources of data collectioncensuses and household surveysinclude refugees; the 2024 census of Morocco will collect data on reasons for migration for purposes of identifying refugees among foreign born populations by country of birth. Special surveys regarding the effects and implications of COVID-19 have recognized the unique vulnerability of refugees to the pandemic. Harmonization and linkage among national administrative data systems remain a challenge in Morocco, but there is also an advantage in building capacity now for migration research as administrative statistical systems are in the process of development.
The case of Norway provides an effective illustration of the leveraging of administrative data systems for studying integration over generations. Vebjørn Aalandslid of Statistics Norway described the history and operations of the population registration system in place in Norway and illustrated the use of the registration data for national public policy analyses as well as for international comparative analysis. An internally consistent population registration system in Norway emerges from the linkage of over 100 national administrative registers, yielding current stock data on registered population groups, such as refugees and international migrants. Persons seeking asylum, however, are excluded from the registration system. The Norwegian population register maintains refugee background permanently within the data file, providing a basis for current and longitudinal analysis of social and economic integration of persons and families with refugee backgrounds, even the next generation. This type of analysis allows researchers to study the impact of refugee status and its continued effects over time.
The COVID-19 pandemic has revealed weaknesses concerning population data systems and their sensitivity to issues faced by forcibly displaced persons. Forcibly displaced personsrefugees, asylees or persons seeking asylum, and persons in flight and or in transit campsare less visible, or invisible, in national statistical systems compared to other foreignborn residents and regular migrants. Many national statistical systems are not well designed to support high-quality and routine official statistical data on the size, demography, and socio-economic situation of forcibly displaced persons. Norway is a case in point. COVID has reminded us of the importance of inclusive population data systems and the integration and interoperability of these systems.
It is also important to note that there is an emerging body of good practices that needs to be studied, refined, and adapted to other settings (where possible). The work of the UN Statistical Commission, described above, to advance common statistical definitions of international migrants, refugees, and displaced persons, is one example. Regarding sources of data, the governments of Lebanon and the State of Palestine recently completed a special population and housing census of Palestinians in Lebanonhelping to clarify the size, demography, and socioeconomic needs of Palestinian refugees in the country. The National Statistics Office in Colombia has been coordinating for several years now with the government's Victims Unit (the main official body responsible for documenting and supporting IDPs in the country) to improve the quality of IDP registration data and incorporate these data into the country's official statistics.
There have also been important initiatives in the use of surveys in the analysis of forced displacement. The World Bank Group and others have piloted several rapid phone-based surveys to understand the impacts of COVID-19 on forcibly displaced persons in Ethiopia, Djibouti, and elsewhere. These efforts demonstrate the potential of leveraging census and administrative data systems to construct usable sampling frames to support rapid surveys. These systems also can allow the monitoring of sociodemographic characteristics of forcibly displaced persons and ensure evidence-based response and action during a pandemic.
There remains a critical need for more investment and technical work in developing national statistical systems so that they serve and support the needs and interests of the forcibly displaced (Baal 2021). As noted below, data protection standards and ethical practices remain weak when it comes to population data systems on forcibly displaced persons. For example2021 2], Shoemaker et al. (2020) have noted that in many humanitarian contexts, data collection efforts and systems are not centered around the rights and needs of forcibly displaced persons, but rather structured around the needs of large institutions; this critique is similar to the argument against "top-down" programs made by Wessels in his paper in this special issue. There exists the possibility that the inclusion of forcibly displaced persons in some administrative data systems and/or legal identity systems effectively translates into vulnerable persons handing over their personal information and biometric identifiers in exchange for humanitarian assistance (Shoemaker et al. 2020). This point resonates with issues raised by Gruskin, which were presented in the session on ethics, discussed below. Sofia Gruskin of the University of Southern California considered the implications of the COVID-19 pandemic for migrants and refugees in the United States. She argued that particular focus should be given to how tools of law and public administration, at all scalesincluding US immigration law, programs pertaining to health and social and financial assistance, criminal law, and so onare being interpreted and implemented in the name of COVID-19. Attention should also be given to variations in application of law and implementation of programs in relationship to the experiences of migrants and refugees in the United States, among groups, across space, and over time. The interconnections among public health programs, migrant policy, and politics are revealed in variations in vaccine hesitancy among migrant groups and go to issues of trust in government and the integrity of public health policy. Research is needed to assess the short-term and long-term effects of these interventions, recognizing the etiology of disease and underlying determinants of health, as well as the resilience of migrants and refugees. Such research, including research on levels of trust in public health programs and health care workers, would inform health delivery programs, such as COVID testing and vaccine distribution.

Ethical issues in migration research
Hamutal Bernstein of the Urban Institute discussed the resources of the US Office of Refugee Resettlement and the Annual Survey of Refugees (ASR) in relationship to ethical dimensions of forced migration research. Ms. Bernstein placed the survey within the context of the sparse amount of national data on the experience and integration of refugees who are resettled in the United States. The Syrian refugee crisis revealed both the (potential) value of the annual survey for social policy analysis as well as the deficiencies of the data collection and the inappropriate interpretation of survey results. Annual reports to Congress, for example, have failed to report margins of error around sample estimates, resulting in misleading conclusions about the relative outcomes for different groups of US refugees. Mr. Steinmuller considered these issues in relationship to issues of confidentiality in aggregated data: for example, in many national contexts, asylum flow data cross-tabulated by country of origin came close to revealing individual migrants. Susan Martin of Georgetown University elaborated on the human rights implications of COVID-19 international travel restrictions on refugees and migrants through a subsequent publication (Martin and Bergmann 2021). The analysis demonstrates that the pandemic, government responses, and resulting economic impacts can lead to the involuntary immobility of at-risk populations, such as aspiring asylumseekers or survival migrants. Closed borders and travel bans can have devastating consequences for people seeking protection from life-threatening situations. Similarly, stay-at-home measures have posed dire challenges for migrants living in precarious, crowded circumstances. They may entrap migrant workers, detainees, and others confined to cramped quarters where the spread of infectious diseases is common. This phenomenon has been present, for example, within US detention centers during the pandemic (see Tosh, Berg and Leon 2021).
The global economic contraction accompanying the pandemic has increased involuntary immobility by reducing both people's resources to finance a move. The United States, now over one year into the pandemic, is experiencing very high demand for labor and very tight labor markets. People's attempts to protect themselves from the virus can result in shifting patterns of mobility, such as increases in cross-border return migration and urban-to-rural movements. Policies that discriminate against migrants in terms of access to healthcare or income support, not uncommon when resources are scarce, can lead to precipitous and dangerous flight. They can also dissuade migrants from seeking critical healthcare and vaccinations, which can, in turn, spread the disease.
Concerns about the need for protection of the human rights of displaced persons during the pandemic spurred a group of academics to draw up 14 principles of protection, which apply to all persons, irrespective of their immigration status (https:// zolberginstitute.org/covid-19/). Actions taken to address the pandemic must be consistent with established international human rights norms including those of non-discrimination, rights to health and to information, due process, and non-return to risks of serious harm. The principles are intended to inform and guide State action, to assist international organizations, and to provide a basis for advocacy and education.
The pandemic has also increased the incidence of mental health issues while at the same time mental health service providers have transitioned to delivering their services virtually. Migrants, along with many low-income communities, often have limited digital tools and internet connection thus limiting their access to virtual mental health care. A Canadian research project led by Michaela Hynie of York University is studying the experiences of recent refugee newcomers and settlement workers in accessing and using mental health services during COVID-19 (Evidence Exchange Network for Mental Health and Addiction 2021). Barriers to accessing care include challenges in building a relationships and trust with the provider, lack of private spaces in which to have virtual meetings, internet issues (dropped calls and connectivity issues), discomfort about having to use less familiar online platforms, and difficulty in maintaining attention in the virtual space. Participants recommended mixing virtual services with face-to-face services when conditions allow and according to the needs of clients, utilizing modalities familiar to refugee newcomers to lower barriers to participation, and providing further education and support. A recent report by a group of Canadian settlement agencies concluded that hybrid service delivery is an inevitable and positive future model for the immigrant and refugee-serving sector and that people, not technology, should be at the center of a future hybrid service delivery strategy (Liu et al. 2021).
The ethical practices of refugee research were closely examined by one of the 2019 workshop presenters, Christina Clark-Kazak, who has worked with researchers and practitioners to develop ethical research guidelines in the field of forced migration. COVID-19 has challenged ethical research practices. Travel and government restrictions have made it difficult for researchers to get to sites of forced displacement to investigate the impact of the pandemic, particularly the impact of actions by authoritarian states against at-risk populations. The ethics boards of universities have imposed strict guidelines on how researchers can access participants; no personal contact is typically allowed. Concerns are raised about the fewer number of participants who can be contacted virtually and the resultant quality of the research findings when the forced migrants themselves are often not contributing. Researchers are also having difficulty accessing service providers to participate in research; they are stressed and exhausted by the increased work demands, with many also trying to support remote schooling of their children. Greater reliance by non-travelling researchers on local researchers at sites of displacement, however, may also build capacity and disrupt power imbalances.

Modelling and projecting forced migration and population displacement
Sarah Staveteig Ford of the US Department of State, outlined the benefits of models of forced migration and displacement for planning for resources and support in response to humanitarian crises. 3 The complexity of "shocks" that drive large-scale forced migration with little warning pose fundamental challenges to modelling, and ultimately in the distribution of their effects on local populations. It is often a challenge for modelers to reliably forecast the scenarios that drive migration, andeven for known or ongoing eventsthere is enormous uncertainty in how complex systems will adapt and what the resulting effects on migration will be. Recent modelling advancements in forced migration studies include the increased use of Bayesian probabilistic approaches and complex agent-based approaches, as well as investment in interdisciplinary perspectives and collaborations. Jean-Christophe Dumont of the Organization for Economic Cooperation and Development (OECD), described models of migration and forced migration developed by the OECD for purposes of information policy response, specifically, early warning systems for population displacement. Results produced by models, however, are typically too dated for intervention to prevent or address displacement. Mr. Dumont expanded on challenges in representing uncertainties, particularly given the dearth of empirical information regarding migration intentions, motivations, fear, and behaviors.
Alexander Kjaerum of the Danish Refugee Council (DRC) described the "predictive analytics" of the DRC models of forced migration, and the impacts of COVID-19 on displacements. For countries with historical data on (forced) migration, the models generate predicted displacements one year out, by country of origin, and estimates of relationships among key drivers of displacement. The outcome is a foundation for the evaluation of scenarios for individual countries, including the analysis of shocks, such as COVID-19, within specific national contexts. Using this approach to modeling/scenario building, the DRC has evaluated the demographic impacts of the pandemic in several regions and nations on sub-Saharan Africa, yielding information that has been useful to donor organizations seeking to help.
Mr. Kjaerum highlighted the need to engage both new users of the models and new approaches to scenario building. Notably, because of the political volatility of issues of immigration within the European context, the DRC has shifted the focus of modeling to prediction of forced migrations and displacement.
In both modelling and practical terms, the COVID-19 pandemic was an event totally "out of scope," rendering many models less than relevant given global, regional, national, and local events regarding infection, morbidity, and mortality. Going forward, models need reconfiguration to include the parameters of a future pandemic, to both predict and understand dynamics of a pandemic (see, e.g., Niu, et al. 2020;Ahsan et al. 2020). Presenters and participants considered how to amplify and communicate the value of models of forced migration and displacement for policy makers, and donor organizations. The utility of forecasts in efforts at prevention was discussed; that is, advocacy for the positive use of modeling. Anticipation of displacement can save lives. On the other hand, the case of COVID-19 revealed that early warning systems can also trigger the closing of borders. This effect may expand the focus from prevention to include the protection of populations vulnerable to displacement.

Research design in evaluation of migrant integration and program evaluation
Pia Orrenius 4 of the Federal Reserve Bank of Dallas highlighted several analytic themes concerning the integration of refugees that emerged from the earlier workshop: conceptualization and measurement of integration must account for selectivity among migrants, including refugees, relative to populations of origin, and differentials, relative to populations and communities of settlement; the process of migration has implications for measurement of integration into host communities; and assessment of the effects of programs promoting integration among refugees requires effective research design, and ideally, longitudinal data collection. Research on refugee and migrant integration should integrate across scales, considering individual level characteristics within the context of migrant communities and culture: dimensions of cultural and community resilience are critical to understanding patterns and pathways of refugee integration and well-being. Results from the ASR reveal a general story of integrationsocial, economic, and civicamong refugees resettled in the United States. Ms. Bernstein emphasized, however, that the population "average" belies the underlying story of significant differences among groups of refugees. There are differences in individual characteristics and integration trends by national origins and cohorts of arrival. Redesigning the ASR has allowed the survey to measure integration more effectively by expanding the concepts to include additional noneconomic outcomes among refugee groups. Reliability of data for refugees, however, declines with longer lengths of residence in the United States. The ASR has untapped potential to provide broad benchmarks for evaluation of programs supporting refugee settlement in the United States.
Illustrations of the use of surveys for program design and evaluation were provided by Jennifer Sime and Graeme Rodgers, both of the International Rescue Committee (IRC). While not prominent in the US context, surveys have been put to good use by the IRC, particularly during the COVID-19 pandemic, for configuring client-centered services for refugees and immigrants who are disproportionally at risk of infection and illness. Effective incorporation of systematic evidence in program design and evaluation is dependent on strong partnerships among stakeholders during refugee settlement and support.
Katie Mullins of Jesuit Refugee Services described the collection of data throughout the largely urbanbased programs of her organization, which serve forced migrants in over 50 countries of origin and transit. Data take a variety of forms, from quantitative tallies of services provided to refugees and displaced persons and families to more qualitative field notes. The impact of the COVID-19 pandemic has been to shift priorities of the organization towards building capacity among migrants and refugees, and as a result generating new forms of information about the lived experience of refugees and displaced persons. Ms. Mullins identified opportunities for utilizing the knowledge generated in the work of Jesuit Refugee Services, including expanding the analytic expertise of its staff, investing in comparisons among countries and cities in which they operate, and sharing, through publication, the empirical record of their work in addressing the needs of forcibly displaced persons and populations.

Implications for the Role of Research in Refugee Policy and Programs of Refugee Resettlement in the United States
Even prior to COVID-19, the world was in a period of increasingly closed borders and anti-immigrant sentiment in many receiving countries (Martin and Bergmann 2021). At first glance, the longer-term impacts of the pandemic on asylum and resettlement seem daunting and mostly negative. With borders shut, immigrant integration slowed in the pandemic, and the quality and efficacy of programs that serve immigrants declined.
Whether we measure immigrant integration through gains in income or English fluency, both likely deteriorated during the pandemic. Job loss, work from home, illness, and school closures contributed to stagnating progress. Pandemic-related lockdowns in local economies resulted in mass layoffs of immigrant workers, removing them from their workplaces which are a source of income and health coverage, as well as where many recent immigrants learn English. Furthermore, lockdowns, school closures, and fear of COVID-19 infection forced nonessential workers and their children and families to stay home, further reducing exposure to English and other learning and community involvement that tends to speed integration and improve living standards. Families overallnative and immigrant alike experienced reduced mobility and confinement that also led to increased incidence of mental health problems, alcohol and drug abuse, and domestic violence.
The immigrant experience in the pandemic was exacerbated, however, because many immigrant families were left out of federal stimulus programs and many unemployed immigrant workers were not able to qualify for expanded unemployment benefits. Hence, while the federal government rolled out a safety net for most low-and middle-income families, many immigrant families were excluded. At the same time, immigrant families suffered disproportionately from the pandemic, reflecting challenges of access to health care entitlements and legal/administrative vulnerabilities. Immigrants, particularly Hispanics, have sustained higher rates of serious illness and death from COVID-19 as compared with natives.
As the economic and health conditions among many immigrant families grew more precarious, the reach and impact of programs that serve immigrants were severely limited by the closing of offices and clinics and the transition from in-person meetings to digital encounters, efforts that were further thwarted by the lack of access to internet and appropriate technology in immigrant households. Public health clinics, rental assistance programs, legal aid services, and other programs were closed or made inaccessible by pandemic-related safety and health measures. Immigrant businesses were likewise less able to compete for resources through initiatives such as the federal Payroll Protection Program (PPP), which doled out hundreds of billions of dollars to small and medium-sized businesses. Immigrant-run businesses often lacked the banking connections, required documentation, access to technology and information necessary to apply and qualify for PPP funds.
At a time when US immigration reached new lows, reflecting both pandemic-related border closures and Trump administration crackdowns, US immigrants might have been expected to have greater economic opportunity, to integrate faster and have access to more resources and aid. The opposite happened. Among groups of low-income and recently arrived immigrants, including refugees, asylum seekers, and undocumented migrants, the pandemic ravaged both economic opportunities and assistance programs. As noted by Gruskin, trust in government among all groups of US migrants has likely deteriorated as well.
In considering the productive role of migration research within this context, a recurring message is the importance and value of partnerships in strengthening the interrelationships between research and evidence and programs supporting refugees and forced migrants. At least five recommendations emerged from the meeting: (i) engaging stakeholders in refugee resettlement processes in evidence-based program design and evaluation; (ii) forming partnerships and fostering trust within communities to reach refugee and migrant household during the pandemic and to address challenges in public policy; (iii) incorporating community-based data within forecasts and models of forced migration and displacement; (iv) developing models and programs that are responsive to knowledge, experience and priorities within refugee and migrant communities; and (v) appreciating the role of statistical and demographic information as well as qualitative data within program implementation at all levels of an organization's operations, from on-the-ground operations to international planning.
As we anticipate a potential renewal of migration flows post-COVID and a new era of refugee resettlement in the United States (and globally), having academics and practitioners recommit to fostering partnerships and to continuing those they have already built could go a long way towards new ideas in this area. Building on these partnerships going forward, as researchers and practitioners, we imagine the need and the opportunity for another space (both real and virtual) for collaboration. This space should be dedicated to bringing together various actors to purse the themes of research, program development, and policy that have emerged from the workshops and the excellent presentations by our colleagues in this session. Whether it is an organization, an institution, or a center to foster research, it must aim to enhance the process of building and knowing our audience and partners, and it should also focus on collaborative, interactive training of practitioners in migration research and the collection, analysis, and interpretation of data. As such, it must be more multi-, inter-, and trans-disciplinary, more integrated, and holistic, more refugee-and migrant-centered and inclusive, grounded in the aspirations and wisdom of migrant and host communities, and generative of informed and humane policy.