3

Analytical and Conceptual Issues

ANALYTICAL CHALLENGES

Ellen Percy Kraly (Malmö University and Colgate University) identified six analytical challenges related to forced migration that have implications for research design, measurement, and data collection:

1. Building analytic flexibility and responsiveness into population information systems to adapt to changing parameters of migration and emerging processes and patterns of human mobility, migration, and displacement.
2. Incorporating the historical-structural and environmental context(s) of human mobility, migration, and displacement.

3. Recognizing interactions between and among drivers of migration and mobility decision making and behaviors, along with social and cultural structures and processes of difference.
4. Moving beyond cause-and-effect analysis to understandings of patterns of relationships among and between drivers of migration while considering the full range and diversity of motivations, behaviors, and outcomes.
5. Understanding changes in mobility motivations and behaviors and how they change over time and across place and space while adopting life-course perspectives on human development and security.
6. Employing social theory as well as formal and social demographic analysis, including linkages between structure and agency.

These challenges reflect the perspective of academic researchers, but researchers constitute just one stakeholder group, she said. Kraly described work she has been doing with Bela Hovy¹ on the Global Compact for Safe, Orderly and Regular Migration² and the Global Compact on Refugees.³ Issues of data and research are embedded in both of the global compacts, which were the product of 2 years of consultation and process that involved multiple voices and stakeholders.

Kraly underscored the need for analytic flexibility in data collection to support measurement that is consistent with research, policy, and programmatic questions and analysis. This requires analyses that are contextual, comparative, and differentiated along social, economic, cultural, and spatial dimensions, including dimensions of difference that have not yet been anticipated. Researchers need to draw on migration theory and research to collect data that complement or even counter data collected specifically for policy and programmatic purposes. Policy “irrelevant” data, drawing on the argument of Oliver Bakewell (2008), may indeed be critically illuminating of the causes and consequences of forced migration, she said.

CONCEPTUALIZING AND MEASURING “INTEGRATION”

Irene Bloemraad (University of California, Berkeley) discussed the concept of integration as an example of the way conceptualization influences measurement. Integration goes by many different terms, including assimilation, incorporation, and adaption, but these terms have somewhat different meanings, she noted. They are associated with different normative assumptions. They differ in the extent to which they connote homogenization, in

___________________

¹ Chief, Migration Section, Department of Economic and Social Affairs (UN DESA).

² For more information, see https://refugeesmigrants.un.org/migration-compact.

³ For more information, see https://refugeesmigrants.un.org/refugees-compact.

which groups become less distinguishable, or the continued existence of difference. They involve models that tend to be more individualized or more group based.

One conceptualization of integration implies that people of migrant origins are indistinguishable from the native-born population of many generations. For example, they may vote at the same rate and have similar political choices, implying that they have become politically assimilated. A different model of incorporation might be that a group chooses not to be involved in electoral politics or cannot be involved in electoral politics because of legal barriers but is invited to participate in a political discussion of newcomer services. In that case, the group is incorporated into the political process but outside of formal electoral politics. Individuals of the group might not be completely integrated or assimilated, but the group itself could be incorporated into a larger community to some degree.

People have different normative judgments as to which form of integration is preferable, Bloemraad observed. For example, she asked, should people of migrant backgrounds be indistinguishable in terms of their wages and occupation from the native-born population, or is it preferable for members of the group to be working but concentrated in a particular occupation? People might be incorporated but not be assimilated, and both of those terms carry with them normative implications, she observed.

The U.S. literature on integration and assimilation embodies different concepts, which makes a difference for measurement. For example, some approaches to integration highlight the lack of differences between groups, which Bloemraad termed a gap approach. But this approach raises the question of which reference groups to use. In the United States, for example, they might be the non-Hispanic White population that has been native-born for many generations, or they might be native-born populations of a similar class background. A further consideration is whether and how to acknowledge differences in gender or age, she added.

Another approach does not involve reference groups. For example, the U.S. Government Accountability Office ([GAO], 2012) has defined immigrant integration as “a process that helps immigrants achieve self-sufficiency, political and civic involvement, and social inclusion.” Bloemraad termed this a process approach in which integration occurs over time and is headed in a particular direction. The problem with this approach is defining an endpoint, she noted. The GAO definition specifies one endpoint, but many others are possible and the endpoint applies only to the migrant group.

A third approach involves the idea of boundaries. Alba and Nee (2003) described a process in which “individuals’ ethnic origins become less and less relevant in relation to the members of another ethnic group . . . and individuals on both sides of the boundary see themselves more and more as alike. . . .Yet assimilation . . . does not require the disappearance of

ethnicity; and the individuals undergoing it may still bear a number of ethnic markers.” In this case, both the migrant group and the receiving society are changing and becoming more alike, even though immigration research often focuses on immigrants and not on changes in the native population. A specific example involves health in the United States, where immigrants on average have healthier outcomes than similarly situated native-born populations. Bloemraad said that one could imagine an integration framework where it would be desirable for the native-born population to move up to the immigrants’ healthy state, rather than what is the empirical reality—the immigrants moving down to the native-born population’s health status.

Finally, a fourth approach involves discontinuous trajectories. This approach could examine whether migrants are in a better situation than they were in their homeland or than they would have been if they had stayed in the homeland. In some contexts, she said, this may be more important than their situations in a receiving society.

Bloemraad noted she was a member of the committee that produced the report The Integration of Immigrants into the United States (NASEM, 2015), which used three concepts of integration. First, it defined integration as the process by which members of the immigrant groups and host societies come to resemble one another, which embodies the gap approach. Second, it examined well-being, which is more of a process approach. Third, it took a boundary approach by observing that integration is a two-way process in which native-born Americans also change in response to immigration over time.

Adopting multiple approaches raises challenges for exactly what to study, Bloemraad said. A gap approach might ask whether the average annual earnings of foreign-born workers are the same as that of native-born workers after controlling for occupation, education, race, gender, and so on. A process approach might ask whether migrants use public benefits when they first arrive but not as much over time. A boundary approach might ask what factors are involved in considering someone truly American. In each case, said Bloemraad, different sets of data will be needed.

Whom to study raises another set of questions, she said, such as whether to study integration in the first or second generation and immediately, or over a lifetime. In some countries, people born in that country are automatically citizens, whereas refugee status can perpetuate across generations in other countries, with quite different effects on integration. Defining which areas of life to study, such as economic, social, civic, cultural, or political, is another decision to make. Studying health outcomes can be less fraught than studying social integration, she said, since people agree that good health is important but differ on preferences for social integration.

A specific example involves language. People have different normative judgments about whether or not immigrants learn to speak a new language,

Bloemraad said. Many people believe that migrants to the United States master English less than in the past, but the data demonstrate otherwise (see Figure 3-1). Nevertheless, this indicator reflects values, since multilingualism rather than English dominance might be a positive in some circumstances, as it is in Canada.

The National Academies of Sciences, Engineering, and Medicine (2015) report used a large number of indicators of integration (see Figure 3-2), some of them controversial, such as the extent to which children were living in two-parent households, Bloemraad said. Other groups also have developed scales of integration using multiple indicators, though many academics are reluctant to do this. One reason for the reluctance is the need to agree on the components of well-being and then measure those components equally across societies or contexts, which can be difficult to do.

Social scientists have shifted in their view of integration, Bloemraad noted. In the past, integration or assimilation was seen as moving toward American cultural values, even if these values were sometimes left undefined. More recent scholarship has focused on the aspects of immigrant cultural values that are protective or may give immigrants an edge over the native-born population. In addition, scholarship has tended to focus on individuals’ attributes, especially human capital and more recently legal status. But immigrants bring many other attributes to the receiving society, such as ethnic economies or migrant cultures.

In describing research needs, Bloemraad noted that more attention needs to be paid to mid-level institutions, such as schools, religious institutions,

unions, and other organizations. In addition, many receiving countries have federal systems that differ among provinces, state, or cities. Trying to figure out levels of analysis among local, national, and international contexts gets very difficult, Bloemraad indicated.

She suggested another research need is better understanding of the macrostructures of receiving societies. Questions include how integration can best be understood in societies that have particular race relations; the role of culture in shaping integration; the effects of legal status on the second generation; and the role of human capital or class, especially compared with factors such as race, and the intersection of these macrostructures. She also suggested the need to understand whether theories developed to explain integration in the United States work elsewhere.

During the general discussion, Holly E. Reed (Queens College, City University of New York) observed that migrants exert their agency in making basic choices about migration. For example, in narrative interviews on Staten Island with the refugee community from Liberia, every refugee reported that he or she had made a choice to go to a camp in Ghana rather than Guinea because that would provide a better chance of resettlement to a third country. This, too, reflects a selection process among potential migrants. Reed noted the lack of knowledge about who stayed behind or

who fled elsewhere; thus, the selection issue is very important to consider. Bloemraad observed that this selection process has implications for policy. For example, success for one group of immigrants may not mean that other groups of immigrants will do as well because of selection processes in the future. In addition to the effects of selection, changing government policies also complicate the task of getting comparative statistics, she pointed out. For example, researchers would like to know whether migrants to Canada who enter through public-private partnerships fare better or worse than those resettled by the government, but selection affects who becomes a government-resettled refugee versus who becomes, for example, a church-sponsored refugee.

Pia Orrenius (Federal Reserve Bank of Dallas) noted that one problem with the scales of assimilation developed by some scholars is that positive outcomes—for example, the health advantages of Hispanic immigrants—can be seen as reflecting a lack of integration. Similarly, legal restrictions on undocumented immigrants can make it impossible to assimilate in the ways a scale might indicate is desirable. A broader question, she added, is whether assimilation is a desirable outcome; since economists believe that the greatest value of exchange is between dissimilar groups, there are no gains from immigration if people are identical. Bloemraad agreed with Orrenius that many people immediately think about integration in terms of diminishing a gap, and that researchers need to think about the extent to which differences should be regarded as positive. However, she noted that differences can also evolve over generations, and she is not convinced that the benefits of having an immigrant generation with distinct skill sets complementary to the native-born population continues to be desirable for the second generation. Bloemraad said the discussion raises questions of whether integration should be considered differently from the first to the second generation, or in the first 10 years of resettlement versus a later 10-year period.

This page intentionally left blank.