Boosting Refugee Outcomes: Evidence from Policy ... - WordPress.com

Boosting Refugee Outcomes: Evidence from Policy, Academia, and Social Innovation Salma Mousa* June 12, 2018

Abstract However measured, refugees are likely to face common barriers towards achieving integration in the West, including language proficiency, difficulty finding a job commensurate with their education and skills, and mental health stressors. This review draws on policy reports and academic studies (descriptive and experimental) to first answer two questions: (1) what do we know about refugee outcomes? and; (2) what factors are associated with these outcomes? I synthesize information on the individual and environmental traits associated with the socio-economic well-being of refugees – ranging from country of origin and gender to ethnic enclaves and rigid labor markets. I then survey pilots and programs aimed at integrating refugees globally, with a focus on the U.S., Canada, and Europe. The evidence base suggests that programs leveraging community support while supplementing income – such as apprenticeships, private sponsorship, and cash transfers dovetailed with financial mentorship – represent promising paths forward.

* Department

of Political Science, Stanford University; [email protected]. I thank Grant Gordon, Jens Hainmueller, Duncan Lawrence, Jeremy Weinstein for their invaluable feedback and guidance.

1

Contents 1 Introduction

3

2 What Do We Know About Refugee Outcomes? 2.1 Economic outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Education outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 3 5

3 What Explains These Outcomes? 3.1 Geographic Assignment . . . . . . 3.2 Delayed processing times . . . . . 3.3 Health . . . . . . . . . . . . . . . 3.4 Financial Literacy . . . . . . . . . 3.5 Ethnic Enclaves . . . . . . . . . . 3.6 Social Ties with Local Community 3.7 Labor Markets . . . . . . . . . . . 4 Which Approaches Seem Promising? 4.1 Outsourcing Services to Volunteers 4.2 Mentorship . . . . . . . . . . . . . 4.3 Apprenticeships . . . . . . . . . . 4.4 Private Sponsorship . . . . . . . . 4.5 Direct Cash Transfers . . . . . . . 4.6 ESL Education . . . . . . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . .

. . . . . . .

. . . . . .

5 Survey of Pilots, Programs, and RCTs 5.1 Pre-departure Preparation . . . . . . . . . . . . 5.2 Processing, Arrival, and Community Engagement 5.3 Training, Re-credentialing, and Employment . . . 5.4 Financial Literacy . . . . . . . . . . . . . . . . 5.5 Micro-finance and Entrepreneurship . . . . . . . 5.6 E.U. Best Practices . . . . . . . . . . . . . . . . 5.7 Canadian Best Practices . . . . . . . . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . . .

7 7 8 9 9 10 11 12

. . . . . .

12 12 14 16 19 22 23

. . . . . . .

25 26 27 29 30 31 31 31

6 Paths Forward and Conclusions

32

7 Appendix 7.1 What motivates volunteers? . . . . . . . . . . . . . 7.2 Re-credentialing Programs in the U.S. . . . . . . . . 7.3 Apprenticeship Alternative: Supported Employment 7.4 Apps for Adult English Learners . . . . . . . . . . .

35 35 36 37 38

2

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

1

Introduction

Often fleeing violence and spending years in under-resourced and over-crowded refugee camps, the 22 million asylum seekers today are among the world’s most vulnerable communities. Unlike other migrants, refugees do not leave their homes voluntarily. The migration experience of refugees frequently takes a toll on mental and physical health even before arriving in the receiving country (Connor 2010). With few social ties and resources, they must then adapt to a foreign language, labor market, and culture – often remaining unemployed, under-employed, or otherwise low-income for years after arrival (Bansak et al. 2018). Despite the magnitude of these barriers, the scale of the challenge (the world is currently experiencing the highest levels of displacement on record), and the politicization of refugee admittance policy in the U.S. and Europe, we know little about how refugees fare once resettled. A cash-strapped U.S. Refugee Resettlement Program further compels a systematic analysis of refugee outcomes and best practices worldwide in order to maximize resettlement agency dollars.1 This document aims to compile and analyze the evidence base on refugee outcomes – and the factors that shape them – globally. However measured, refugees are likely to face common barriers towards achieving integration in the West, including: a lack of language proficiency, difficulty finding a job commensurate with their education and skills, having foreign credentials recognized, and weak social ties. This review draws on policy reports, state-collected survey data, and academic studies to first answer two questions: (1) what do we know about refugee resettlement outcomes? and; (2) what factors are associated with these outcomes? I then highlight emerging best practices, programs, and promising pilots aimed at resettling and integrating refugees worldwide, with a focus on leveraging community engagement. This is a challenging endeavor given the lack of “ground truth” data on refugees. For instance, while a useful and creative first step among migration scholars, most studies on refugees in the U.S. rely on imputing refugee status from surveys that have yet to be externally validated with a known refugee population.2 The trends I highlight here, especially those from the U.S., should be therefore taken as suggestive. I primarily consider the refugee-specific literature but note where programs or studies instead focus on other groups, such as immigrants, that may nevertheless be informative.

2

What Do We Know About Refugee Outcomes?

2.1 Economic outcomes Governments typically view full-time employment as necessary for self-sufficiency, but approaches to encouraging employment vary across, and sometimes within, countries. In the U.S., the Bureau for Population, Refugees and Migration (PRM) contracts nine Voluntary Agencies (VolAgs) to resettle refugees vetted by the United Nations

1 At the current pace, the U.S. will fall far short of meeting it’s presidentially-set target of admitting 45,000 refugees – itself a target well below the Presidential Determination of 110,000 in 2017. At 20,000 estimated refugees, this number will be the lowest since the Federal Refugee Resettlement Program was created in 1980. The State Department similarly slashed funding to resettlement agencies and encouraged them to downsize in December 2017 (Torbati and Rosenberg 2017). 2 Studies like Evans and Fitzgerald (2017) and Fix, Hooper and Zong (2017) impute refugee status using country-year combinations in the American Community Survey. Refugee status is assigned to each country-year combination for which “refugee admissions in the DHS and WRAPS data exceed 40 % of the estimated foreign-born population identified in the ACS data” (Capps et al. 2015a). Without being externally validated against a representative dataset where refugee status is known, however, evidence that relies on imputation methods should be taken as suggestive but not definitive. For instance, Fix, Hooper and Zong (2017) apply this method using data from four U.S. states and conclude that initial location only weakly shapes outcomes. Yet initial state and municipal assignments at the time of arrival, as well as information on secondary migration, are unknown. The descriptive findings in these studies are important first steps in uncovering refugee outcomes, but more work is needed against verified “ground truth” data to understand the direction of imputational bias using this technique.

3

and U.S. Citizenship and Immigration Services (USCIS).3 Through some 250 VolAg offices across the country, these agencies greet arrivals at the airport, find and furnish suitable accommodation, and provide casework support for a median of 90 days.4 One of the few pieces of information that PRM tracks post-arrival is employment rates at the 90 and 180 day marks. VolAg funding is also tied to this metric – leading to an incentive structure that prioritizes fast employment. The “work–first” paradigm assumes that once marginalized groups join the labor force, cultural, psychological, and linguistic integration will follow. This approach implicitly values the speed rather than the quality of employment. The counterfactual of whether refugees would secure higher-quality jobs were they to delay employment (and perhaps invest in up-skilling) remains unclear. While the U.S. program is designed around fast employment, the “skills–first” Canadian and European models instead stress upfront investment in language skills and job training. We have little causal evidence on which approach is more successful for socially, economically, and linguistically integrating refugees. We can, however, compare employment rates and average earnings for some Western receiving countries. Refugees tend to find work fastest in the U.S., followed by Canada, continental Europe, and Scandinavia (Pressé and Thomson 2007). Half of U.S. refugees find work within eight months, with refugee men surpassing employment rates for U.S.-born men (67% vs. 62% percent) and refugee women equally as likely to work as U.S.born women (54%). Again, these figures may be upwardly biased due to non-representative data. Nevertheless, Burmese, Iraqi, and Somali men seem to have lower-than-native employment probabilities, but they are also the most recently settled (Capps et al. 2015b). We observe the same gender gap, overall high refugee employment rate (82% employed after five years), and weaker performance of Middle Eastern refugees in Canada (Salant and Benton 2017). Country of origin is thus strong determinant of occupation status for refugees, as it is for immigrants (Vigdor 2010). Shifting to Europe, only a quarter of refugees in Sweden find work after two years, and most in subsidized jobs (Desiderio 2016). We thus do not have sufficient information to determine, even descriptively, whether refugees are better off under the work–first model in the U.S., with its flexible labor market, relative to skills–first approach in a rigid European labor market. How long does it take refugees to converge on employment rates with other immigrant groups? Refugees take five years to reach parity with family reunification migrants in Canada, six years in the E.U. (50% employment, higher than family reunification migrants in France, Italy, Switzerland and Belgium), 15 years (65% employment) in Sweden, and 10 –15 years in Norway (60 % employment) (Capps et al. 2015b; Desiderio 2016). Convergence is slower when comparing refugees with economic migrants and natives: refugees take 15 years to reach 70% employment and converge with economic migrants in Europe (Desiderio 2016). Second-generation Australian refugees seem exceptional, overshooting the Australian average (and their parents) on almost all labor force measures, though without reaching native-level earnings (Hugo et al. 2011). Unfortunately, we know little about second-generation refugee outcomes elsewhere. Employment rates bely two related phenomena: low average earnings and the under-employment of refugees. Nearly half (44%) of U.S. refugees resettled between 2009 – 2011 were low-income, and their household income remained below the U.S. average even after 20 years. This is a recent phenomenon: no such income gap existed in 2000 (Capps et al. 2015b). Those that do find employment are overrepresented in unskilled low wage sectors like agriculture, construction, cleaning services, hotels and restaurants, and retail trade – around 25% of employed refugees work in skilled jobs in the U.S. relative to roughly 40% of other immigrants (de Matos and Liebig 2014; Connor 2010). Underemployment is sticky: the share of over-qualified refugees stays roughly constant at 8 weeks in the U.S. and 21 months (57% vs. 54%) (Cheung and Phillimore 2014). We observe the same phenomenon of 3 In

Europe, this role is usually filled by municipalities and state agencies, while Canada has both government-resettled and privately sponsored tracks. 4 Some states provide only 30 days of these “Reception and Placement” Services, while Utah provides the most at two years.

4

underemployment, low earnings, and overrepresentation in temporary and part-time work among refugee men and women in Canada, Australia, and Europe, even relative to other immigrant groups (Desiderio 2016; Immigration and Citizenship Canada 2016; Salant and Benton 2017; Hugo et al. 2011). Analyzing data on n = 445 refugees in the 2003 New Immigrant Survey (a sample of immigrants with permanent residency in the U.S.), Connor (2010) finds a persistent disparity in occupational level and earnings when compared to other migrants – nearly half of which cannot be explained by education, English skills, and neighborhood. Perhaps because baseline earnings are so low for refugees, the wage premium for duration in the host country and language proficiency is high. The median refugee household earns $31,000 more after 20 years in the U.S. compared to five years, and each additional year spent in the U.S. is associated with a 3.4% spike in the employment rate (Capps et al. 2015b; Beaman 2012). Employment rates among Canadian refugees double over time: from 20% employment at 6 months to 40% at 2 years (Immigration and Citizenship Canada 2016). Language proficiency is a strong predictor of earnings and employment opportunities, albeit with small marginal effects on permanent employment (Hooper, Desiderio and Salant 2017; Connor 2010; Cheung and Phillimore 2014). However, only 7% of refugees resettled in the U.S. between 2002 – 2013 report speaking “good” English (33% spoke “some”) and over half are “limited English proficient” even after 20 years in the U.S (Capps et al. 2015b; Desiderio 2016). Promisingly, refugees tend to be much younger than other migrant groups and more likely to stay in the host country – pointing to a large potential to be productive with the right training and education (Hugo et al. 2011). Lastly, a note on wealth. The wealth gap between natives and foreigners no longer converges after roughly two decades the way it did for cohorts who arrived in the 1990’s (Akresh 2011). Refugees seem to accumulate less wealth over time than immigrants (up to 75% less by estimates based on the 2003 New Immigrant Survey), but demonstrate higher patterns of home equity than other types of immigrants (Akresh 2011). Duration in the U.S., country of origin, and marital status matter: married immigrants increase their net worth by around $800/year with no comparable changes for single immigrants (including refugees). Age at arrival also matters – positive net worth declines by about 10% for each additional year of age upon arrival for all immigrants (including refugees). As expected, initial wealth is a strong predictor of future wealth, as is inter-marriage with natives. Unfortunately, education and work experience in the country of origin are discounted in the U.S., underscoring the dual challenges of underemployment and re-credentialing among refugees (Hao et al. 2001; Hao 2004; Akresh 2011).

Takeaways: A“refugee gap” in occupation status and average earnings persists even in countries where employment rates eventually converge with natives. Refugees who are employed, tend to be overqualified. Language proficiency, duration in the host country, and recognized education and work experience are strongly correlated with employment outcomes.

2.2 Education outcomes The literature on refugee educational attainment focuses on language proficiency and early childhood education – little is known about high school and tertiary performance. Host language proficiency is a strong predictor of earnings and employment opportunities for refugees across Europe and North America (Hooper, Desiderio and Salant 2017; Connor 2010; Cheung and Phillimore 2014). Yet an estimated 7% of refugees resettled in the U.S. between 2002 – 2013 report speaking “good” English (33% spoke “some”) and over half are “limited English proficient” even after 20 years in the U.S (Capps et al. 2015b; Desiderio 2016).5 Refugee responses from the 5 These

findings are based on imputing refugee status from the American Community Survey, so results should be taken as rough

estimations.

5

2003 New Immigrant Survey indicate that enumerator-reported English levels are also lower than those of other immigrants (2.7 vs. 3.0 on average, on a 4-point scale) (Connor 2010). Barriers around access to bilingual instruction, literacy, and linguistic distance from the mother tongue language are described in Section 4.6. Refugees are also likely to face additional challenges as adult language learners (Qureshi 2016). Perhaps due to linguistic barriers, refugee adults have an average half a year less of schooling, and half as much education undertaken in the U.S., relative to other immigrants (Connor 2010). Parental language proficiency is important for parental engagement in schooling and children’s language acquisition. English competence among young refugees, in turn, is associated with a reduced likelihood of depressive symptoms in young refugees (Fazel et al. 2012). Unfortunately, one-third of refugee children live in linguistically isolated households where no member over 14 is proficient in English (Hooper et al. 2016). An observational study in the U.K. points to direct access to English language support for high-schoolers, a welcoming school attitude, and the presence of specialist teachers as key for refugee educational attainment (Hek 2005). A positive finding here is that once resettled, refugee children tend to acquire conversational ability in the language of their new country faster than their parents (Zhou 2001). The linguistic, cognitive, and emotional benefits of high-quality ECECs (Early Childhood Education Centers) appear amplified for refugee children as dual-language learners and children whose families benefit from access to comprehensive services and early exposure to the U.S. education system (Hooper et al. 2016; Morland et al. 2016). Preschool education among immigrant children is associated with higher English proficiency, reading and math scores, and a 20% reduction in the immigrant-native achievement gap according to one study of Mexican immigrant families (n = 13,843) (Crosnoe 2007; Magnuson, Lahaie and Waldfogel 2006). While comparable to other immigrants, the pre-school enrollment rate of refugee children is well below rates for U.S. born children – except for refugees from Iran, Colombia, Russia, Haiti, and Vietnam (Hooper et al. 2016; Takanishi 2004; Crosnoe 2007; Bass, Shields and Behrman 2004). The variation in pre-school participation rates by ethnic groups may be partially be explained by parental beliefs and accessibility to affordable preschool programs (Magnuson, Lahaie and Waldfogel 2006). School and neighborhood quality matter for academic achievement. Immigrants assigned to high-quality high schools with more native speakers do better, while those enrolled in lower-quality pre-schools do not gain the maximum language and cognitive benefits possible (Cattaneo and Wolter 2012; Gould, Lavy and Paserman 2004; Leseman 2007). Community quality can exacerbate negative effects of co-ethnicity. Leveraging randomized housing assignments in Sweden, Åslund et al. (2011) find that a standard deviation increase in the fraction of highly educated co-ethnics in the local community raises the probability of graduating from upper-secondary school on time by 2.4 percentage points and improves student performance by 0.8 percentile ranks – closing the achievement gap between immigrant and native boys by 6 – 9%. The size of the (educated) ethnic community is more important for boys and for children whose parents are less educated, two groups that have the poorest school outcomes. However, immigrant families with academically low-performing children tend to move to ethnic enclaves with lower fractions of educated members. This sorting bias of immigrants with “unfavorable” characteristics selfselecting into enclaves is so severe that co-ethnic concentration is associated with lower high school graduation rates in Sweden and 18% lower annual earnings on average in Denmark (Åslund et al. 2011; Damm 2009). Refugee children face several risk factors in educational attainment, but also have some advantages. Refugee children are more likely than other children to live in a two-parent family, whose dual income can compensate for below-average earnings among refugees (Hooper et al. 2016). The presence of two parents and extended family members also implies more adults available for child supervision and guidance. Refugee parents are also better educated than immigrant parents but not American parents (one-third of refugee parents have a college education

6

and over four-fifths have a high school degree according to Hooper et al. (2016)). Refugee parents from Iran, Russia, and Armenia are the most likely to have completed a four-year college degree or higher – well above the U.S. average.6 We have scant evidence of the effect of early education interventions on future employment. A 2017 study suggests that age of arrival matters: refugees who enter the U.S. before age 14 appear to graduate high school and enter college at the same rates as natives. Conversely, those who enter as older teenagers seem to have lower attainment potentially because of language barriers and status as unaccompanied minors (Evans and Fitzgerald 2017). While these findings align with evidence on the importance of age at arrival, one should again interpret these results cautiously given the difficulty of accurately inferring refugee status from non-representative administrative data.

Takeaways: Access to early education centers (e.g. pre-schools) significantly boosts language proficiency, socialization, and cognitive development in immigrant children, but they are less likely to attend (high-quality) pre-schools. School quality and fraction of language-proficient classmates improves educational achievement. Sorting effects lead to ethnic enclaves being associated with worse learning outcomes. Little is known about high school or university achievement among refugees, although years of schooling seem lower than other migrants.

3

What Explains These Outcomes?

Factors that inhibit economic and social integration include: geographic assignment, transit and processing time, temporary and insecure housing especially when located away from centers of economic activity, mental health stressors, financial products unsuitable for refugee-specific needs, a lack of referrals and accreditation documents, and the absence of trusted social networks.

3.1 Geographic Assignment Three sets of factors affect the employment probability of refugees: geographic setting, personal traits, and the synergies between the two. Location fixed effects are known to be highly important for economic integration (Connor 2010; Bansak et al. 2018). All else equal, some cities provide better economic and network opportunities that can boost employment. Likewise, some individual characteristics, like language and educational attainment, are associated with high employment rates regardless of resettlement location. Lastly, there are interaction effects between places and people. Certain traits render a refugee a better match for a particular city. For instance, Francophone refugees are more likely to be employed when assigned to French-speaking cantons as opposed to German-speaking cantons. The algorithm developed by Bansak et al. (2018) learns just such synergies and provides estimates for returns to employment given a particular refugee case, the set of available resettlement locations, and agency-specific capacity constraints. Running the model on historic data suggested increases in employment rates by around 40% in the U.S. and 75% in Switzerland, with no losses for any refugee cases. The algorithm is set to be tested against the status quo assignment system using future arrivals in Switzerland in 2018. If realized gains align with the back-tested predictions, the use of data-driven techniques to as a key input in the placement process could change the resettlement pipeline. 6 Gender

patterns of parental education vary significantly: fathers from some origins (e.g., Ethiopia, Sudan, Afghanistan, Cambodia, Liberia, and Somalia) are much better educated than mothers, while those from Ukraine, Yugoslavia, Armenia, Bosnia, Cuba, Russia, Ukraine are less educated than their wives.

7

Once assigned to a city, refugees often struggle to find affordable and well-located housing, especially in the absence of pre-existing social and family networks. Several European countries, like Switzerland, Denmark, and Sweden, randomly assign refugees to a province or municipality to avoid overburdening any one locality when resettling refugees (Åslund et al. 2011; Foged and Peri 2016; Hainmueller, Hangartner and Lawrence 2016a). The U.S. and Canada take social networks into account (Beirens and Fratzke 2017). In both cases, however, we see secondary movement that is both hard to prevent and to predict with accuracy. We do know that there are often spatial mismatches between service providers concentrated in multi-ethnic, downtown areas on the one hand and refugees (especially Government Assisted Refugees, or “GARs”) who settle outside these areas in suburban, rural, or ethnic communities on the other (Pressé and Thomson 2007). It is thus unsurprising that 11% of GARs in a Canadian government survey (n = 541) moved in the first year and sought services from another resettlement assistance office (known as Service Provider Organizations). Of the GARs who indicated that they moved, 40% indicated it was to find employment, 22% indicated it was to be closer to friends, and 19% indicated it was to be closer to family (Immigration and Citizenship Canada 2016). While some refugees can rely on sponsors for information on appropriate housing – and more often than not, a place to stay for free – government-assisted refugees struggle to find adequate and affordable housing without this support (Hyndman 2017). Housing considered affordable for refugees is often located in low-quality neighborhoods far from public transportation and co-ethnic networks (Connor 2010), and may be perceived as disreputable by new arrivals in stress-inducing ways. GARs in particular struggle to pay off transportation loans, which may correlate with geographic isolation (Mousa 2017a). On the other end of the rent spectrum, large families are susceptible to being housed in unaffordable lodging – thus starting their life in the U.S. in debt. Often with only days or weeks notice before a travel date is set, VolAgs are forced to spend part of refugees’ “welcome money” (a federal stipend proportional to family size) on hotels (Mousa 2017a). Regardless of what drives secondary migration, frequent residence changes (especially for boys) can worsen mental health outcomes for children, although there is no discernible difference in living in an urban vs. rural area on adult mental health (Fazel et al. 2012).

Takeaways: Geographic assignment is one of the strongest determinants of economic integration. Placement algorithms that combine refugee traits with location fixed effects to optimize for employment are encouraging. Finding well-located and affordable housing is arduous for many governmentassisted refugees. Initial housing away from economic hubs, service providers, and (co-national) social networks may encourage secondary migration within the host country.

3.2 Delayed processing times The lives of refugees are often put on hold for months or years before obtaining a decision on their refugee status. This waiting period is often de-motivating, depressing, and unproductive (Mestheneos and Ioannidi 2002). Analyzing refugees in Switzerland between 1994 and 2004, a causally identified study finds that one additional year of waiting reduces the subsequent employment rate by 4 to 5 percentage points, a 16 – 23% drop compared to the average rate (Hainmueller, Hangartner and Lawrence 2016b). The results are consistent with the idea that the effect of wait time is mostly driven by psychological mechanisms rather than skill atrophy, although more work is needed to investigate the effects of refugees’ skill depreciation over time. Unlike Europe, wait times for U.S.-bound refugees take place before arrival over a (minimum) two year processing period. Long wait times also affect host populations: the motivation and interest of private sponsors in Canada greatly decreases with processing times (Hyndman 2017). While urging quick registration times, both the E.U. and the Canadian government recommend providing refugees with cultural orientation courses, language and soft skills training for educated refugees, and

8

job-matching and referral services to use wait times productively (Yu, Ouellet and Warmington 2007; Sommarribas et al. 2016).

3.3 Health The majority of refugee-related health studies concern mental health, a strong correlate of social and economic integration. In a 2001 study (n = 1,364), job experience in Canada was the strongest predictor of employment for men whereas depression proved an important predictor of employability for women (Hou and Beiser 2006). Refugees resettled in the West are estimated to be ten times more likely to have post-traumatic stress disorder than age-matched general populations in those countries (a prevalence of 9 – 11%) and double as likely to suffer depression and anxiety as labor migrants (44% vs. 20% and 40% vs. 21%) according to a n = 24,051 meta-analysis (Fazel, Wheeler and Danesh 2005). Higher GNP in the host country is related to lower rates of depression and anxiety in labor migrants but not in refugees (Fazel, Wheeler and Danesh 2005). Adult refugees also appear reluctant to report mental and somatic conditions to resettlement authorities in the pre-departure period, and to medical officials in the post-departure period, despite high rates of insurance coverage among some communities (Sommarribas et al. 2016; Taylor et al. 2014). On the other hand, encouraged by favorable eligibility rules, children of refugees participate in public health programs at higher rates than immigrant children, although they are also more likely to be poor (Hooper et al. 2016). The poorest refugees have the highest rates of benefit use, though the range of benefits varies by state (Hooper et al. 2016).

Takeaways: Refugees are high-risk for depression, anxiety, and PTSD, which negatively impacts employability. Refugee children access health services at high rates, but their parents are more reluctant.

3.4 Financial Literacy Refugees face specific financial difficulties relative to other migrants that can be compounded by the urgency of their situation, a lack of social support and knowledge about the host country, and limited opportunities to liaise with family members. These difficulties include a lack of documentation (including proof of address), language barriers, a lack of trust in the financial sector, and a high opportunity cost for shift workers to visit with banks or service providers (Atkinson and Messy 2013; Orozco 2015). Among immigrants, this disadvantage appears to be passed on to children: OECD PISA research indicates that students with an immigrant background have lower levels of financial literacy than other students even when comparing across students who speak the same language at home and have a similar socio-economic background, math and reading performance (PISA 2012). Even if most transactions are undertaken in cash, refugees may need support to know how to: calculate exchange rates, budget and make effective use of financial resources, keep records of money sent and received, and smooth irregular cash flows. Refugees may not be aware of the fees they are paying to make money transfers even when sending cash through informal channels (OECD 2016). Most work on the financial literacy of refugees focuses on receiving countries in the Global South. In this context, refugees often require a mix of services including emergency loans, grants, insurance and educational input (Azorbo 2011). Even then, refugees have low rates of running sustainable businesses and repaying loans (EastonCalabria and Omata 2016). Refugees interacting closely with NGOs have also been found to confuse loans with grants (Nourse 2003). The UNHCR rolled out a more incremental approach beginning with financial literacy and ending with micro-loans (“the Graduation Approach”) with only short-term positive results as of 2013 (Egypt 2013). Nonetheless, traditional access to credit could be beneficial for some refugee groups – such as older adults 9

who ran small businesses in their countries of origin resettled in economically developed countries (Hooper, Desiderio and Salant 2017). In the U.S., a car loan program benefiting around 200 refugees settled with the International Rescue Committee in San Diego, CA led to a default rate of 1.95% – below the national average for subprime auto loans at 11.96% – in 2017 (Morrissey 2017). The program’s success relies on low interest rates (relative to subprime lenders), mandatory financial education courses, regular budgeting sessions with a financial counselor, and earned income verification, suggesting that loan programs might be most successful when paired with financial literacy support and when targeted at full-time employees. Scholars and international organizations (including the UNHCR) are shifting to micro-savings – the only sure way to secure future assets – rather than micro-finance to tackle financial literacy among migrants and refugees. Financial literacy interventions improve savings outcomes for the most vulnerable migrants, while the less vulnerable develop micro-savings and lending groups within their own communities (Seshan and Yang 2014; Doi, McKenzie and Zia 2014; Ashraf et al. 2015; Easton-Calabria and Omata 2016). Financially literate refugees, however, would likely still benefit from savings products, electronic payment facilities and access to credit and insurance, among other products (OECD 2016). Accessible information on building a credit history and starting a business seem important, especially given qualitative research in the U.S. reporting that refugee entrepreneurs cite complex legal obstacles as a reason for not having started a business sooner (Nibbs 2016).

Takeaways: Refugee-specific hurdles to financial literacy and good savings practices are intensified among low-skilled, vulnerable sub-groups – and these groups are the most likely to benefit from financial literacy interventions. Financial literacy support coupled with low-interest loans tied to particular assets could benefit high skilled U.S. refugees or those with a business ownership background. The micro-savings realm is promising for both low- and high-skilled refugees.

3.5 Ethnic Enclaves Around 60% of refugees have pre-existing ties in the U.S. The network-creating effects of common nationality among immigrants can be informative in understanding the potential use of these ties (Bandiera, Barankay and Rasul 2008).7 Ethnic enclaves can improve economic integration through three channels: 1. Disseminating job information, which increases job-worker match quality and thus the hourly wage rate (Damm 2009; Topa 2001; Beaman 2012); 2. Ethnic trade or overrepresentation in a particular sector of the economy (Bevelander and Lundh 2007; Alesina, Harnoss and Rapoport 2016; Ottaviano and Peri 2006); 3. Lower exposure to discrimination than in the mainstream economy (Portes and Bach 1985; Waldinger and Lichter 2003). There is no scholarly consensus on the effect of enclaves, perhaps because effects are highly contextual. Positive labor market effects seem to be concentrated among two groups in particular: (1) low-skilled, low-income, and low-education refugees and immigrants (who are more likely to move into enclaves than their high-capital compatriots according to Borjas (1998)); and (2) refugees settled in enclaves who arrive separately from large co-national cohorts. Experimental evidence from Scandinavia suggests that a standard deviation increase in ethnic concentration increases earnings for low-skilled immigrants by 13% while living in a low-quality enclave depresses earnings 7 The

strength of these ties varies widely, ranging from first-order relatives that can provide housing and employment support, to superficial acquaintances little more than a nominal contact on asylum documents.

10

for co-ethnic migrants in both Sweden and Denmark (Edin, Fredriksson and Åslund 2003; Damm 2009). Lowskilled immigrants seem to have an easier time integrating economically in “escalator regions” like big cities but with lower returns than native-borns (Andersson 1996; Connor 2010; Phillimore and Goodson 2006). On the socio-psychological front, positive enclave effects may be driven in part by the positive effect of living and socializing with co-ethnics on psychological functioning (Fazel et al. 2012). The competition produced by arriving at the same time as a large co-national cohort may reduce employment rates and earnings. A study of 1,700 male adult male refugees resettled by the IRC between 2001 and 2005 finds that a one standard deviation increase in the number of co-nationals who arrive in the U.S. one year prior to a newly arrived refugee lowers her probability of being employed by 4.8 percentage-points and average hourly earnings by $0.82. The negative competition effect more than offsets the positive effects of an additional year of residence (Beaman 2012). Conversely, an increase in the number of established network members (arrived 3 – 5 years prior) increases the employment probabilities and average earnings of recently arrived refugees by the same amount (Beaman 2012). The results suggest that by providing job information, networks affect wages both through the employment rate and job quality. In other cases, however, negative enclave effects have been observed (Portes and Zhou 1993). Knowledge transfer may also disseminate information detrimental to labor market success, like welfare eligibility (Bertrand, Luttmer and Mullainathan 2000).

Takeaways: The causal effects of ethnic enclaves on employment and other integration metrics is unclear, but suggestive evidence points to a positive effect for low-skilled refugees while high-skilled refugees tend to move away from enclaves. The portion of employed and educated enclave residents appears to shape educational attainment and employment outcomes for new arrivals.

3.6 Social Ties with Local Community Social contacts – not necessarily with co-ethnics – and a welcoming reception have been found to improve integration and well-being across a variety of settings (Ager and Strang 2008). A study of 6,538 Mexican immigrants in the U.S. found that the same individual is more likely to be employed and hold a higher paying nonagricultural job when her network is exogenously larger (Munshi 2003). Living in close proximity to family and friends decreased the quality of employment for refugees in Alberta, Canada, but those who sought familial and extra-familial aid in finding a job found better quality employment than their peers who did not have such networks (Salant and Benton 2017). Social networks can help refugees access functional resources like help with employment and housing and strengthen English proficiency (Phillimore and Goodson 2006; Cheung and Phillimore 2014). The economic disadvantage of having no social contacts fades after 21 months in the U.S., which may explain why refugee participation in religious or cultural groups is initially high but drops off quickly (Hugo et al. 2011; Cheung and Phillimore 2014). A strong connection with one’s neighborhood and social support in the host country are associated with lower rates of depression in refugees (Fazel et al. 2012; Nazzal et al. 2014). It is unclear how ties with natives shape outcomes relative to (non-native) co-ethnic ties. It appears, at least in Canada, that refugees socialize most with co-ethnics and extended families (47% and 26%) rather than native-born friends (26%) or even sponsors (7%). (Salant and Benton 2017). One concern is that the emphasis on rapid selfsufficiency in North America undermines investments in building personal connections with community members or learning English (Beirens and Fratzke 2017). Relatedly, the rates of refugee participation in civic or community organizations in the U.S. was low enough in 2017 to prompt one VolAg to cease measuring this outcome at the two-year mark (Mousa 2017a).

11

Takeaways: Social connections with neighbors, community members, and local organizations have positive effects on language acquisition, psychological well-being, and securing jobs and housing. How ties with natives vs. co-ethnics differentially shape outcomes is unknown.

3.7 Labor Markets At least three structural factors impede refugee employment outcomes. First, high minimum wages, extensive job protections, and small informal sectors in European welfare regimes are prohibitive barriers to refugee employment. A Swedish study (n = 350,000 refugees) found that a 1% increase in the minimum wage increases the length of unemployment for refugees by 3.5 days on average as well as raising the unemployment rate by 0.33%. The effect for age-matched young natives is half as large (Lundborg and Skedinger 2014). Second, even if firms are willing to hire refugees, they face significant hurdles including navigating complex immigration and labor legislation and understanding the value of foreign-acquired qualifications and skills. These barriers are often insurmountable for small firms without dedicated human resources departments or experience in hiring foreigners (Hooper, Desiderio and Salant 2017). There is a pressing need to streamline the process of recognizing foreign credentials for refugees, especially in light of encouraging evidence that immigrants who have their foreign degrees recognized in the U.S. and Europe are more likely to find well-matched jobs even after accounting for the origin of the qualifications and the field of study (de Matos and Liebig 2014; Halpern 2008). Third, discrimination in Western labor markets is pervasive, as evidenced by myriad experiments comparing resume call-back rates for applicants with ethnic vs. native-sounding names (Riach and Rich 2006; Kaas and Manger 2012; Blommaert, Coenders and Van Tubergen 2014; Bertrand and Mullainathan 2004; Adida, Laitin and Valfort 2016). Discrimination and prejudice are closely tied to the nature of inter-group contact (Enos 2017; Paluck, Green and Green 2017; Pettigrew and Tropp 2006; Allport 1979). Causal evidence from Greece and Austria suggests that contact with refugees under optimal conditions (uncompetitive, frequent, positive) reduces votes for far-right, anti-refugee parties, while brief exposure at the border can increase it (Hangartner et al. 2017; Steinmayr 2016). Moreover, Hofinger and Hoser (2017) find that 68% of Austrians in municipalities that received large numbers of refugees believe that hosting refugees “worked well” or “very well” in their municipality, while roughly half of the population report interacting with immigrants at least several times per month (n = 704). A prediction follows from these results: discrimination is most likely to affect refugee integration where meaningful, positive, and frequent contact with refugees is lowest.

Takeaways: Rigid labor markets, difficulties in recognizing foreign credentials (especially without documentation), and discrimination against foreigners and ethnic minorities all serve as obstacles to employment for refugees. Frequent, positive contact with refugees likely lowers anti-refugee hostility.

4

Which Approaches Seem Promising?

4.1 Outsourcing Services to Volunteers Funding cuts and an on-going migration crisis have increased the need for, and enthusiasm of, volunteers working in the refugee space. Volunteers have stepped up globally: German refugee organizations saw volunteer enrollment increase by 70% and British voluntary organizations historically bear the brunt of state withdrawal of support for 12

asylum seekers (Karakayali and Kleist 2015; Patel and Kerrigan 2004). VolAgs in the U.S. similarly saw increases in volunteer registrations (although not all are approved) by as much as 200% (Mousa 2017a). The surge in volunteerism in response to the Syrian fall-out suggests that engagement is geared toward helping refugees rather than volunteering in itself (Karakayali and Kleist 2016). Evidence from the U.S., Canada, and Germany suggests that volunteers tend to be predominantly female and in their twenties or over 60 (Government 2009). In Germany, they are typically of migrant background and non-religious relative to the societal average (Karakayali and Kleist 2015). What kinds of tasks do volunteers take on, and are they effective? Volunteers in the refugee space globally tend to begin with a limited, logistical mandate that gradually expands to include personal, essential care. German volunteers started out by facilitating visits and communication with officials, and offering translation and language lessons, and then began to donate and distribute essentials (including food and clothing) in lieu of the state according to a survey of 466 volunteers and 70 organizations (Karakayali and Kleist 2015). Syrian volunteers working with Iraqi refugees in Damascus identified vulnerable refugees, provided information on services available to them, and assisted with accessing these services by facilitating appointments. After becoming engaged, volunteers began offering space in their own houses, conducting regular home visits, sharing their own food and clothes, and handling emergencies. Leveraging the opportunity to build trust with refugees (Mccarthy and HaithCooper 2013), the UNHCR then trained volunteers on issues like gender-based violence, self-reliance, first aid, and psychosocial issues (Mirghani 2013). Volunteers are effective at alleviating the burden from agencies: some U.S. resettlement agency offices outsource 40% of furnishing requirements to volunteers, the German state now intentionally relies on volunteers in some areas, and Syrian volunteers assisted an average of 6,000 cases per month between 2003 – 2008.8 Many governments fund programs to match volunteers with new arrivals (Table 1) to cover basic, short-term needs (e.g. driving to appointments) in addition to long-term social adjustment (e.g. job search help, language practice). The literature on how best to motivate volunteers is reviewed in the Appendix (section 7.1). Clear mission statements, input in decision-making, group-based work to share experiences, and career progression options all serve to improve volunteer satisfaction and performance.

Takeaways: We are in the midst of a global volunteering movement aimed at helping Syrian refugees, although support is waning. Volunteers often start out handling administrative tasks, but quickly move on to personalized and crucial areas of care. Governments have come to rely on committed volunteers to cover some basic needs in addition to long-term adjustment support.

8 Best

practices in this arena include: weekly meetings between volunteers and organizational staff for sharing ideas and case management scenarios, an informational “hotline” for refugees, covering transportation and communication costs via a monthly allowance for volunteers, and establishing a Facebook group to share experiences and recruit others (Mirghani 2013).

13

Table 1: Volunteer-Driven Resettlement Programs Worldwide

Program

Type

Country

Target Group

Outcomes

Result

Handling 1st reception centrally for 14 days, w/ interpreting services, provision of clothing, medical care, counselling & cultural orientation supported by volunteers

Govt. Program

Germany

Refugees

Integration-enhancing arrival reception

German best practice, refugee-reported positive opportunity to get bearings, learn about the country & daily routines, & identify the right contact people

“HOST” program matching volunteers w/newcomers for navigational, social & admin support, job search help, language practice

Govt. Program

Canada

Refugees & Immigrants

Language proficiency, community knowledge, navigational independence, social & professional network

Matched over 18,000 newcomers by 2017 (inc. 9,000 refugees)

Peer support from settled refugees & volunteers (e.g. help signing up for sports clubs, tutoring, biking lessons, field trips)

Govt. Program

E.U. (most expansive in Belgium)

Refugees

Integration-enhancing social support

E.U. best practice

“Bank of Offers” w/ variety of inkind services offered on a voluntary basis to support integration. Promoted by religious orgs, companies & volunteers

Govt.Private Program

Poland

Refugees

Integration-enhancing arrival reception

Polish best practice

Lutheran Immigration Services’ BRYCS program: dozens of programs nationally coordinating between refugee service providers, religious organizations, and local volunteers to link refugees w/ info, training & education programs, transportation, and American “buddies”

Private Program

U.S.

Refugees

Language proficiency, children enrolment in Head Start, social support, psychological well-being

Substantial up-take (1000s of refugees) and improved outcomes in some sub-programs

4.2 Mentorship Mentoring is broadly defined as a “developmental relationship” whereby an individual advises someone less experienced for the sake of the latter’s growth and advancement (Kram and Isabella 1985). Mentoring is associated with enhanced work effectiveness (Kram and Isabella 1985), job success (Roch 1979; Fagenson 1989), salaries and bonuses (Roche 1979), promotions (Stumpf and London 1981), and self-reported career mobility, recognition, and satisfaction (Scandura 1992). Unfortunately, none of these studies are causally identified. Proteges and non-proteges likely differ on unobserved motivation.

14

With this constraint in mind, mentorship is thought to work through two main pathways. First, psychosocial support allows a protege to cope with the stress of career management, hold positive attitudes toward the work environment, and view the mentor as a role model and coach (Kram and Isabella 1985; Ostroff and Kozlowski 1993; Wallace 2001a; Baugh, Lankau and Scandura 1996). In this vein, “social learning” theories suggest that the example-setting, friendship, and counseling provided by mentors improves the mentee’s sense of competence (Eby, Lockwood and Butts 2006; Noe, Greenberger and Wang 2002). Second, mentorship can enhance earnings and career function in more direct ways. Mentors impart advice about the culture and protocols within an organization that enable proteges to rapidly acclimate. Mentors also advocate on behalf of proteges, recommending them for leadership positions, pay raises, and communicating their accomplishments to senior management (Chao, Walz and Gardner 1992). This blend of accelerated skill acquisition and reputation-building can raise protege earnings (Wallace 2001b; Dreher and Cox Jr 1996) and open up doors to promotions (Fagenson 1989; Scandura 1992; Murrell and Tangri 1999). This support is possible because of the senior individual’s status, experience, and organizational influence (Allen et al. 2004; Ragins 1989; Reichman and Sterling 2001). This explanation aligns with the importance of social capital to employment highlighted in a survey (n = 525) of Canadian refugees, which finds that social networks – not human capital – are most predictive of employment (Lamba 2003). Although in need of updating and testing, best practices for mentorship programs include (Scandura 1992): • Perceived input into the matching process, correlated with both mentor and mentee satisfaction and motivation (Allen, Eby and Lentz 2006; Viator 2001) • Skill or sector-based matches (Barron, Black and Loewenstein 1987) • Shared cultural background, enhances identification with role model (Ragins 1997) • Hours spent in high-quality training, positively associated with mentorship satisfaction for proteges (Allen, Eby and Lentz 2006) Canada has pioneered sector-matched career mentorship programs targeted at immigrants and refugees. Partners include professional associations, educational institutions, employers facing labor shortages, and cities. Sector councils work with community colleges to provide bridging programs for technical upgrading in their sector (e.g. Information Technology Association of Canada and provincial hospital associations). By encouraging their employees to be mentors, employers gain access to potential recruits (65% of apprentices are recruited via employee references (Lerman, Eyster and Chambers 2009), become eligible for tax credits, fulfill corporate social responsibility prerogatives, and often receive diversity leadership training on behalf of the mentorship program (Alboim 2002). If Canadian employers hire an individual who has been on welfare assistance for the previous six months, they can apply for wage subsidies. Interestingly, interviews with staff at the Immigrant Employment Council of British Columbia (IECBC) reveal that most corporations are not aware of these funds when they sign up to use the platforms described below (Mousa 2017b). Two Canadian programs stand out, both of which are employer-facing and based in Ontario and British Columbia. Job Connect works with resettlement agencies to input the profiles of government-assisted refugees (GARs) and immigrants into their database. Employers then browse and filter potential recruits. Employers identify candidates and request a connection, facilitated by Job Connect, and set up an interview. Note that Job Connect is not a job-matching tool or job board, but rather a free portal for employers seeking diverse and motivated employees. Most recruitments are made for skilled jobs, despite the fact that employers are most in need of low-skill labor. This is because clients are typically unable to meet language proficiency requirements imposed by Job Connect – candidates should reach a level six Canadian Language Benchmark (CLB) out of ten, but most lie between levels zero and two according to IECBC staff. Government statistics indicate that 56% of 2009 arrivals (including immigrants) spoke no English or French (Government 2009). Moreover, clients face obstacles in accepting low-

15

skilled positions due to transportation and childcare issues, in addition to low wages that make public benefits more attractive. Low-skill jobs that have been filled have been at farms, construction sites, and Arabic-speaking bakeries. Since launching in February 2017, Job Connect has posted 325 refugee profiles in B.C. (Mousa 2017b). The second program, Mentor Connect, matches refugees and immigrants to mentors based on work experience and education. Matched pairs embark on a 12-week relationship, meeting for an hour every week. The mentor must introduce the mentee to at least three people within the industry by the end of the program. The referral system at the heart of this model suggests that Mentor Connect relies on social networks and industry exposure to improve employment outcomes. By 2008, the national Mentoring Partnership Program graduated 2,000 skilled immigrant mentees, with 72% of these finding employment within a year. However, the independent effect of enrollment is unknown and several mentorship programs are restricted to professionals.9

Takeaways: Mentorship is associated with both psycho-social and career function benefits that may enhance career mobility and earnings. It is unclear which mechanism, if any, is leading to these outcomes. Mentorship programs tend to be targeted toward high-skill professions. Matches based on sector, shared cultural background, and personal choice in the selection process are best practices.

4.3 Apprenticeships Apprentices are applied students that enter into a structured training program of classroom and paid on-the-job training under the guidance of a mentor. Classroom instruction is often offered at community or vocational colleges. As their skills increase, so do their wages. Upon completion of the program, apprentices earn an industryrecognized credential and are usually hired into a job that marks the start of a career. Lengths of apprenticeships vary from around one to four years (typically 3 – 4 years for manufacturing and construction apprenticeships), with costs ranging anywhere from $20,000 to $250,000 per apprentice. Most of this money goes to compensating the apprentice (Smith and Kemmis 2013). Other important costs cover program start-up, tuition and educational materials, mentors’ time, and overhead. Including non-registered apprenticeships, most apprentices (80%) in the U.S. work full-time (Helper et al. 2016). The payoff for workers is clear: 91% of American apprentices find employment after completing their program, and their average starting wage is above $60,000 (Steinberg and Schwartz 2014). A study of ten U.S. states finds that apprentices earn 1.4 times more than non-participants matched for pre-apprenticeship history six years on, returning $28 in benefits for every $1 in costs and yielding net gains of $125,000 over a worker’s career (Reed et al. 2012). Apprenticeships enjoy higher returns relative to other training programs, including two-year community college programs, in a variety of contexts (Hollenbeck and Huang 2016; Brodaty, Crépon and Fougère 2001). Similar increases associated with apprenticeships are observed globally (McIntosh 2007; Euwals and Winkelmann 2004). Of course, there are serious selection problems with these studies that could vary across sectors. Construction apprenticeships boost wages by 32% in the U.K. with no observed effect for those in retail – it is hard to know how much of this difference is explained by selection vs. the program itself (McIntosh 2007). Employers see returns, too. Apprentice sponsors gain a pipeline of loyal, reliable, skilled workers, increase productivity, and improve the bottom line (Lerman, Eyster and Chambers 2009). Because apprentices start out earning about 40 – 50% less than a fully trained employee, businesses save money on wages and thus offset some 9 For

instance, the Toronto-based TRIEC program sets the language requirement at a lofty level eight (out of ten), requires postsecondary education (accredited locally if possible), and focuses exclusively on high-skill professions like accountancy and IT.

16

Figure 1: Apprenticeship Up-Take in the U.S.

of the costs of offering an apprenticeship program, including mentor time, training equipment, and tuition for classroom-based instruction. Over 60% of U.S. employers surveyed said saving money on wages was either a “very important” or “somewhat important” benefit of apprenticeship, with 86% saying they would “strongly recommend” hiring an apprentice (n = 974) (Helper et al. 2016). The productive value of apprentices is apparent cross-nationally. British apprentices increase their firm’s economic output by an average of $366 per week, Canadian employers make $1.47 for every $1 invested in apprenticeships, and Swiss apprentices generate $450 million in profits for the 30% of companies that participate in their national apprenticeship schemes, perhaps the world’s best (for Economics and Research 2013; Forum 2009; of Foreign Affairs 2013). Similar returns to investment for companies have been observed across the OECD (Hoeckel and Schwartz 2010). Funding for U.S. apprenticeship programs relies heavily on employers, often in collaboration with trade unions. In the construction industry, it is common for both employers and workers to contribute a modest amount to finance the classroom component of apprenticeships. With some exceptions, sponsors receive no public funding for the work-based learning or classroom-based components of apprenticeship (Richard 2012). However, employers are typically able to take advantage of low-cost tuition programs at community colleges subsidized by state and local governments. A few states like North Carolina and Connecticut use financial incentives to encourage the creation of apprenticeships.10 Because of these positive results, the U.S. Department of Labor invested $265 million to meet President Obama’s goal of diversifying and doubling the number of apprenticeship slots to 750,000 by the end of 2018 (Steinberg and Schwartz 2014). The half million mark was passed in 2016 (Steinberg and Schwartz 2014). President Trump signed a executive order in June 2017 to double the $200 million in taxpayer money spent on apprenticeships by re-allocating money from existing job training programs (Kellman 2017). Many states are increasing funding for 10 The South Carolina government funded annual employer tax credits of $1,000 per apprentice per year beginning in 2007.

Connecticut offers employers 50% of the wages of apprentices up to $4,800, but only for apprentices in the manufacturing, construction, or plasticsrelated trades.

17

technical assistance, tax credits to employers, and career and technical training to prepare students for apprenticeships (Steinberg and Schwartz 2014). Over 1,000 occupations are included in the U.S. registered apprenticeship system, traditionally in craft trades. Five of the top six occupations are in construction (electrician, carpenter, plumber, construction craft laborers and pipe fitters), with 36% of sponsors and 40% of apprentices in construction (Smith and Kemmis 2013). Apprenticeships are slowly expanding from core skilled trades to industries like health care and IT with a need for skilled workers, and openness to women and minorities (Helper et al. 2016). Despite their proven benefits, American companies are not familiar with apprenticeships according to an Obama administration report (Kellman 2017). The U.S. apprenticeship system is small compared to those of other advanced economies as a result, with very limited government support. Consider that England started 510,000 new apprentices in 2012 while the U.S. started 147,000 despite a sixfold difference in population size (Steinberg and Schwartz 2014). 70% of young people in Switzerland and Germany enter the workforce through an apprenticeship, but only 12% of American workers report undergoing training programs at all. The half-million apprentices in the U.S. make up a mere 1.5% of the workforce (Smith and Kemmis 2013). Apprenticeships come with some challenges. First, and most significantly, non-completion is a continuing concern for about half of apprenticeship sponsors (Smith and Kemmis 2013). Second, unlike apprenticeships programs in most other countries, the U.S. system rarely reaches students in the 15–19 age group (the median age is 27–28, with a third above the age of 34 when they start). Third, “poaching” by competitors post-completion is a concern, but a waning one according to a 2013 study of n = 974 apprentice sponsors (Smith and Kemmis 2013). Fourth, about a third of American employers cite instruction costs and the costs to experienced workers’ time to instruct apprentices as a minor problem (Lerman, Eyster and Chambers 2009). Many also say they want help in finding and screening applicants, as well as in finding related instruction. Who becomes an apprentice, and are refugees likely to benefit? American sponsors cite referrals by current employees (65%), educational institutions (40%), and community organizations (including pre-apprenticeship programs) (20%) as recruitment sources (Lerman, Eyster and Chambers 2009). Yet the market may be harder for disadvantaged groups like refugees. In Norway, ethnic minority applicants of non-western origin have to outperform their majority peers in order to have the same chance of obtaining an apprenticeship (Helland and Støren 2006). Nonetheless, disadvantaged groups thrive when they can access “earn and learn” programs, especially when provided by firms with greater training capacity (Card and Sullivan 1987; Fuhrer and Schweri 2010; Mohrenweiser 2012). Bridge programs have also been helpful in preparing refugees for apprenticeships, as thresholds in English and math skills need to be met (Richard 2012). Taking up a job with no prior learning can improve confidence and soft skills among those with poor skills, but the quality of the match with employers becomes so critical that on-the-job learning programs are still preferred (Hasluck and Green 2007; England 2015). In Europe, internships and apprenticeships are seen as a way to improve language, soft skills, and social integration for newly arrived refugees while getting a foothold in their host country’s labor market (especially for those without formal proof of work experience or qualifications). These opportunities also offer a valuable first salary to new arrivals. Resettlement agencies like the IRC have similarly graduated hundreds of clients through apprenticeship programs. • TRIEC’s “Canadian Bridge” program: Employers pay a $1,500 per month stipend to interns who work for 4, 6, 9 or 12 months while gaining Canadian work experience in their field. • The U.K’s National Health Service (NHS) professional re-certification program graduated 821 refugee doctors supported over the 2011-13 period, out of which 25% entered a medical position at a level corresponding to his/her qualifications immediately after.

18

• Swedish Fast Track Initiative included “bridging courses” that graduated 941 participants in teacher courses and 224 participants in higher administrative courses ( ≤ 1% of the eligible population). Using a matching technique to compare those in the program with those similar on relevant dimensions but not enrolled, the program is associated with: higher earnings (445,400 SEK/year vs. 388,800 SEK/year), employment rates in relevant fields (44% vs. 27%) and profession according to education (47% vs. 35%). • Porsche and Siemens offer paid internships for refugees with graduate degrees and German or English proficiency. Internships included workplace orientation, skills assessment and training, and an anti-discrimination staff campaign. Porsche technical youth apprenticeship program employed 26 out of 30 refugee apprentices in full-time positions. Siemens employed 16 refugees in a 6-month pre-qualification course to prepare for an apprenticeship.

Takeaways: Apprenticeships are proven to be productive and revenue-generating for both employers and apprentices. Up-take remains low but growing in the U.S., largely due to a lack of corporate awareness of these benefits (97% of apprentice sponsors would recommend the program in a 2007 survey of 974 firms). The causal effects are unknown due to self-selection of motivated individuals, but refugees can and do find stable, skilled work with bridging programs leading to apprenticeships.

4.4 Private Sponsorship The primary alternative to the government-settled refugee track is private sponsorship. Sponsorship is a commitment by a group of private citizens to assist with resettling a refugee case through an organized contribution of in-kind goods, services, and financial assistance. When resettlement duties are shared with government agencies, this arrangement is known as co-sponsorship. Private sponsorship is gaining traction among host countries around the globe – with the UK, Ireland, Argentina, and New Zealand announcing programs in 2017 (Immigration and Canada 2017). Simply put, private sponsorship combines mentorship with cash transfers and some core services, like securing housing. The most expansive and established private sponsorship program in the world is that of Canada. There, privately sponsored refugees (PSRs) enter the labor market more quickly than government-assisted refugees (GARs) (71% vs. 45%) and earn double on average ($20,000) one year on – converging on an average income of $35,000 only after a decade (Hyndman 2017; Immigration and Citizenship Canada 2016; DeVoretz, Pivnenko and Beiser 2004). There is a large selection effect, however. Not only must refugees request PSR status, but they are significantly more likely to be better educated and speak English or French, sponsors often arrange employment for PSRs prior to arrival, PSRs tend to live with their sponsors (at least initially), and 62% of them are sponsored by family members. Given these disparities in support systems, it is unsurprising that GARs are double as likely to be receiving social assistance at the two-year point than their PSR counterparts (73% vs. 33 %) (Immigration and Citizenship Canada 2016). GARs are also disproportionately impacted by delays in accessing language training and hurt by reductions in government financial support upon finding a job (Yu, Ouellet and Warmington 2007; Immigration and Citizenship Canada 2016). The exception to this trend are GARs aged 20 – 29, who find work faster than PSRs of the same age, perhaps due to high pressure to enter the labor market immediately given few financial assets. PSRs and GARs are nearly identical, however, in application rates for Canadian citizenship (Pressé and Thomson 2007). Full-fledged private sponsorship – private citizens providing core services and support in lieu of the government – has not taken off in the U.S. Co-sponsorship, however, is underway on a small, localized, and often ad-hoc

19

level among American resettlement agencies. Co-sponsorship is similar to Canada’s PSR model in that it often includes financial responsibilities, such as paying rent or utilities, and providing material goods like furnishings for an apartment or children’s school supplies. Co-sponsorship differs from private sponsorship in that the primary responsibility for core services and well-being remains the resettlement agency. Somewhat surprisingly, co-sponsorship in Canada (known as the BVOR hybrid program) has been far less popular than private sponsorship, resettling only 808 BVORs between 2010 and 2014 compared to almost 40,000 PSRs.11 Similar to their Canadian counterparts, privately sponsored refugees in the U.S. earn higher median incomes than governmentsettled refugees – $16,600 vs. $11,900 among the 2009–2013 arrivals (Desiderio 2016; Eby et al. 2011). Sponsorship assistance from community groups has become increasingly ad-hoc, however, as VolAgs work mainly through national institutions (Nawyn 2005). Moreover, the screening process is slow for U.S.-bound refugees (18 to 24 months), which can decrease both sponsor enthusiasm and client motivation (Hainmueller, Hangartner and Lawrence 2016b; Hyndman 2017). The appetite to assist refugees boomed among some American civic organizations in response to the Syrian refugee crisis. 18 mayors pledged to resettle more refugees (Scott 2015) and several cultural groups, such as the Syrian American Council, the Arab American Institute, the U.S. Conference of Catholic Bishops and the Islamic Society of North America, signed open letters expressing interest in supporting resettled refugees (Foley 2015a). Private sponsorship by cultural groups has a mixed past in the U.S.: in 1990, the U.S. Coordinator for Refugee Affairs created a one-year unfunded refugee pilot program for Soviet Jewish refugees through MOUs signed in February 1990 with the Hebrew Immigrant Aid Society and the Conference for Jewish Federations to privately resettle 8,000 refugees. These refugees were ineligible for public assistance programs, however, and were resettled to small towns with poor employment prospects, leading to a failure of program renewal (Center 2015). Nonetheless, a provision in the PRM-VolAg contract allows for resettlement agencies to outsource non-essential services to sponsors registered with a community association – most often church congregations. What could an American co-sponsorship program look like? There is legal precedent and federal authority for an American private sponsorship program, whereby the President can allocate slots for privately-sponsored refugees (invoked by President Reagan to admit 16,000 refugees). Such a program must comply with the principle of “additionality,” i.e. any privately sponsored refugees should be in addition to – not instead of – government settled refugees. Privately sponsored refugees would also have to undergo existing systems for screening pre-arrival. Refugees with no U.S. ties could then be matched with a trained and vetted American sponsor (an individual or organization), or sponsored by a pre-existing U.S. tie. Resettlement agencies could train and monitor sponsorships, ensuring that basic needs are being met, avoiding exploitation or abuse, and serving as a safety net should the sponsorship relationship dissolve (IRAP 2016). Agencies could also match sponsors to refugees based on language, culture, or occupation sector to optimize the psychological and economic returns to network effects. Whereas ethnic cultural organizations are experienced sponsors in Canada, the small-scale co-sponsorship programs in the U.S. currently tend to attract White, primarily English-speaking church congregations (Mousa 2017a). Structuring a co-sponsorship model around the interest expressed by civic groups that share a cultural background with new arrivals could prove fruitful. Localities, universities, or civic organizations can drive campaigns to recruit cosponsors. Toronto’s Ryerson University, OCAD University, University of Toronto, and York University launched just such a campaign. A year after its launch, the universities’ refugee resettlement effort raised CAD 4.5 million, formed 102 sponsorship groups and supported 150 refugee families at an estimated cost of CAD 27,000 per family (Keung 2016).

11 The high demand for the PSR program and the misalignment of UNHCR and Canadian selection processes reduced the eligible pool

of candidates and popularity of the BVOR program. The data we have, however, suggest that BVORs tend to be younger and more educated than GARs, but less likely to speak an official language (Immigration and Citizenship Canada 2016).

20

Some sponsors could be exclusively financial, pooling together resources to support other private sponsors in covering medical screening and other costs, or donating directly to the resettlement agency (IRAP 2016). Donating to screening costs could accelerate the process without skipping security or health checks (Foley 2015b). A hybrid model could combine financial commitments from distant sponsors with the volunteerism and mentorship of proximate civic groups (like co-ethnic religious groups) with low financial resources. An additional model of co-sponsorship could stress career mentorship, and take place at the municipal level or via corporations. Under this arrangement, refugees could be provided special job opportunities like sector-matched apprenticeships. Co-sponsorship programs like this could tackle one of the biggest disadvantages of the Canadian model – overdependence and “friendship fatigue” (Kantor and Einhorn 2017). Another innovation can be tapering off financial assistance near the end of the sponsorship relationship, to avoid the financial and psychological stress associated with benefits cliffs (Mousa 2017a).

Takeaways: Privately sponsored refugees fare better, but the independent effect of private sponsorship is unknown as these refugees arrive with more “favorable” traits. The funds committed by sponsors could free up refugee arrivals to focus on language, acclimating, and the job search rather than making rent, but may lead to over-reliance. RCT evidence is needed to disentangle the contributions of the mentorship vs. cash aspects of sponsorship. Table 2: Volunteer-Driven Resettlement Programs Worldwide

Program

Sample

Commitments

Duration

Pros

Cons

PrivatelySponsored Refugees (Canada), often through faith groups

45,000 approved and 16,000 resettled (2017), groups of 5-10 individuals or families

$50,000 that can be tied or untied at sponsors’ discretion, registering for services (replace agency), social adjustment & mentorship, job help as a soft goal

12 months

Self-sufficiency and learning by doing at best (when sponsors are hands-off, encourage ESL, and help clients reach their own goals). Pros and cons to untied cash. Cash assistance crucial to freeing up clients to focus on English, finding work, and acclimating. Easy to fundraise. Arabic language community huge help (e.g. bank staff, doctors.) and co-ethnic co-sponsors more effective for this reason

Over-reliance (extra infusions of cash after 12 months) and dissatisfied sponsors – these tend to misunderstand their roles as sponsors, instead internalizing client problems rather than teaching by doing and stepping away. Cons of untied cash: financial mismanagement possible

IRIS sponsorship pilot (New Haven, CT), church groups

58 refugee families matched with groups of 5 –10 co-sponsors

Secure short-term housing, PRM furniture requirements & rental assistance for up to 6 months ($4,000-7,000), core service enrollment, check in regularly w/ case worker

12 months

Money is “huge,” most important component. Co-sponsorship worked well for high-risk (ICM) cases

Over-reliance, social isolation, and worse economic outcomes than non co-sponsored groups. Often resettled in remote locations close to co-sponsors, co-sponsors doing things for clients rather than teaching them how, resentment built up because co-sponsors had specific expectations of gratitude

21

4.5 Direct Cash Transfers An important component of co-sponsorship is a financial commitment that can be tied (e.g. rent payments) or untied (e.g. monthly stipend). There is reason to be optimistic about the impact of direct cash transfers based on program evaluations and randomized control trials among the poor in both developing and industrialized contexts. Most relevant to the U.S. refugee space, a series of negative income tax (“NIT”) experiments in the U.S. and Canada in the 1970s (n = 9,115) targeted poor families and provided benefits pegged to the national poverty line ($10,000 in 2014 dollars for a family of four in the U.S). As families got closer to the poverty line via earned income, the cash decreased until it was lost entirely. Note that this incentive structure is similar to the current U.S. public benefits system. Studies of these NIT programs found little evidence of reduced working hours (Burtless et al. 1986), while those who did reduce hours instead invested in education or childcare (Widerquist 2005). A variety of positive impacts unrelated to work were also recorded such as improved nutrition especially for children, increased home ownership, and improved test scores for children (Burtless et al. 1986). A review of 165 cash transfer programs in the developing world reached the same conclusions: cash transfers, on average, improve the health and schooling of children and savings, investment, and health in adults (Bastagli et al. 2016; Duflo 2003; Baird et al. 2012; Macours, Schady and Vakis 2008; Amarante et al. 2011; De Mel, McKenzie and Woodruff 2012; Blattman, Fiala and Martinez 2013). For these reasons, UNICEF, UK Aid, the World Food Programme, USAID, and others have launched flagship cash transfer programs with millions of beneficiaries collectively. Two common concerns are often cited in response to direct cash payments. First, one might worry that cash transfers could disincentivize work. Yet there is little evidence that cash transfers affect on the number of hours worked, with some studies finding increases in working hours as household members migrate to obtain better jobs (Salehi-Isfahani and Mostafavi 2016; Fiszbein and Schady 2009). Most studies that report a significant effect on adults of working age found an increase in work participation (3/4 studies) and intensity (4/6) (Hagen-Zanker et al. 2016). Transfers below a living wage also seem successful: a controlled UNICEF study distributing 300 rupees ($4.50) per month to Indian recipients intentionally sized the transfer below the level required for basic needs so as not to make them a “substitute for employment.” The study found improved nutrition, improved school attendance, increases in assets, and increases in hours worked (Bharat, UNICEF et al. 2014). A second concern is that direct cash may encourage some recipients to squander cash on non-essential or harmful goods. A World Bank meta-analysis concluded that the poor do not systematically abuse cash transfers, with studies consistently showing no significant impact or a significant negative impact of transfers on “temptation goods” spending (Evans and Popova 2014). We may find, however, that the information asymmetries experienced by refugees as opposed to the working poor in native populations may lead to uninformed or otherwise suboptimal spending decisions. A next step in this research agenda is to study how transfers affect financial behavior in countries with low wealth and large amounts of high cost debt such as the U.S., according the the Innovations for Poverty Action (IPA)’s U.S. Finance Initiative. Cash transfers are not revolutionary – half of world’s countries incorporate direct cash into their social safety nets (Gentilini et al. 2014) – but unconditional and untargeted basic income is. Current basic income pilots are ongoing in Finland ($650/month for all residents), Oakland, CA ($1,500 – 2,000/ month for 50 residents) and Stockton, CA (announced in October, 2017). Untied stipends also exist in Alaska (around $2000/year) to widespread support and among Eastern Cherokee in North Carolina ($4000/year for 350 members) linked to lower rates of substance abuse, school absence, and minor crimes (Costello et al. 2010). The design of core transfer features, particularly the size of the transfer, duration of its receipt, and method of transfer (e.g. mobile money) is crucial to achieving greater impacts (Haushofer and Shapiro 2013). Programs should be designed creatively and carefully to meet the distinct needs of each target group.

22

Takeaways: Cash transfers, especially those large and regular enough to bring beneficiaries up to and above the poverty line, are associated with improved health, nutrition, and educational attainment in children and greater savings, assets investments, and mental health in adults in developing and developed contexts. Several RCTs show that cash transfers are unlikely to reduce the number of hours worked. More work is needed to test the direct effects of cash transfer among refugees, who may be less informed about consumption options in their new environment.

4.6 ESL Education The percentage of public school students in the United States who are “ELLs” (English Language Learners) has been rising year after year, reaching 9.4%, or roughly 4.6 million students in 2015 (Statistics 2015). One of the greatest challenges facing the teachers of ELLs lies in the diversity of their linguistic, ethnic, and educational backgrounds, with ELLs also entering the U.S. at every age and grade level. An estimated 80% of ELL students are Spanish-speaking, but a much more diverse picture is painted when considering refugees, especially adults that face additional barriers to language attainment (Calderón, Slavin and Sánchez 2011). The policy landscape further constrains opportunities to learn English. Public investments in second language acquisition is low: only a tenth of adults with limited English proficiency (LEPs) are enrolled in ESL courses (Foundation 2008). Moreover, propositions passed in Arizona, California, and Massachusetts imposing English-only language classes actually worsened English proficiency and academic achievement in the long run (Goldenberg 2015). With these constraints in mind, what factors are associated with second language acquisition? Predictors of success can be categorized into three pools: (1) individual traits; (2) learning environment; and (3) methods of instruction. Variation in students’ literacy within their native language is a strong predictor of their ability to learn and become fluent in English. The shorter the “linguistic distance” between the students’ native tongue and English, the easier it is to learn English (Koda 2007). Illiteracy in one’s first language thus presents a significant obstacle to second language attainment. Relatedly, those with interrupted formal education, and especially refugee children who have never attended school, tend to have literacy and subject matter gaps that go beyond language learning, particularly if they enter U.S. schools in the upper grades (Flege and Liu 2001). Another important individuallevel factor is motivation. This was the strongest predictor of success for users of Duolingo, and motivation of parents was cited as an untapped resource for schools in a seminal review of ESL commissioned by the U.S. Department of Education (Flege and Liu 2001; Vesselinov and Grego 2012). Motivation itself may be negatively shaped by perceiving oneself as a member of a low-status language group (August, Shanahan and Escamilla 2009). Motivation may be low among some refugees, given that most did not choose to emigrate (Richmond 1988). Perhaps the most crucial determinant of language acquisition is age. Linguists have fiercely debated the “critical period hypothesis:” the idea that an ideal time window exists to acquire language, after which further language acquisition becomes much more difficult and effortful. Meta-analyses indicate that “younger = better in the long run,” but with several caveats. Around 5% of adult bilinguals master a second language even though they begin learning well into adulthood, motivation can compensate somewhat for neurobiological capacities weakened by age (Klein 1995; Moyer 2004; Piller 2002), and adults’ performance can improve measurably over time with a substantial amount of native speaker input (Flege and Liu 2001). Some linguistic aspects appear to be more affected by age than others: for instance, adult ELLs will nearly always retain an identifiable foreign accent, including those who display perfect grammar. In short, English test performance is generally linearly related to age of arrival up to puberty; after puberty, performance is low but highly variable (Johnson and Newport 1989; Qureshi 2016).

23

There is clear evidence that tapping into first-language literacy can confer advantages to ELLs. Several metanalyses (cumulative n = 405 studies) reached two conclusions. First, instructional programs work best when they provide opportunities for students to develop proficiency in their first language (Genesee et al. 2005; Greene 1997; Rolstad, Mahoney and Glass 2005; Slavin and Cheung 2005). Second, instruction in one’s native language as well as in English improves English reading proficiency compared with students instructed only in English (n = 68 studies). Instructional programs that invest time and resources into the development of the first language literacy are more effective than those that are English medium or English only programs. These findings concur with those of virtually all of the other research syntheses in this area (Genesee et al. 2005; Greene 1997; Rolstad, Mahoney and Glass 2005; Slavin and Cheung 2005). Policy evidence further bolsters the case for bilingual instruction. Propositions severely limiting bilingual instruction in three states lowered English proficiency and academic achievement among students immersed in English-only instruction, as well as stunting the supply of bilingual instructors for decades to come.12 Unfortunately, American legislation from the Bilingual Education Act of 1968 to No Child Left Behind in 2002 has shifted continuously away from bilingual instruction and toward English-only education, which further contributes to educational inequity suffered by emergent bilinguals (García and Kleifgen 2010). Bilingual instruction is associated with a “moderate” effect size (Plonsky and Oswald 2014) and has not been found to delay or hinder English attainment (Cummins 2000). This finding concords with qualitative evidence linking higher education achievement among second generation immigrant students with maintaining their native languages and cultures (Portes and Rumbaut 2001). Environments both inside and outside of the classroom shape language attainment as well. Consider that poverty is a strong predictor of language fluency: children on welfare hear an average of 176 words an hour while affluent children hear 487, and a difference of 30 million words in exchanges with adults separates the two groups by the age of three (Putnam 2016; Hart and Risley 1995). Language decays in similar ways to other educational attainments among disadvantaged groups. In a 5-year panel study of immigrant children (n = 470) in Boston and San Francisco, GPAs dropped consistently over time and father’s employment status was a strong predictor of academic achievement (Suárez-Orozco, Suárez-Orozco and Todorova 2009). School quality also matters, with immigrant students more likely to attend low-performing schools and only 12% of public school teachers are trained to teach ELLs (Koda 2007). Less is known about the effect of environmental factors on adults, but the time, motivation, and emotional health constraints imposed by poverty likely play a role.

Takeaways: Bilingual instruction has been found to be consistently more effective than English-only immersion. Proficiency in one’s native language improves second language proficiency, further accelerated by linguistic distance to English. Motivation and age are key determinants of English language acquisition, as well as environmental factors like school quality and poverty.

Meta-analyses identified the following pedagogical best practices, with a focus on children. For a review of apps aimed at adult ELLs, see Section 7.4 in the Appendix. • Teaching English using the learner’s native language. Bilingual classroom contexts are hugely varied, but “heritage language schools” are common in multicultural societies (Creese and Blackledge 2010) 12 A Stanford University study found that in Arizona, the reading achievement gap between English-learners and non-English-learners

increased by about 1.5 grade levels for 4th and 8th graders. In California, the gap increased almost as dramatically in 8th grade and slightly in 4th. In Massachusetts, the achievement gap increased somewhat at both grade levels between ELLs and native speakers (Goldenberg 2015).

24

• Types of instruction that help native English speaking children learn to read in English are advantageous for ELLs as well; however the effect sizes are lower and adjustments are needed (n = 49 studies). This may be linked to the finding that similar proportions of ELLs and monolingual speakers are classified as poor readers and word-level skills (e.g., decoding, spelling) of both groups are very similar. Note that this is not the case for text-level skills (e.g., reading comprehension, writing). These skills rarely reach levels equal to monolingual English speakers (n = 107 studies) (August, Shanahan and Escamilla 2009). • Vocabulary instruction, divided by practical content areas (Short and Fitzsimmons 2007). Frequent exposure to a word in multiple forms; ensuring understanding of meaning(s); providing examples of its use in phrases, idioms, and usual contexts, pronunciation, spelling, and word parts; and, when possible, teaching its cognates, or a false cognate, in the learner’s primary language. This helps remove vocabulary-related constraints to passing standardized tests (Calderón, Slavin and Sánchez 2011) • Instruction that covers the key components of reading identified by the National Reading Panel (phonemic awareness, phonics, fluency, vocabulary, and text comprehension) (August, Shanahan and Escamilla 2009) • Open-ended questions and multiple exposure to words during shared reading help children know how to use those words (Robbins and Ehri 1994) • To become good readers – to be able to recognize words and comprehend a text simultaneously – English learners require practice at both decoding and fluency (Haynes and Grabe 2010) • There is no single approach to writing instruction that will meet the needs of all students. In general, the research base on cross-linguistic influences in the acquisition of writing is sparse (Calderon and MinayaRowe 2010) • Adequate assessment systems. As the National Literacy Panel has found, assessments used to gauge language-minority students’ language proficiency and to make placement and reclassification decisions are inadequate in most respects (August, Shanahan and Escamilla 2009)

5

Survey of Pilots, Programs, and RCTs

Below is a review of select experiments and pilots worldwide. I note where sample sizes are known, where no treatment effects were found in related interventions, and where programs were targeted at both immigrants and refugees. Best practices refer to government tried, tested, and recommended programs. Table 5.1 outlines interventions for the pre-departure period, suggesting productive uses of wait times such as language instruction or cultural orientation. While the European Commission prioritizes the pre-departure period, U.S. VolAgs have virtually no contact with refugees for the two years (or more) before they arrive (Table 5.1).13 Table 5.2 presents access points with local host communities for the first month after arrival. Shifting to employment-focused interventions, Table 5.3 outlines bridging courses, job training, recredentialing programs targeted mainly at the medium to high skilled. The final two tables cover financial literacy, micro-finance, and entrepreneurship programs (Tables 5.4 and 5.5). Average intervention durations tend to vary by issue area in the following way (based on EU practices): • • • • • •

Pre-departure phase: 2 days – 2 weeks Welcome reception, post-arrival orientation: 2 – 6 weeks Gradual transition from paying partial to full rent: 12 months Career development/VESL: 3 – 12 months Internships and business incubators: 1 – 3 months Re-credentialing via testing: 3 – 5 days of testing

13 See:

the 2017 Action Plan on the Integration of Third-Country Nationals.

25

• Financial literacy seminars: 1 – 2 hours

5.1 Pre-departure Preparation

Intervention

Type

Country

Sample

Dependent Variable(s)

Result

Cultural orientation via webinar or Skype with bi-cultural trainer

Govt. Program

Netherlands, Finland, Norway

8,000 refugees (Norway)

Awareness of host country society, expectations about arrival & resettlement, minimization of culture shock

E.U. best practice

Refugee country seminars for municipalities and resettlement service providers

Govt. Program

Norway

820 municipal workers

Awareness of countries and cultures of resettled refugees (26 refugee country seminars held as of May 2017)

E.U. best practice

Online help w/ identifying potential employers, job search skills & filing for professional recertification

Province Program

Quebec

Refugees

Faster accreditation, job search assistance

Evaluation unknown, program on-going

60-hour English-language course piloted by PRM in camps in Kenya, Thailand, and Nepal

Govt. Program

U.S.

U.S.-bound refugees

Basic English proficiency

Improved preparation and selfconfidence

Early Intervention Program: language learning, qualifications recognition, job search help in early stages of application

Govt. Pilot

Germany

700 refugees

Language proficiency, credentials recognition, identifying jobs

Succeeded and rolled out nationally

26

5.2 Processing, Arrival, and Community Engagement

Intervention

Type

Country

Sample

Dependent Variable(s)

Result

Handling 1st reception centrally for 14 days, w/ interpreting services, provision of clothing, medical care, counselling & cultural orientation

Govt. Program

Germany

Refugees

Integration-enhancing arrival reception

German best practice, refugee-reported positive opportunity to get bearings, learn about the country & daily routines, & identify the right contact people

Direct access to affordable housing while “shifting to rent” (bail glissant ) in France. Incrementally pay rent proportional to income over 12 months before paying full rent

Govt. Program

France (direct housing: Netherlands and Finland)

Refugees

Housing independence at 12 months

French, Dutch and Finnish best practice

Peer support from settled refugees & volunteers (e.g. help signing up for sports clubs, tutoring, biking lessons, field trips)

Govt. Program

E.U. (most expansive in Belgium)

Refugees

Integration-enhancing social support

E.U. best practice

“Bank of Offers” w/ variety of inkind services offered on a voluntary basis to support integration. Promoted by religious orgs, companies & volunteers

Govt.Private Program

Poland

Refugees

Integration-enhancing arrival reception

Polish best practice

Lutheran Immigration Services’ BRYCS program: dozens of programs nationally coordinating between refugee service providers, religious organizations, and local volunteers to link refugees w/ info, training & education programs, transportation, and American “buddies”

Private Program

U.S.

1000’s of refugees

Language proficiency, children enrolment in Head Start, social support, psychological well-being

Substantial up-take and improved outcomes in some sub-programs

27

28

5.3 Training, Re-credentialing, and Employment

Intervention

Type

Country

Sample

Dependent Variable(s)

Result

The IMPRINT consortium works directly with employers to mentor, train (via VESL), re-credential, and place immigrants

NGO Program

U.S. Canada

Vocationallytrained immigrants

Re-credentialing, English proficiency, mentorship, job placement

16-year national programs nation-wide, served 11,300+ immigrants with some programs expanded to refugees recently

Uni. of Metropolia’s Technology and Engineering Pilot Program: tests math, digital competence, & engineering tests in Arabic, English & Finnish

Govt. Pilot

Finland

30 refugees w/ technology or engineering background

Skills accreditation

Almost all participants certified, rolled out nationally. Also promising: E.U. will have 765,000 unfulfilled tech vacancies in 2020

Fast tracking skilled refugees into shortage occupations via vocational training paths set up by companies and the public sector (21 occupations)

Govt. Program

Sweden, Germany

High-skill refugees

Training, certification, mentoring, vocational language proficiency, and work placement

E.U. best practice as employers trust the results of vocational training they designed

Paid internships at Siemens & Porsche for refugees w/ graduate degrees and German/English proficiency (inc. workplace orientation, skills assessment & training, and anti-discrimination staff campaign)

Govt.Private Pilot

Germany

113 high-skill refugees

Training, accreditation, apprenticeship, employment

Second edition rolled out in Nov. 2016, 11/13 beneficiaries at Porsche now employed there

Kiron Open Higher Education: 23 universities offer online & inperson classes, no German required initially

Govt.Private program

Germany

1,250+ refugees

Remote e-learning and accreditation of university credits obtained this way

Evaluation unknown, ongoing program

World Education Services alternative qualifications assessment: reconstructs academic histories and advises on equivalency of credentials when documentation unavailable

Private Program

Canada

Syrian refugees

Academic attestation & re-credentialing for Syrian refugees

Pilot phase

Norwegian Agency for Quality Assurance in Education (NOKUT) accreditation program based on academics’ assessment and written tests

Govt. Program

Norway

500+ refugee applicants

Accreditation

220/500 applicants obtained re-credentialing, over half of those then found relevant jobs in Norway and 4/5 were satisfied with the service

Swedish bridging courses for recognizing qualifications of highly qualified immigrants (medical professionals, teachers, and lawyers)

Govt. Program

Sweden (sister programs in Norway and UK)

High-skill immigrants & refugees

Professional accreditation

Foreign-qualified teachers who participated in the courses were 18% more likely to find relevant employment

&

29

5.4 Financial Literacy

Intervention

Type

Country

Sample

Dependent Variable(s)

Result

Financial education seminar held for Indian migrants (Seshan and Yang 2014)

Field Experiment

Qatar

232 Indian migrant couples (157 treated husbands)

Financial/ savings knowledge & practices

Treated participants were 44.6% more likely to listen to financial matters radio show, 13.2 percentage-points more likely to make financial decisions with their wives (advocated by the seminar), and more likely to demonstrate sound financial and savings practices 13-17 months after the intervention

World Bank RCT to induce savings among migrants through financial education aimed either at migrants, the families of migrants, or both the migrant and their family (Doi, McKenzie and Zia 2014)

Field Experiment

Indonesia

400 migrant households (296 treated)

Financial/savings knowledge & practices

Treatment increased knowledge of migrant insurance & cost of a remittance transaction by 25.6 and 7.9 percentagepoints. Treated households 10 percentage-points more likely to have prepared cash records, saved, and opened a bank account in the past 6 months

Varying migrant control over El Salvador-based savings by offering different types of accounts across treatment groups (Ashraf et al. 2015)

Field Experiment

U.S.

898 migrants from El-Salvador

Savings

Migrants offered the greatest degree of control accumulated the most savings at the partner bank, compared to others offered less or no control over savings

Financial literacy programs (Karunarathne and Gibson 2014)

Field Experiment

Australia & New Zealand

Pacific islander, east Asian & Sri Lankan migrants

Financial literacy knowledge, use of cheaper remittance products, amount & frequency of remittances

Small increase in knowledge but no effect the use of cheaper remittance products, or on the amount and frequency of sending money home

30

5.5 Micro-finance and Entrepreneurship

Intervention

Type

Country

Sample

Dependent Variable(s)

Result

Monthly start-up classes allowing refugees to connect w/ local business community, take classes in Arabic, receive mentoring & legal assistance

Germany

Govt. Pilot

Refugees in Berlin

Start-up support

Still in pilot phase by German Chamber of Trade and Industry (IHK)

Incubator for Immigrants program (similar to German program above)

Netherlands

Govt. Pilot

Refugee entrepeneurs

Start-up support

40 business plans from asylum seekers and refugees less than a year after launching (mid-2017)

UNHCR “graduation approach’ micro-finance program, where refugees pass incremental steps from financial literacy training, to micro-grants, to micro-loans. Treatments: self-employment track (SET) & wage-employment track (WET)

UNHCR program

Egypt

1,000 lowincome Syrian refugees

Income, job placement, start-up creation & survival

Income increased by 18 – 27% (still cannot meet basic needs), only 17-35% of WETS retained jobs after 6 months, 78-97% of SETs opened businesses but majority folded in 6 months

5.6 E.U. Best Practices While not yet developed into EU-wide programs, below is a compilation of recommendations proposed by the E.U. to member-states in order to improve resettlement outcomes (Desiderio 2016): • Increased communication and cooperation between stakeholders (e.g. steering committees, quarterly meetings) • Adequate information and cultural orientation at pre-departure stage to prepare candidates for transfer and manage expectations • Early medical assessments and coordination with local health staff • Direct and immediate access to housing, ideally without passing through reception centers • Social involvement from volunteers, e.g. engaging local churches and community organizations and preparing them for refugee arrivals • Geographic assignment near urban centers

5.7 Canadian Best Practices The Canadian government produced a similar set of suggestions aimed at improving resettlement outcomes (Immigration and Citizenship Canada 2016): • Directly communicating with all refugees about the services available to them to rebuff outdated or inaccurate information from family and friends

31

• Making translation services available at all settlement services agencies (refugees cite this as the largest barrier to accessing settlement services). Too often, language support is ad-hoc and pro-bono. Service providers should partner so that one translator (especially in a lesser spoken language) can offer assistance to different programs/offices based on demand • Covering transportation costs, especially for refugees in rural areas (e.g. bulk bus and taxi passes) to help access services and employment opportunities • On-the-job training with a Canadian company or organization, which can provide value to potential employers and allow refugees to learn workplace norms. This can include expanding internship and mentorship programs, but depends on employers’ willingness to take part and can be over-subscribed. Predicted labor shortages should be leveraged • Streamlining the process of recognizing foreign credentials • Job-matching support or subsidized hires • Preliminary skills assessment as part of orientation program, but specific enough to be useful for job matching purposes

6

Paths Forward and Conclusions

Despite a large potential to contribute and thrive, refugees in the U.S., Canada, and Europe often struggle to find adequate employment and can take decades to escape low-income traps. Many of the risk factors are wellknown: lack of language proficiency, recognition of foreign credentials, and affordable housing located near urban centers, among others. The impact of other factors, like ethnic enclaves or work–first resettlement models, are less apparent. More longitudinal data are needed for a wider array of outcomes (e.g. university admission rates, secondary migration) to establish ground truths about refugee trajectories – and how these differ by sub-group in this incredibly diverse population. The high degree of heterogeneity within the refugee community poses clear challenges in programs design and targeting. Leveraging geographic placement algorithms that interact personal characteristics with place to optimize employment probabilities, as well as offering bilingual ELL instruction where possible, are two interventions that can accommodate this diversity. Six intervention spaces in particular seem encouraging given the evidence reviewed here. These interventions work within current U.S. policy constraints geared toward the 180-day employment outcome, by aiming to enhance labor market prospects without delaying employment. Table 6 below provides estimates regarding the feasibility, cost-effectiveness, and likely effect sizes of interventions in these spaces. These recommendations adjudicate between interventions backed by evidence in the U.S. and elsewhere, and those that tackle spaces known to be important for integration (e.g. housing placement) where scant systematic evaluations exist. 1. Re-credentialing, possibly via external testing by potential employers in the private sector, re-credentialing NGOs, or universities that test particular skill sets (e.g. accounting, coding). Given that re-credentialing itself does not require training, these programs leverage existing human capital at low costs. Caseworkers could also be trained to identify high-capacity individuals to fast-track for re-credentialing.14 2. Placement assignments. City and neighborhood effects seem significant, and randomization could occur in the placement process (subject to ethical constraints) across and within cities.

14 The

success of different service provision models themselves has received scant attention in this review. There appears to be large amounts of variation in service provision models at the local level, and various models (e.g. integrative vs. piece-meal) likely have a significant effect on resettlement outcomes (Simich et al. 2005). For this reason, adjusting and standardizing service provision models could be a fruitful avenue.

32

3. Direct cash and investment capital for refugees. This includes matched micro-savings, conditional loans (e.g. tied to educational attainment) or business start-up support for those with a business ownership background. 4. Vocational ESL, as an independent program or as part of a re-credentialing or training package. English proficiency is strongly correlated with employment rates and earnings, and VESL is an established (although under-evaluated) method of enhancing the employability of refugees in their professional fields. 5. Private co-sponsorship that connects refugees to a network of individuals in a host community. This can range from formal co-sponsorship with high financial and time commitments (perhaps matched on occupational sector) to more light-touch engagement of local churches, schools, and volunteer groups (e.g. “buddy” systems that cover activities like supermarket trips, tutoring, and apartment-hunting) that can provide practical advice, emotional support, job search support, and even supplementary financial support. 6. Apprenticeships. Work experience in the host country is often highly valued, and apprenticeships are expanding in the U.S. Apprenticeship tracks for refugees focused on local boom industries have proven successful in the U.S. and Europe, but a model for scaling up to regional or national schemes is needed. Resource constraints in the American resettlement program, the availability of large pools of volunteers, and the efficacy of volunteer-led programs abroad encourage outsourcing some services to local communities. Programs that combine community support with financial assistance seem particularly promising. The evidence reviewed here suggests that private co-sponsorship, apprenticeships, and direct cash transfers are well-positioned to overcome the network, economic, and linguistic constraints faced by refugees. The relative importance of mentorship vs. financial support – the key components of sponsorship – remains unknown, however. In addition, programs that include financial assistance would benefit from shared learning and tie-ins with programs that target lowincome Americans. For instance, both populations risk losing eligibility for badly-needed welfare assistance by accepting direct cash transfers. Creative work-arounds could include matched micro-savings, or targeting small families in states with low welfare coverage in order to reduce the opportunity cost of foregoing welfare assistance. Whichever routes are taken up by those of us working on refugee integration, an inconsistent data landscape and on-going migrant crisis render the need for scholars to partner with service providers more important than ever.

33

34

Target Group(s)

Professionalized & high-skill refugees, teenagers and adults of all ages

All refugees

High motivation adult refugees

Refugees w/ vocational or professional backgrounds

All refugees (might be some preferences for co-religious or co-ethnic groups)

Mid to high skilled refugees

Intervention Space

Re-credentialing

Placement location

Investment capital (entrepreneurial support & educational loans)

Vocational ESL

Private cosponsorship (light touch volunteering to formal sponsorship)

Apprenticeships

Medium, depends on associated tuition costs or wage subsidies

Low for community outreach (info sessions at schools & churches, “buddy” systems). High for formal cosponsorship, but both could reduce reliance on formal service provision

Low–Medium, depends on breadth of vocations covered & whether VESL provided as part of re-credentialing package

Medium, depends on loan sizes & repayment rates

Low additional costs given that housing assignment is part of processing anyway & Air B‘n’B partnership could drive down costs

Low–Medium if recredentialing without additional training (depends on costs of partnerships)

Cost?

Only if partnering with national corporations (exc. franchises), otherwise very localized

Private co-sponsorship depends on availability of willing partners, lighttouch local outreach is scalable condition on local-level cooperation

Yes, conditional on critical mass of clients in the targeted vocational fields

Scalable conditional on efficiently identifying suitable candidates

Yes, housing assistance already part of processing for all refugees

If done through corporate partners or NGO’s that design certifications, yes, but if done by creating expert panels & university testing systems from scratch, less so

Scalable?

evaluation

Yes, e.g. employer and self-reported skills assessments, and credentials later on

Yes, e.g. self-reported happiness, connections to jobs, housing, & civic opportunities, reliance on resettlement agencies in 3-6 month range

Yes, e.g. technical language proficiency certification

Yes, e.g. creation of business plan or obtaining license for entrepreneurs, educational enrolment & attainment among students

Possibly, e.g. rates of transportation network & resettlement agency office use, employment quality 3 months later

Yes, e.g. test pass rate or applicant certification rate

Short-term possible?

Large given that apprenticeships often lead fo full-time, skilled employment

Yes, depending on depth of program, could provide functional social support that helps employment, language attainment & mental health

Likely large effects on language proficiency & well-matched employment if employers recognize program value

Depends on quality of candidates but likely large

Housing location & quality are key to myriad outcomes, so likely large effect sizes on at least some outcomes

Likely to have a large average effect size on binary outcomes like test-passing since baseline underemployment is so high. Better job opportunities will likely have family spillover effects

Effect size?

Table 3: Feasibility Checks for Possible Intervention Spaces

Descriptive evidence that apprenticeships increase credentialing, full-time employment, and earnings

Coordinated & well-informed local reception is an E.U. ‘best practice’ – supports refugees while keeping local community engaged and prepared for newcomers (missing in the U.S.), leverages grassroots desire to help. Social support greatly needed for government-resettled refugees

VESL successfully baked into IMPRINT re-credentialing courses in the U.S. (12,000+ immigrants served since 2000) & training programs designed with employers across the E.U. that increase employment rates and client satisfaction

Small-scale incubators and entrepreneur mentorship programs successful in Europe, NYEDC currently piloting educational loans ($5,000 on average) to 150 refugees

Housing location is key for service use, transportation & job access, but there is little systematic evidence estimating the effect of housing characteristics on these outcomes (secondary migration also an issue)

Strong evidence that re-credentialing is urgent to avoid eroding human capital and accessing the labor market. IMPRINT re-credential programs targeting specific professions in the U.S. have re-certified 12,000+ immigrants, E.U. initiatives with companies (Siemens, Porsche), universities & public sector (doctors, lawyers, teachers) boost job placement

Evidence?

7

Appendix

7.1 What motivates volunteers? Identifying what motivates volunteers is crucial to resettlement agencies that hope to leverage these resources. Common trends emerge when assessing volunteer-level traits and organization-level practices associated with volunteer motivation, satisfaction, and job performance among the general population of volunteers. At the broadest level, a mixture of social rewards, altruism, and a strong affinity for group activities are common drivers to volunteer (Schlegelmilch and Tynan 1989). Indices that capture organizational support, participation efficacy, and peer relationships have subsequently been validated in contexts as different as the U.S. and China (Wong, Chui and Kwok 2011; Galindo-Kuhn and Guzley 2002). Perhaps the most ubiquitous measure of volunteer satisfaction is the Volunteer Functions Inventory. Motivations for volunteering are assessed using the 30-item VFI (Clary et al., 1998) that focuses on six functions (assessed with five items each) that volunteering can serve:

Figure 2: Volunteer Function Index

A meta-analysis of over 20,000 volunteers across a range of issue areas shows that the Values factor obtained the highest mean score whereas the lowest scores were observed for the Career and Enhancement factors. Studies conducted with samples with a mean age under 40 years obtain higher scores on Career and Understanding scales when compared to studies in older samples. The group of studies with fewer than 50% women yield higher mean scores on the Social scale than studies with more than 50% women in the sample. The nine studies reporting reliability coefficients range from .78 and .84, suggesting little measurement error (high test-retest reliability) in the scores. Of the 26 studies that performed factor analysis, 18 confirmed the original structure of six factors (Chacón et al. 2017). How predictive are these factors of volunteer behaviors? Recent studies indicate that motivational satisfaction, task satisfaction, and management satisfaction can make the difference between those who remain in the organization and those who leave after a 12-month tracking period (Vecina, Chacon, & Sueiro, 2009). Strongly identifying with the organization often matters, manifested in believing its goals and values, willing to make an effort to benefit the organization, and the desire to remain a member (Meyer & Allen, 1984, 1997; Mowday, Steers, & Porter, 1979). This variable maps well onto the intention to remain in the longer term and can compensate for moderate, temporary decreases in satisfaction and improve responses to frustrating events (Vecina et al. 2012; Garner and Garner 2011). Many studies show that commitment, rather than satisfaction, is the variable which clearly differentiates two extreme groups of volunteers: those who abandon their volunteering during the first year

35

and those who remain after seven years (Vecina, Chacon, & Sueiro, 2010). “Organization-based self-esteem” (OBSE) captures this commitment to the organization well: the perception individuals have of themselves as important, meaningful, effectual, and worthwhile within their organization (Vecina et al. 2012). Important volunteer-level motivations thus include commitment to the organization’s values, a sense of being valued and participating in decision-making, and working within groups rather than individually. From an organizational perspective, traits associated with satisfaction included clearly articulated mission statements, structured support for volunteers, and arranging for volunteers to share experiences and work in groups (Drucker 1989). It follows that the match between volunteers and organizations matters as well. The functional approach to volunteerism holds that outcomes from volunteering (e.g., satisfaction and intentions to remain a volunteer) are a function of the match between a volunteer’s motivations and opportunities to meet those motivations found in the environment (i.e., the volunteer activities, position, or organization). Indices capturing match success seem capable of predicting outcomes better than volunteer-level traits or organizational-level opportunities. The importance of good matches seem more important in less structured and smaller organizations (Stukas et al. 2009). Based on these findings, the following are recommendations to approach volunteer recruitment and motivation in a more intentional way: • Training and orientation that demonstrates the value of volunteer work to the organization • Clearly articulate the organization’s mission and affirm the importance of volunteer work to achieving these humanitarian goals • Recognize good performance, hold performance reviews at least once a year, and provide career ladders or expanded responsibilities for young volunteers (conversely, re-allocating poor performers especially in environments with long wait lists) • Consulting volunteers and involving them in decision-making that affects their work and the work of the organization, allowing them to express their ideas and rate their satisfaction with tasks and projects (e.g. check-ins with volunteer coordinators) • Organizing volunteers into working groups that meet regularly to share experiences and work on common problems, with opportunities to share achievements and best practices with the organization

7.2 Re-credentialing Programs in the U.S. Below are a few established re-credentialing programs run by immigrant-focused NGOs in the U.S., some of which have expanded to working with refugees (Singer 2012): 1. Upwardly Global, currently with offices in Chicago, San Francisco, and New York, uses volunteers from U.S. companies to serve as peer mentors to help immigrants with interviewing techniques and experiential knowledge. They also have screen, present, and place candidates in jobs at partner businesses. UG educates employers about the potential pool of employees, the U.S. visa system, and how to evaluate foreign credentials. Their employee engagement program has trained more than 2,800 immigrant professionals from 127 countries since 2000. 2. The Welcome Back Initiative (WBI), in 10 locations across the country, works exclusively with professionallytrained immigrants in the healthcare sector. WBI connects these professionals with typically underserved communities compatible with their language and cultural skills. WBI retrains, recredentializes, and licenses healthcare workers to practice in the U.S. and market their skills to U.S. employers. Housed mainly in community colleges, and offering VESL courses, the WBI has served 11,000 immigrant health professionals since 2001. 36

3. World Education Services (WES) validates foreign credentials for job-seekers (and students) as well as for employers and licensing boards. Their Global Talent Bridge program reconstructs clients’ academic history and advises on licensing for information technology, engineering, accounting and other professionals. Both Upwardly Global and the Welcome Back Initiative work with WES to get their credentials assessed and recognized by employers. WES is currently piloting a program for Syrian refugees in Canada. 4. Together, these organizations and others (including the Massachusetts Immigrant and Refugee Advocacy Coalition) comprise the coalition known as IMPRINT. IMPRINT’s mission is to work with business, government, higher education, and other partners to raise awareness about the talents and contributions of immigrant professionals in the U.S. and to streamline the complexities of professional licensing and re-credentialing of skilled workers. into the U.S. workforce. 5. The New York City Economic Development Corporation (NYCEDC) is set to launch its Immigrant Bridge Program, which will provide credential assessment, training, internship referrals and job search support. It will also offer low-interest micro-loans to immigrants to fund tuition, training, exam fees, and any other educational costs to upskill in their fields. The NYCEDC expects to lend to around 150 people during the first 18 months with an average loan of $5,000.

7.3 Apprenticeship Alternative: Supported Employment Supported employment aims to place “unemployable” clients into competitive jobs while providing on-going intensive support services, such as job coaching, role “shadowing,” and mentoring. Overall, supported employment seems to substantially improve employment outcomes and performance among at-risk populations and modestly improve them among the unemployed at large (Bond 2004; Sumnall and Brotherhood 2012). Education also tends to improve, but usually as a direct result of credentialing built into the program (Mawn et al. 2017). The sooner clients enroll, the better. Studies of Swedish job support services and Austrian job-coaching (n = 32,000) indicate that program effects (reducing residual unemployment duration by 32%) are most concentrated within the client’s first year of unemployment (Weber and Hofer 2004; Sianesi 2004). Some concerns worth noting include: sub-minimum wages, disincentives to participate due to the interplay of benefits, and a focus on entry-level rather than skilled jobs. Revenue streams often include wage subsidies by NGOs, which are generally expensive (Grimm and Paffhausen 2015). This may still be acceptable if the jobs created are sustainable, but evidence regarding whether this is really the case is scarce. A major cost component is the dead weight loss produced by the fact that many of the workers hired under a subsidized rate would have been hired anyway. An experimental study in Sri Lanka found a strong correlation between pre-program hiring intentions and program uptake (De Mel, McKenzie and Woodruff 2013). However, earlier studies reported benefits of $1.97 to every $1 in cost with $13,815 in gross wages and fringe benefits to the employee whom others viewed as “permanently unemployable” (Hill et al. 1987). A British job assistance and wage subsidy experiment likewise raised employment by about 5 percentage points but long-run evidence was mixed (Blundell et al. 2004). Individual-level traits also shape success. Non-cognitive skills and motivation are important determinants of success and these can be improved more successfully and at later ages than basic cognitive skills (Heckman 2000). On the other hand, those with deeply entrenched barriers and without positive attitudes are among the hardest to help (Hasluck and Green 2007). Helping the most disadvantaged is resource intensive, usually requiring individually tailored support over a prolonged period. Individuals may make considerable progress without reaching employment. There is a question as to whether employment is a realistic goal for some groups, and by extension, how best to target interventions. It should be stressed that studies on supported employment have mainly been run on at-risk groups such as the severely mentally ill and the disabled, meaning that low baselines may lead to large effect sizes that are not directly transferrable to other groups. 37

Successful supported employment programs (based on roughly 15 RCTs) include: • Support to employers to help them understand the strengths and weaknesses of different disadvantaged groups • On-the-job mentoring or buddying • Competitive employment (not sheltered work) as an attainable goal for clients • Rapid job search as opposed to lengthy pre-employment assessment, training, and counselling • Individualized job placements according to client preferences • At least 6 months of programming, with follow-up up until 2 years

Takeaways: Supported employment (mentored work placements) improves wages, hours worked, and professionalism for extremely disadvantaged segments of American society, but unlivable wages and low-skilled work remain a problem.

7.4 Apps for Adult English Learners Online language programs are growing substantially in users and revenue. The worldwide market for digital English learning products (the most popular language) was $2.8 billion in 2015 and projected to grow to $3.8 billion by 2020 (Seave 2016). Many companies are competing for market share, including busuu, Duolingo, Babbel, and Rosetta Stone. However, entry has slowed considerably – the last major competitor to enter was Duolingo in 2011 (Seave 2016). These companies may have different features to help speed up learning and varying price schemes, but struggle to distinguish themselves among potential new users. Nonetheless, different apps have comparative advantages in teaching styles and skill sets targeted. Just as well: language learning is a complex process where any given skill (e.g. reading) can be further broken down into distinct components (e.g. word recognition, decoding, and comprehension). Learners also have different goals, with some prioritizing conversational fluency where others aspire to written proficiency. The most successful path to English learning may thus be a combination of apps tailored to the user’s needs, making sure to cover speaking practice offline or using in-app chat features. The apps reported below are all supported on mobile and web browsers, targeted at adults, and offer instruction from a range of languages (including Arabic, Turkish, and French) to English.

38

Table 4: Adult Language Learning Apps

Platform

Skills

Cost

User Base

Approach

Cons

Duolingo

Pronunciation, vocab, sentence construction, reading, listening

Free

120 million

“10,000 sentences” approach, repetition, gamification, “smart” app that focuses on your weakest areas

Less effective on mobile than computer (can be cheated), needs to be supplemented with conversational practice to get beyond A2 level

Lang-8

Writing

Free or Premium

750,000

Users upload text they’ve written and have it corrected by a native speaker. Users then correct others’ text in their native language

Posts by non-premium accounts can go unchecked, quality of corrections inconsistent

Busuu

Sentence construction, reading, listening

Free or Premium ($10/month)

60 million

Different voices, focus on writing skills, 10,000 sentences approach, gamification, chat with natives feature (web only)

Needs to be supplemented with conversational practice, less focus on repeated drills

Memrise

Vocabulary

Free or Pro ($5/month)

20 million

User chooses from a set of cues to remember a new word (mnemonics techniques) and retain it long-term, gamification, “smart” app that focuses on weak words

No speaking or grammar training, web version is inferior to app

Lingvist

Reading, listening

writing,

Free or Pro ($5/month)

1 million

Shows words in useful contexts, spaced repetition to encourage memorization, typing new language to activate muscle memory and spelling, can get to B1/B2 level

No speaking, limited to typing exercises

Babbel

Vocabulary, nunciation

pro-

Free or Pro ($7 – $13/month)

20 million

Focuses on building conversational skills, occasionally sets off immersive lessons by explaining grammatical concepts at length, pronunciation practice

Can be buggy, especially phone app version (e.g. must pronounce words exactly right to move on)

References Adida, Claire L, David D Laitin and Marie-Anne Valfort. 2016. Why Muslim integration fails in Christianheritage societies. Harvard University Press. Ager, Alastair and Alison Strang. 2008. “Understanding integration: A conceptual framework.” Journal of refugee studies 21(2):166–191. Akresh, Ilana Redstone. 2011. “Wealth accumulation among US immigrants: A study of assimilation and differentials.” Social Science Research 40(5):1390–1401.

39

Alboim, Naomi. 2002. Fulfilling the promise: Integrating immigrant skills into the Canadian economy. Caledon Institute of Social Policy Ottawa. Alesina, Alberto, Johann Harnoss and Hillel Rapoport. 2016. “Birthplace diversity and economic prosperity.” Journal of Economic Growth 21(2):101–138. Allen, Tammy D, Lillian T Eby and Elizabeth Lentz. 2006. “Mentorship behaviors and mentorship quality associated with formal mentoring programs: closing the gap between research and practice.” Journal of Applied Psychology 91(3):567. Allen, Tammy D, Lillian T Eby, Mark L Poteet, Elizabeth Lentz and Lizzette Lima. 2004. “Career benefits associated with mentoring for protégés: A meta-analysis.”. Allport, Gordon W. 1979. The nature of prejudice. Basic books. Amarante, Verónica, Marco Manacorda, Edward Miguel and Andrea Vigorito. 2011. Do cash transfers improve birth outcomes? Evidence from matched vital statistics, social security and program data. Technical report National Bureau of Economic Research. Andersson, Roger. 1996. “The geographical and social mobility of immigrants: escalator regions in Sweden from an ethnic perspective.” Geografiska Annaler. Series B. Human Geography pp. 3–25. Ashraf, Nava, Diego Aycinena, Claudia Martínez A and Dean Yang. 2015. “Savings in transnational households: a field experiment among migrants from El Salvador.” Review of Economics and Statistics 97(2):332–351. Åslund, Olof, Per-Anders Edin, Peter Fredriksson et al. 2011. “Peers, neighborhoods, and immigrant student achievement: Evidence from a placement policy.” American Economic Journal: Applied Economics 3(2):67– 95. Atkinson, Adele and Flore-Anne Messy. 2013. “Promoting financial inclusion through financial education.” OECD Working Papers on Finance, Insurance and Private Pensions (34):1. August, Diane, Timothy Shanahan and Kathy Escamilla. 2009. “English language learners: Developing literacy in second-language learners?Report of the National Literacy Panel on Language-Minority Children and Youth.” Journal of Literacy Research 41(4):432–452. Azorbo, Michelle. 2011. Microfinance and Refugees: lessons learned from UNHCR’s experience. UNHCR, Policy Development and Evaluation Service. Baird, Sarah J, Richard S Garfein, Craig T McIntosh and Berk Özler. 2012. “Effect of a cash transfer programme for schooling on prevalence of HIV and herpes simplex type 2 in Malawi: a cluster randomised trial.” The Lancet 379(9823):1320–1329. Bandiera, Oriana, Iwan Barankay and Imran Rasul. 2008. “Social capital in the workplace: Evidence on its formation and consequences.” Labour Economics 15(4):724–748. Bansak, Kirk, Jeremy Ferwerda, Jens Hainmueller, Andrea Dillon, Dominik Hangartner, Duncan Lawrence and Jeremy Weinstein. 2018. “Improving refugee integration through data-driven algorithmic assignment.” Science 359(6373):325–329. Barron, John M, Dan A Black and Mark A Loewenstein. 1987. “Employer size: The implications for search, training, capital investment, starting wages, and wage growth.” Journal of Labor Economics 5(1):76–89.

40

Bass, Sandra, Margie K Shields and Richard E Behrman. 2004. “Children, families, and foster care: Analysis and recommendations.” The future of children pp. 5–29. Bastagli, Francesca, Jessica Hagen-Zanker, Luke Harman, Georgina Sturge, Valentina Barca, Tanja Schmidt and Luca Pellerano. 2016. “Cash transfers: what does the evidence say? A rigorous review of impacts and the role of design and implementation features.” Overseas Development Institute, London . Baugh, S Gayle, Melenie J Lankau and Terri A Scandura. 1996. “An investigation of the effects of protégé gender on responses to mentoring.” Journal of Vocational Behavior 49(3):309–323. Beaman, Lori A. 2012. “Social networks and the dynamics of labour market outcomes: Evidence from refugees resettled in the US.” The Review of Economic Studies 79(1):128–161. Beirens, Hanne and Susan Fratzke. 2017. “Taking Stock of Refugee Resettlement: Policy objectives, practical tradeoffs, and the evidence base.”. Bertrand, Marianne, Erzo FP Luttmer and Sendhil Mullainathan. 2000. “Network effects and welfare cultures.” The Quarterly Journal of Economics 115(3):1019–1055. Bertrand, Marianne and Sendhil Mullainathan. 2004. “Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination.” The American Economic Review 94(4):991–1013. Bevelander, Pieter and Christer Lundh. 2007. “Employment integration of refugees: The influence of local factors on refugee job opportunities in Sweden.”. Bharat, SEWA, UNICEF et al. 2014. “A little more, how much it is? piloting basic income transfers in Madhya Pradesh, India.” January, Sewa Bharat and UNICEF, New Delhi . Blattman, Christopher, Nathan Fiala and Sebastian Martinez. 2013. “The economic and social returns to cash transfers: evidence from a Ugandan aid program.” Columbia University, Departments of Political Science and International & Public Affairs . Blommaert, Lieselotte, Marcel Coenders and Frank Van Tubergen. 2014. “Discrimination of Arabic-named applicants in the Netherlands: an internet-based field experiment examining different phases in online recruitment procedures.” Social forces 92(3):957–982. Blundell, Richard, Monica Costa Dias, Costas Meghir and John Reenen. 2004. “Evaluating the employment impact of a mandatory job search program.” Journal of the European economic association 2(4):569–606. Bond, Gary R. 2004. “Supported employment: evidence for an evidence-based practice.” Psychiatric rehabilitation journal 27(4):345. Borjas, George J. 1998. “To ghetto or not to ghetto: Ethnicity and residential segregation.” Journal of Urban Economics 44(2):228–253. Brodaty, Thomas, Bruno Crépon and Denis Fougère. 2001. Using matching estimators to evaluate alternative youth employment programs: evidence from France, 1986–1988. In Econometric evaluation of labour market policies. Springer pp. 85–123. Burtless, Gary et al. 1986. The work response to a guaranteed income: A survey of experimental evidence. In Conference Series;[Proceedings]. Vol. 30 Federal Reserve Bank of Boston pp. 22–59. Calderon, Margarita Espino and Liliana Minaya-Rowe. 2010. Preventing long-term ELs: Transforming schools to meet core standards. Corwin Press. 41

Calderón, Margarita, Robert Slavin and Marta Sánchez. 2011. “Effective instruction for English learners.” The Future of Children 21(1):103–127. Capps, Randy, Kathleen Newland, Susan Fratzke, Susannah Groves, Greg Auclair, Michael Fix and Margie McHugh. 2015a. “Integrating refugees in the United States: The successes and challenges of resettlement in a Global Context.” Statistical Journal of the IAOS 31(3):341–367. Capps, Randy, Kathleen Newland, Susan Fratzke, Susannah Groves, Greg Auclair, Michael Fix and Margie McHugh. 2015b. “Integrating refugees in the United States: The successes and challenges of resettlement in a Global Context.” Statistical Journal of the IAOS 31(3):341–367. Card, David and Daniel M Sullivan. 1987. “Measuring the Effect of Subsidized Training Programs on Movements In andOut of Employment.”. Cattaneo, Alejandra and Stefan C Wolter. 2012. “Migration policy can boost PISA results: Findings from a natural experiment.”. Center, Niskanen. 2015. “Statement for the Record of the Niskanen Center Submitted to The House Committee on the Judiciary Hearing on“The Syrian Refugee Crisis and Its Impact on the Security of the U.S. Refugee Admissions Program”.”. Chacón, Fernando, Gema Gutiérrez, Verónica Sauto, María Luisa Vecina and Alfonso Pérez. 2017. “Volunteer Functions Inventory: A systematic review.” Psicothema 29(3):306–316. Chao, Georgia T, Patm Walz and Philip D Gardner. 1992. “Formal and informal mentorships: A comparison on mentoring functions and contrast with nonmentored counterparts.” Personnel psychology 45(3):619–636. Cheung, Sin Yi and Jenny Phillimore. 2014. “Refugees, social capital, and labour market integration in the UK.” Sociology 48(3):518–536. Connor, Phillip. 2010. “Explaining the refugee gap: Economic outcomes of refugees versus other immigrants.” Journal of Refugee Studies 23(3):377–397. Costello, E Jane, Alaattin Erkanli, William Copeland and Adrian Angold. 2010. “Association of family income supplements in adolescence with development of psychiatric and substance use disorders in adulthood among an American Indian population.” JAMA 303(19):1954–1960. Creese, Angela and Adrian Blackledge. 2010. “Translanguaging in the bilingual classroom: A pedagogy for learning and teaching?” The modern language journal 94(1):103–115. Crosnoe, Robert. 2007. “Early child care and the school readiness of children from Mexican immigrant families.” International Migration Review 41(1):152–181. Cummins, Jim. 2000. “Putting language proficiency in its place: Responding to critiques of the conversational/academic language distinction.” English in Europe: The acquisition of a third language pp. 54–83. Damm, Anna Piil. 2009. “Ethnic enclaves and immigrant labor market outcomes: Quasi-experimental evidence.” Journal of Labor Economics 27(2):281–314. de Matos, Ana Damas and Thomas Liebig. 2014. “The qualifications of immigrants and their value in the labour market.” Matching Economic Migration with Labour Market Needs pp. 187–228. De Mel, Suresh, David McKenzie and Christopher Woodruff. 2012. “One-time transfers of cash or capital have long-lasting effects on microenterprises in Sri Lanka.” Science 335(6071):962–966. 42

De Mel, Suresh, David McKenzie and Christopher Woodruff. 2013. “The demand for, and consequences of, formalization among informal firms in Sri Lanka.” American Economic Journal: Applied Economics 5(2):122– 50. Desiderio, Maria Vincenza. 2016. “Integrating Refugees into Host Country Labor Markets: Challenges and Policy Options.”. DeVoretz, Don J, Sergiy Pivnenko and Morton Beiser. 2004. “The economic experiences of refugees in Canada.”. Doi, Yoko, David McKenzie and Bilal Zia. 2014. “Who you train matters: Identifying combined effects of financial education on migrant households.” Journal of Development Economics 109:39–55. Dreher, George F and Taylor H Cox Jr. 1996. “Race, gender, and opportunity: a study of compensation attainment and the establishment of mentoring relationships.” Journal of applied psychology 81(3):297. Drucker, Peter F. 1989. “What business can learn from nonprofits.” Harvard business review 67(4):88–93. Duflo, Esther. 2003. “Grandmothers and granddaughters: old-age pensions and intrahousehold allocation in South Africa.” The World Bank Economic Review 17(1):1–25. Easton-Calabria, Evan and Naohiko Omata. 2016. “Micro-finance in refugee contexts: current scholarship and research gaps.”. Eby, Jessica, Erika Iverson, Jenifer Smyers and Erol Kekic. 2011. “The faith community?s role in refugee resettlement in the United States.” Journal of Refugee Studies p. fer038. Eby, Lillian T, Angie L Lockwood and Marcus Butts. 2006. “Perceived support for mentoring: A multiple perspectives approach.” Journal of Vocational Behavior 68(2):267–291. Edin, Per-Anders, Peter Fredriksson and Olof Åslund. 2003. “Ethnic enclaves and the economic success of immigrants: Evidence from a natural experiment.” The quarterly journal of economics 118(1):329–357. Egypt, Beit Al Karma Consulting. 2013. “Mid-Term Evaluation of UNHCR Graduation Programme in Egypt.”. England, Public Health. 2015. Local action on health inequalities: reducing the number of young people not in employment, education or training (NEET). U.K. Government (PHE Gateway number: 2014334). Enos, Ryan D. 2017. The space between us: Social geography and politics. Cambridge University Press. Euwals, Rob and Rainer Winkelmann. 2004. “Training intensity and first labor market outcomes of apprenticeship graduates.” international Journal of manpower 25(5):447–462. Evans, David K and Anna Popova. 2014. “Cash transfers and temptation goods: a review of global evidence.”. Evans, William N and Daniel Fitzgerald. 2017. The economic and social outcomes of refugees in the United States: Evidence from the ACS. Technical report National Bureau of Economic Research. Fagenson, Ellen A. 1989. “The mentor advantage: perceived career/job experiences of protégés versus nonprotégés.” Journal of organizational behavior 10(4):309–320. Fazel, Mina, Jeremy Wheeler and John Danesh. 2005. “Prevalence of serious mental disorder in 7000 refugees resettled in western countries: a systematic review.” The Lancet 365(9467):1309–1314. Fazel, Mina, Ruth V Reed, Catherine Panter-Brick and Alan Stein. 2012. “Mental health of displaced and refugee children resettled in high-income countries: risk and protective factors.” The Lancet 379(9812):266–282. 43

Fiszbein, Ariel and Norbert R Schady. 2009. Conditional cash transfers: reducing present and future poverty. World Bank Publications. Fix, Michael, Kate Hooper and Jie Zong. 2017. “How Are Refugees Faring? Integration at US and State Levels.”. Flege, James Emil and Serena Liu. 2001. “The effect of experience on adults’ acquisition of a second language.” Studies in second language acquisition 23(4):527–552. Foged, Mette and Giovanni Peri. 2016. “Immigrants’ effect on native workers: New analysis on longitudinal data.” American Economic Journal: Applied Economics 8(2):1–34. Foley, Elise. 2015a. “Refugee Groups Say Communities Are Ready And Willing To Do More.”. Foley, Elise. 2015b. “These Groups Have An Idea To Help Syrian Refugees: Let People Sponsor Them.”. for Economics, Centre and Business Research. 2013. “THE IMPACT OF APPRENTICESHIPS.” Centre for Economics and Business Research . Forum, Canadian Apprenticeship. 2009. IT PAYS TO HIRE AN APPRENTICE: Calculating the Return on Training Investment for Skilled Trades Employers in Canada. Canadian Apprenticeship Forum. Foundation, Gates. 2008. “Innovation in ESL Technology: Mobile Learning Technology.”. Fuhrer, Marc and Jürg Schweri. 2010. “Two-year apprenticeships for young people with learning difficulties: a cost-benefit analysis for training firms.” Empirical research in vocational education and training 2(2):107– 125. Galindo-Kuhn, Roseanna and Ruth M Guzley. 2002. “The volunteer satisfaction index: Construct definition, measurement, development, and validation.” Journal of Social Service Research 28(1):45–68. García, Ofelia and Jo Anne Kleifgen. 2010. Educating emergent bilinguals: Policies, programs, and practices for English language learners. Teachers College Press. Garner, Johny T and Lindsey T Garner. 2011. “Volunteering an opinion: Organizational voice and volunteer retention in nonprofit organizations.” Nonprofit and Voluntary Sector Quarterly 40(5):813–828. Genesee, Fred, Kathryn Lindholm-Leary, William Saunders and Donna Christian. 2005. “English language learners in US schools: An overview of research findings.” Journal of Education for Students Placed at Risk 10(4):363–385. Gentilini, Ugo, Maddalena Honorati, Ruslan Yemtsov et al. 2014. “The state of social safety nets 2014.” The World Bank, Washington, DC . Goldenberg, Claude. 2015. “Congress: Bilingualism Is Not a Handicap.”. Gould, Eric D, Victor Lavy and M Daniele Paserman. 2004. “Immigrating to opportunity: Estimating the effect of school quality using a natural experiment on Ethiopians in Israel.” The Quarterly Journal of Economics 119(2):489–526. Government, Canadian. 2009. “Evaluation of the Host Program.”. Greene, Jay P. 1997. “A meta-analysis of the Rossell and Baker review of bilingual education research.” Bilingual Research Journal 21(2-3):103–122.

44

Grimm, Michael and Anna Luisa Paffhausen. 2015. “Do interventions targeted at micro-entrepreneurs and small and medium-sized firms create jobs? A systematic review of the evidence for low and middle income countries.” Labour Economics 32:67–85. Hagen-Zanker, Jessica, Francesca Bastagli, Luke Harman, Valentina Barca, Georgina Sturge and Tanja Schmidt. 2016. “Understanding the impact of cash transfers: the evidence.” ODI Briefing. London: Overseas Development Institute . Hainmueller, Jens, Dominik Hangartner and Duncan Lawrence. 2016a. “When lives are put on hold: Lengthy asylum processes decrease employment among refugees.” Science advances 2(8):e1600432. Hainmueller, Jens, Dominik Hangartner and Duncan Lawrence. 2016b. “When lives are put on hold: Lengthy asylum processes decrease employment among refugees.” Science Advances 2(8):e1600432. Halpern, Peggy. 2008. “Refugee economic self-sufficiency: An exploratory study of approaches used in office of refugee resettlement programs.” US Department of Health and Human Services . Hangartner, Dominik, Elias Dinas, Moritz Marbach, Konstantinos Matakos and Dimitrios Xefteris. 2017. “Does Exposure to the Refugee Crisis Make Natives More Hostile?”. Hao, Lingxin. 2004. “Wealth of Immigrant and Native-Born Americans.” International Migration Review 38(2):518–546. Hao, Lingxin et al. 2001. “Immigrants and Wealth Stratification in the US.” Joint Center for Poverty Research Working Paper 216:1–10. Hart, Betty and Todd R Risley. 1995. Meaningful differences in the everyday experience of young American children. Paul H Brookes Publishing. Hasluck, Chris and Anne E Green. 2007. What works for whom?: a review of evidence and meta-analysis for the Department for Work and Pensions. Number 407 Corporate Document Services. Haushofer, Johannes and Jeremy Shapiro. 2013. “Household response to income changes: Evidence from an unconditional cash transfer program in Kenya.” Massachusetts Institute of Technology . Haynes, Margot and William Grabe. 2010. “Reading in a second language: Moving from theory to practice.” Studies in Second Language Acquisition 32(4):648. Heckman, James J. 2000. “Policies to foster human capital.” Research in economics 54(1):3–56. Hek, Rachel. 2005. “The role of education in the settlement of young refugees in the UK: The experiences of young refugees.” Practice 17(3):157–171. Helland, Håvard and Liv Anne Støren. 2006. “Vocational education and the allocation of apprenticeships: Equal chances for applicants regardless of immigrant background?” European sociological review 22(3):339–351. Helper, Susan, Ryan Noonan, Jessica R Nicholson and David Langdon. 2016. “The Benefits and Costs of Apprenticeships: A Business Perspective.” US Department of Commerce . Hill, Mark L, P David Banks, Rita R Handrich, Paul H Wehman, Janet W Hill and Michael S Shafer. 1987. “Benefit-cost analysis of supported competitive employment for persons with mental retardation.” Research in Developmental Disabilities 8(1):71–89.

45

Hoeckel, Kathrin and Robert Schwartz. 2010. “Learning for Jobs OECD Reviews of Vocational Education and Training.” Austria: Organisation for Economic Co-operation and Development (OECD) . Hofinger, C., Zandonella M. and B Hoser. 2017. “Integrationsmonitor Oberoesterreich 2016.”. Hollenbeck, Kevin and Wei-Jang Huang. 2016. “Net impact and benefit-cost estimates of the workforce development system in Washington state.”. Hooper, Kate, Jie Zong, Randy Capps and Michael Fix. 2016. “Young children of refugees in the United States: Integration successes and challenges.”. Hooper, Kate, Maria Vincenza Desiderio and Brian Salant. 2017. “Improving the Labor Market integration of Migrants and Refugees.”. Hou, Feng and Morton Beiser. 2006. “Learning the Language of a New Country: A Ten-year Study of English Acquisition by South-East Asian Refugees in Canada.” International Migration 44(1):135–165. Hugo, GJ, Sanjugta Vas Dev, Janet Wall, ME Young, Vigya Sharma and Kelly Parker. 2011. “Economic, social and civic contributions of first and second generation humanitarian entrants.”. Hyndman, Jennifer. 2017. “The State of Private Refugee Sponsorship in Canada: Trends, Issues, and Impacts.”. Immigration, Refugees and Citizenship Canada. 2017. Five countries working on new refugee sponsorship programs. Cision. Immigration, Refugees and Evaluation Division Citizenship Canada. 2016. “Evaluation of the Resettlement Programs (GAR, PSR, BVOR and RAP).”. IRAP, International Refugee Assistance Project. 2016. “Private Sponsorship of Refugee Resettlement in the United States: Guiding Principles and Recommendations.”. Johnson, Jacqueline S and Elissa L Newport. 1989. “Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language.” Cognitive psychology 21(1):60–99. Kaas, Leo and Christian Manger. 2012. “Ethnic discrimination in Germany’s labour market: a field experiment.” German Economic Review 13(1):1–20. Kantor, Jodi and Catrin Einhorn. 2017. “Canadians Adopted Refugee Families for a Year. Then Came ‘Month 13.’.”. Karakayali, Serhat and J Olaf Kleist. 2015. “EFA-Studie.” Sturkturen und Motive der ehrenamtlichen Flüchltingsarbeit (EFA) in Deutschland. 1. Forschungsbericht. Ergebnisse einer explorativen Umfrage vom November/Dezember 2014 . Karakayali, Serhat and J Olaf Kleist. 2016. “Volunteers and asylum seekers.” Forced Migration Review (51):65. Karunarathne, Wasana and John Gibson. 2014. “Financial literacy and remittance behavior of skilled and unskilled immigrant groups in Australia.” Journal of Asian Economics 30:54–62. Kellman, Laurie. 2017. Trump orders more cash, industry input, for apprenticeships. Associated Press. Keung, Nicholas. 2016. “Ryerson celebrates success of Syrian sponsorship challenge.”.

46

Klein, Elaine C. 1995. “Second versus third language acquisition: Is there a difference?” Language learning 45(3):419–466. Koda, Keiko. 2007. “Reading and language learning: Crosslinguistic constraints on second language reading development.” Language learning 57(s1):1–44. Kram, Kathy E and Lynn A Isabella. 1985. “Mentoring alternatives: The role of peer relationships in career development.” Academy of management Journal 28(1):110–132. Lamba, Navjot K. 2003. “The employment experiences of Canadian refugees: Measuring the impact of human and social capital on quality of employment.” Canadian Review of Sociology/Revue canadienne de sociologie 40(1):45–64. Lerman, Robert, Lauren Eyster and Kate Chambers. 2009. “The Benefits and Challenges of Registered Apprenticeship: The Sponsors’ Perspective.” Urban Institute (NJ1) . Leseman, Paul. 2007. “Early education for immigrant children.” The Transatlantic Task Force on Immigration and Integration, Migration Policy Institute/Bertelsmann Stiftung . Lundborg, Per and Per Skedinger. 2014. “Minimum wages and the integration of refugee immigrants.”. Macours, Karen, Norbert Schady and Renos Vakis. 2008. “Cash transfers, behavioral changes, and cognitive development in early childhood: evidence from a randomized experiment.”. Magnuson, Katherine, Claudia Lahaie and Jane Waldfogel. 2006. “Preschool and school readiness of children of immigrants.” Social Science Quarterly 87(5):1241–1262. Mawn, Lauren, Emily J Oliver, Nasima Akhter, Clare L Bambra, Carole Torgerson, Chris Bridle and Helen J Stain. 2017. “Are we failing young people not in employment, education or training (NEETs)? A systematic review and meta-analysis of re-engagement interventions.” Systematic reviews 6(1):16. Mccarthy, Rose and Melanie Haith-Cooper. 2013. “Evaluating the impact of befriending for pregnant asylum seeking and refugee women.”. McIntosh, Steven. 2007. A cost-benefit analysis of apprenticeships and other vocational qualifications. Department for Education and Skills Nottingham. Mestheneos, Elizabeth and Elizabeth Ioannidi. 2002. “Obstacles to refugee integration in the European Union member states.” Journal of Refugee Studies 15(3):304–320. Mirghani, Zahra. 2013. “Healing through sharing: an outreach project with Iraqi refugee volunteers in Syria.” Intervention 11(3):321–329. Mohrenweiser, Jens. 2012. “Which firms train disadvantaged youth?” Empirical research in vocational education and training 4(2):115–130. Morland, Lyn, Nicole Ives, Clea McNeely and Chenoa Allen. 2016. “Providing a Head Start: Improving Access to Early Childhood Education for Refugees.” Migration Policy Institute . Morrissey, Kate. 2017. “Refugee auto loan program may be model to help low-income families with financing.”. Mousa, Salma. 2017a. “Interviews with 28 refugee cases and 20 staff members, at three resettlement agency offices and at refugee clients’ homes in Modesto, Turlock, and Oakland CA and Salt Lake City, UT.”.

47

Mousa, Salma. 2017b. “Phone interview with program manager at Immigrant Employment Council of British Columbia (IECBC).”. Moyer, Alene. 2004. Age, accent, and experience in second language acquisition: an integrated approach to critical period inquiry. Vol. 7 Multilingual Matters. Munshi, Kaivan. 2003. “Networks in the modern economy: Mexican migrants in the US labor market.” The Quarterly Journal of Economics 118(2):549–599. Murrell, Audrey J and Sandra Schwartz Tangri. 1999. “Mentoring at the margin.” Mentoring dilemmas: Developmental relationships within multicultural organizations pp. 211–224. Nawyn, Stephanie J. 2005. “Faithfully Providing Refuge: The Role of Religious Organizations in Refugee Assistance and Advocacy.” The Center for Comparative Immigration Studies pp. 565–585. Nazzal, Kristel Heidi, Marzieh Forghany, M Charis Geevarughese, Venus Mahmoodi and Jorge Wong. 2014. “An innovative community-oriented approach to prevention and early intervention with refugees in the United States.” Psychological services 11(4):477–485. Nibbs, Faith. 2016. “Moving Into the Fastlane: Understanding refugee upward mobility in the context of resettlement.”. Noe, Raymond A, David B Greenberger and Sheng Wang. 2002. Mentoring: What we know and where we might go. In Research in personnel and human resources management. Emerald Group Publishing Limited pp. 129–173. Nourse, Timothy H. 2003. “Microfinance for Refugees Emerging Principles for Effective Implementation.” Microfinance and refugees . OECD. 2016. “Responses to the refugee crisis: Financial education and the long-term integration of refugees and migrants.”. of Foreign Affairs, Federal Dept. 2013. “Earn While You Learn: Switzerland?s Vocational Education and Training System A Model for Apprenticeships in the United States.” Embassy of Switzerland in the United States of America . Orozco, Manuel. 2015. Economic Status and Remittance Behavior Among Latin American and Caribbean Migrants in the Post-recession Period. In Immigrant Vulnerability and Resilience. Springer pp. 207–234. Ostroff, Cheri and Steve WJ Kozlowski. 1993. “The role of mentoring in the information gathering processes of newcomers during early organizational socialization.” Journal of Vocational Behavior 42(2):170–183. Ottaviano, Gianmarco IP and Giovanni Peri. 2006. “The economic value of cultural diversity: evidence from US cities.” Journal of Economic geography 6(1):9–44. Paluck, Elizabeth Levy, Seth Green and Donald P Green. 2017. “The Contact Hypothesis Revisited.” SSRN . Patel, Bharti and Saoirse Kerrigan. 2004. “Hungry and Homeless: The impact of the withdrawal of state support on asylum seekers, refugee communities and the volunteer sector.”. Pettigrew, Thomas F and Linda R Tropp. 2006. “A meta-analytic test of intergroup contact theory.” Journal of personality and social psychology 90(5):751.

48

Phillimore, Jenny and Lisa Goodson. 2006. “Problem or opportunity? Asylum seekers, refugees, employment and social exclusion in deprived urban areas.” Urban Studies 43(10):1715–1736. Piller, Ingrid. 2002. “Passing for a native speaker: Identity and success in second language learning.” Journal of sociolinguistics 6(2):179–208. PISA, OECD. 2012. “Results: Students and Money: Financial Literacy Skills for the 21st Century (Volume VI).”. Plonsky, Luke and Frederick L Oswald. 2014. “How big is ?big?? Interpreting effect sizes in L2 research.” Language Learning 64(4):878–912. Portes, Alejandro and Min Zhou. 1993. “The new second generation: Segmented assimilation and its variants.” The annals of the American academy of political and social science 530(1):74–96. Portes, Alejandro and Robert L Bach. 1985. Latin journey: Cuban and Mexican immigrants in the United States. Univ of California Press. Portes, Alejandro and Rubén G Rumbaut. 2001. Legacies: The story of the immigrant second generation. Univ of California Press. Pressé, Debra and Jessie Thomson. 2007. “The resettlement challenge: Integration of refugees from protracted refugee situations.” Refuge: Canada’s Journal on Refugees 24(2). Putnam, Robert D. 2016. Our kids: The American dream in crisis. Simon and Schuster. Qureshi, Muhammad Asif. 2016. “A meta-analysis: Age and second language grammar acquisition.” System 60:147–160. Ragins, Belle Rose. 1989. “Barriers to mentoring: The female manager’s dilemma.” Human relations 42(1):1–22. Ragins, Belle Rose. 1997. “Diversified mentoring relationships in organizations: A power perspective.” Academy of Management Review 22(2):482–521. Reed, Debbie, Albert Yung-Hsu Liu, Rebecca Kleinman, Annalisa Mastri, Davin Reed, Samina Sattar, Jessica Ziegler et al. 2012. An effectiveness assessment and cost-benefit analysis of registered apprenticeship in 10 states. Technical report Mathematica Policy Research. Reichman, Nancy J and Joyce S Sterling. 2001. “Recasting the brass ring: Deconstructing and reconstructing workplace opportunities for women lawyers.” Cap. UL Rev. 29:923. Riach, Peter A and Judith Rich. 2006. “An experimental investigation of age discrimination in the French labour market.”. Richard, Doug. 2012. “The Richard review of apprenticeships.”. Richmond, A. 1988. “1988: Sociological theories of international migration: the case of refugees. Current Sociology, 36 (2), 7-25.”. Robbins, Claudia and Linnea C Ehri. 1994. “Reading storybooks to kindergartners helps them learn new vocabulary words.” Journal of educational psychology 86(1):54. Roch, GR. 1979. “Much ado about mentors.” Harvard business review 57(1):14–20. Roche, Gerard R. 1979. “Probing opinions.” Harvard Business Review 57(1):15. 49

Rolstad, Kellie, Kate Mahoney and Gene V Glass. 2005. “The big picture: A meta-analysis of program effectiveness research on English language learners.” Educational policy 19(4):572–594. Salant, Brian and Meghan Benton. 2017. “The economic integration of refugees in Canada: a mixed record?”. Salehi-Isfahani, Djavad and Mohammad-Hadi Mostafavi. 2016. “Cash transfers and labor supply: evidence from a large-scale program in Iran.”. Scandura, Terri A. 1992. “Mentorship and career mobility: An empirical investigation.” Journal of organizational behavior 13(2):169–174. Schlegelmilch, Bodo B and Carolyn Tynan. 1989. “Who volunteers?: An investigation into the characteristics of charity volunteers.” Journal of Marketing Management 5(2):133–151. Scott, Eugene. 2015. “Mayors strike back against governors in refugee fight.”. Seave, Ava. 2016. “In The Battle Of Online Language Learning Programs, Who Is Winning?”. Seshan, Ganesh and Dean Yang. 2014. “Motivating migrants: a field experiment on financial decision-making in transnational households.” Journal of Development Economics 108:119–127. Short, Deborah and Shannon Fitzsimmons. 2007. Double the work: Challenges and solutions to acquiring language and academic literacy for adolescent English language learners: A report to Carnegie Corporation of New York. Alliance for Excellent Education. Sianesi, Barbara. 2004. “An evaluation of the Swedish system of active labor market programs in the 1990s.” Review of Economics and statistics 86(1):133–155. Simich, Laura, Morton Beiser, Miriam Stewart and Edward Mwakarimba. 2005. “Providing social support for immigrants and refugees in Canada: Challenges and directions.” Journal of immigrant health 7(4):259–268. Singer, Audrey. 2012. “Investing in the Human Capital of Immigrants, Strengthening Regional Economies.” Washington, DC: Brookings Institution. Retrieved October 20:2014. Slavin, Robert E and Alan Cheung. 2005. “A synthesis of research on language of reading instruction for English language learners.” Review of Educational Research 75(2):247–284. Smith, Erica and Ros Brennan Kemmis. 2013. Towards a model apprenticeship framework: a comparative analysis of national apprenticeship systems. ILO. Sommarribas, Adolfo, David Petry, Noemie Marcus and Birte Nienaber. 2016. Resettlement and Humanitarian Admission Programmes in Europe–what works? Technical report University of Luxembourg and the European Migration Network. Statistics, National Center Education. 2015. “English language learners: Fast Facts.”. Steinberg, Sarah Ayres and Ben Schwartz. 2014. “Ethnic enclaves and the economic success of immigrants: Evidence from a natural experiment.” Center for American Progress Report . Steinmayr, Andreas. 2016. “Exposure to refugees and voting for the far-right:(unexpected) results from Austria.”. Stukas, Arthur A, Keilah A Worth, E Gil Clary and Mark Snyder. 2009. “The matching of motivations to affordances in the volunteer environment: An index for assessing the impact of multiple matches on volunteer outcomes.” Nonprofit and Voluntary Sector Quarterly 38(1):5–28. 50

Stumpf, Stephen A and Manuel London. 1981. “Management promotions: Individual and organizational factors influencing the decision process.” Academy of Management Review 6(4):539–549. Suárez-Orozco, Carola, Marcelo M Suárez-Orozco and Irina Todorova. 2009. Learning a new land. Harvard University Press. Sumnall, Harry and Angelina Brotherhood. 2012. Social reintegration and employment: evidence and interventions for drug users in treatment. Number 13 Publications Office of the European Union Luxembourg City. Takanishi, Ruby. 2004. “Leveling the playing field: Supporting immigrant children from birth to eight.” The future of Children pp. 61–79. Taylor, Eboni M, Emad A Yanni, Clelia Pezzi, Michael Guterbock, Erin Rothney, Elizabeth Harton, Jessica Montour, Collin Elias and Heather Burke. 2014. “Physical and mental health status of Iraqi refugees resettled in the United States.” Journal of immigrant and minority health 16(6):1130–1137. Topa, Giorgio. 2001. “Social interactions, local spillovers and unemployment.” The Review of Economic Studies 68(2):261–295. Torbati, Yeganeh and Mica Rosenberg. 2017. “Exclusive: State Department tells refugee agencies to downsize U.S. operations.”. Vecina, María L, Fernando Chacón, Manuel Sueiro and Ana Barrón. 2012. “Volunteer engagement: Does engagement predict the degree of satisfaction among new volunteers and the commitment of those who have been active longer?” Applied Psychology 61(1):130–148. Vesselinov, Roumen and John Grego. 2012. “Duolingo effectiveness study.” City University of New York, USA . Viator, Ralph E. 2001. “The association of formal and informal public accounting mentoring with role stress and related job outcomes.” Accounting, Organizations and Society 26(1):73–93. Vigdor, Jacob L. 2010. from Immigrants to Americans: The Rise and fall of fitting In. Rowman & Littlefield. Waldinger, Roger and Michael I Lichter. 2003. How the other half works: Immigration and the social organization of labor. Univ of California Press. Wallace, Jean E. 2001a. “The benefits of mentoring for female lawyers.” Journal of Vocational Behavior 58(3):366–391. Wallace, Jean E. 2001b. “The benefits of mentoring for female lawyers.” Journal of Vocational Behavior 58(3):366–391. Weber, Andrea and Helmut Hofer. 2004. “Employment effects of early interventions on job search programs.”. Widerquist, Karl. 2005. “A failure to communicate: What (if anything) can we learn from the negative income tax experiments?” The Journal of Socio-Economics 34(1):49–81. Wong, Lok Ping, Wing Hong Chui and Yan Yuen Kwok. 2011. “The volunteer satisfaction index: A validation study in the Chinese cultural context.” Social indicators research 104(1):19–32. Yu, Soojin, Estelle Ouellet and Angelyn Warmington. 2007. “Refugee integration in Canada: A survey of empirical evidence and existing services.” Refuge: Canada’s Journal on Refugees 24(2). Zhou, Min. 2001. “Straddling different worlds.” Ethnicities: Children of immigrants in America pp. 187–227. 51