International Journal of Evidence Based Coaching and Mentoring
2019, Vol. 17(2), pp.122-137. DOI: 10.24384/3d5c-w176

Academic Paper

As Iron Sharpens Iron: A mentoring approach to labour market integration for humanitarian migrants

Akiva Weiss (Erasmus University Rotterdam and ITMO University Saint-Petersburg)
Marina Tulin (Amsterdam Institute for Social Science Research, University of Amsterdam)

PDF

Introduction

Recent surges in immigration to European Union (EU) countries has increased interest in labour market integration initiatives. Gainful employment is critical for new waves of humanitarian migrants −refugees, asylum seekers awaiting status determination, and those falling under subsidiary protection− to integrate into their host country’s job force. In fact, labour market integration is frequently cited as the bedrock of socio and cultural integration (Sniderman, Hagendoorn, and Prior 2004) and is especially critical for countries whose low fertility rates requires immigrants to occupy higher skilled jobs, provide benefits for retirees, as well as have a decisive role in the cultural relevance of customs (Coleman 2006). Although significant structural measures aiming to facilitate labour market integration have been taken in many EU since 2015, several impediments remain.

Humanitarian migrants still face significant linguistic, informational, and social barriers toward gainful employment. Linguistic barriers often force immigrants to rely on their own social networks resulting in less opportunities for growth in well-paid sectors (Hatami & Weber, 2012) The role of information permeates a range of social cues and institutional knowledge which is often critical to extending positive signals during the hiring process, as well as optimizing educational resources (Ermini et al., 2017). The latter is especially important since humanitarian migrants have less higher education than their native counterparts and even those with requisite education and prior professional experience frequently require upskilling (Degler Liebig, and Senner 2017), and additional licensure (Klingler & Marckmann, 2016). Facing hurdles to translate home-country achievements into host-country credentials is likewise common (Desiderio, 2016).  Attaining requisite education is hindered by a range of factors such as entrance qualifications which often have language requirements. Compounding informational and societal barriers is social isolation which may adversely affect psycho-emotional states and motivation to invest in a host country’s unique culture (Li, et al., 2016). These informational and social correlates are not so much additive as they are interdependent (Berg, 2018); facilitating a meaningful intervention thus requires addressing the intersectionality of numerous social, cultural and country-specific bureaucratic matters.

Labour market integration may be envisioned as a series interconnected layers including social capital, language, skills, meaningful contact with the local population. Social capital helps facilitate connections often crucial towards developing social ties and networks that reduce uncertainty and informational asymmetries, enhanced language abilities broaden workforce opportunities, whilst skills sharpen employability. Moreover, meaningful contact from mentoring may mitigate value and role conflicts for newcomers, both often critical towards inculcating a positive sense of self and emotional well-being (Awujo, 2016).

The aim of this paper is to test whether a particular integration strategy, mentoring, is able to mitigate labour market impediments resulting from informational, educational, linguistic or societal factors. Mentoring can be defined as a type of developmentally-oriented coaching in which an experienced mentor acts as a role model (Eby et al., 2008). Mentoring is often separated into two broad categories: instrumental, which targets a specific goal, and socio-psychological, which nurtures emotional needs. Both categories help to leverage the mentee emotionally, socially, and at times professionally. When asylum seekers arrive without linguistic or familial roots, nor the tools to navigate institutions and culture-specific social cues, mentors can use their knowledge to bridge informational and social inequalities.

In the literature, mentoring has been shown to target many of the disadvantages shown to adversely affect labour market integration. Mentors positively impact the social skills needed to succeed in the labour market (Lerner 2005) career outcomes and may increase intrinsic job rewards (McDonald and Lambert 2014). At the structural level mentoring has been shown to affect age at immigration (Corak 2012) and income inequality (Schneeweis 2011). Qualitative studies have also provided insight into the processes through which mentoring positively affects socio-emotional capacity and secure identities (Rhodes 2005, 2008;) both of which, we argue, will impact labour market outcomes. The mentoring literature has also shown a benefit to newcomers: Månsson & Delander (2017) find an increase in yearly income for mentored male refugees in Sweden, whilst Liebig (2007) suggests that mentoring is an efficient way to develop social networks crucial for refugees finding employment in Denmark and self-employment in Australia.

Whilst the positive effects of mentoring on socio-economic outcomes are well-established, to the authors’ knowledge there have been no empirical studies linking mentoring to Arab Spring humanitarian migrant labour market outcomes. Addressing this gap, this paper adds value to the literature by studying the impact of mentoring, among a particular subset (Arabic speaking humanitarian migrants) in three European Union countries. Respondents were asked whether they had a mentor, whether that mentor aided in bridging institutional barriers, inculcating meaningful interpersonal contacts, and whether the mentor was integral towards attaining gainful employment. The results show that mentoring has a positive effect on the labour market participation of humanitarian migrants, especially when the mentoring program includes an educational component, and when mentors help mentees create meaningful social connections with natives of the host country.

The remainder of the paper arrives at these conclusions in the following sequence: a literature review gives birth to hypotheses concerning mentoring’s effects on social capital, the role of information, and identity. Following the literature review, the empirical design is described, and the results presented. Finally, a discussion of the results is presented alongside limitations and suggestions for forward research.

Terminology

Refugees are humanitarian immigrants who fall under the 1951 Geneva Refugee Convention definition of persecuted. Determination of status often takes several months and may take
years. In popular parlance, those awaiting status determination are called asylum seekers.
In addition, many humanitarian immigrants who are not afforded refugee status are given
subsidiary status and remain in their host country. Integration is a concern for all types of
humanitarian immigrants, whether being granted refugee status or not. We therefore test humanitarian immigrants in general, and control for type (i.e., granted refugee status, awaiting status) in the analyses.

Literature and Hypotheses

To be considered a successful labour market intervention strategy mentoring must positively affect the determinants, largely derived from the Labour Economics, Educational, Applied Psychology, and Acculturation literature, shown to aid effective integration. As theorized in the introductory section, mentoring may simultaneously address these determinants thereby avoiding a differential model of integration whereby inclusion at one level fails to permeate to a wider diffusion of social and political opportunities. We explore this simultaneity, revealing how the net return of mentoring humanitarian migrants may be realized through fostering instrumental information, social capital, and inculcating interpersonal connections between humanitarian migrants and the host population.

Labour market integration

Integration is often conceptualized as hierarchical process whereby economic assimilation becomes the basis for social and educational: if one contributes fiscally, they are valuable to society and have earned the right to be valued (Sniderman et al. 2004). In macroeconomic terms, effective economic integration is frequently measured by net fiscal impact (NFI): the difference between tax-based contributions minus the cost of services and benefits received.

Research has not found a substantial impact from humanitarian migrants, positive or negative, on the NFI. A 2016 European Commission report, for example, found the macroeconomic impact of refugees calculated as moderate to negligible, with short-term rises in public spending, whilst a recent IMF study found a small increase in GDP growth from refugees over the short-term, with uncertainty surrounding longer term effects on the economy (IMF 2016). Dustmann et al., (2016) finds that employment probabilities of refugees increase by year, in Europe, although the refugee-native employment gap loses its statistical significance only at between 15 to 19 years and is dependent on age at migration, pre-existing institutional conditions, and family structure.

What is clear is that employment is widely considered the most significant factor catalyzing durable integration and is crucial for the net fiscal contribution of humanitarian migrants to be positive (OECD 2013). What then are the barriers towards humanitarian migrants obtaining gainful employment? In the literature three factors stand out: lack of social ties which hinder employment opportunities; asymmetries in knowledge of educational infrastructure; and insufficient meaningful inter-ethnic contact providing opportunities for a sense of belonging and development of ego self within host country society. The following section address these three barriers to securing a job. We begin with the role of social capital.

Social capital

Social capital, defined here as the information, trust, and norms of reciprocity inhering in one’s social networks (Woolcock and Narayan 2000) may lower transaction costs by reducing the uncertainty employers face whilst screening applicants (Smith 2005), as well as make job searching more efficient for applicants (Mouw 2003). Lin (2017) extrapolates the fourfold pathways whereby social capital may reasonably impact labour market outcomes: through the flow of information, influence of agents, social credentials and reinforcement. Working under imperfect information hiring managers can use social ties and social cues as a signal that a potential employee is hirable thereby reducing risk and the concomitant transaction costs associated with recruiting and hiring. In addition, social credentials may generate added value beyond the personal capital of job-related duties, value that may multiply as clients recognize the emotional support and entitlement that such ties suggest (Lin 2017).

For humanitarian migrants, lack of social capital is frequently the result of arriving in a country without an extended social network. This absence may adversely affect sense of belonging, access to institutions, and impede landing higher paid and prestigious jobs (Lin 2002). Whilst some humanitarian migrants have connections from diaspora communities this rarely provides the same benefit as inter-ethnic connections (Lancee 2012). Overcoming these disadvantages frequently necessitates not merely bridging inequalities grounded in historical and institutional circumstances but actively promoting interactions with the host population (Verhaeghe et al., 2013). These interactions may increase the flow of information, job contacts (Goldthorpe 2007) and wages (Edin et al. 2003).

A strand of scholarship has identified best mentoring practices for inter-cultural mentoring for immigrant newcomers (Reeves, 2017), particularly those that foster a positive entrepreneurial climate through inculcating social ties (Jordan, 2018), accelerate integration through in-country networks (Oberoi, 2016) and foster social affiliations leading to increased trust and reduced negative stereotypes about ability (Camras, 2004). Supporting these recommendations, Raithelhuber (2019) finds that mentoring for unaccompanied refugee minors may lead to “substantial social participation at central institutions” which may be critical given the frequent bias towards native applicants in European markets (Kaas & Manger, 2012). Moreover, recent European Commission proposals have promoted employment coaching and mentoring as an active labour market policy to raise social and knowledge capital (EC 2016).  On the other hand, Allen (2009) found that sponsorship’s positive effect on social capital had a mixed-gender effect on refugee earnings.

The effect of social capital appears to vary with duration of mentoring, age and gender of mentee, as well as state of the labour market. Allen’s sample population, for instance, was derived from Somali and Sudanese refugees in a single U.S. city at a time when refugee integration was not a hot issue in the United States. In contrast, our sample consists of Arabic-speaking refugees from multiple cities, most of whom had arrived during the Arab Spring migration when integration programs aimed at promoting social capital and labour market success were a priority for our sample countries: Germany, Italy, and Greece. Based on its potential to foster social capital, frequently shown to improve labour market outcomes, the first hypothesis is formulated as follows:

H1: Individuals who were mentored have a higher probability of employment than those who were not mentored.

Mentoring and education

Empirical evidence points to education playing a significant role in labour market integration as well as overall growth, societal prosperity, social cohesion, and inter-generational mobility (Woessmann, 2016). Many of these conclusions cascade from human capital theories that link education to earnings (e.g., Mincer 1974), or signal theory, where skills developed in the classroom transmit positive signals during job interviews conducted under imperfect information (Ermini et al., 2017).

Informational asymmetries play a formidable role in educational achievement. They impact how immigrants navigate the education system, engage their child’s learning, and enable parent-teacher coordination, thereby playing an active role in their child’s achievement. Because parental support is often needed for success in the classroom, lack of first-hand knowledge, frequently a result of language capacity or insufficient diffusion of domain-specific knowledge (Kanno and Varguese 2010) prevents equalization with the local population. To education’s positive impact on employment, an integration program needs to address institutional factors and information barriers influencing educational achievement.

Institutional factors

Institutional factors typically refer to tracking and resource allocation. Tracking −the separation of pupils by curriculum based on academic ability− often leads to educational inequality and imparts a more significant role to family effects, particularly in systems with early tracking (Borgna and Contini 2014). In a landmark cross-country study, Crul and Schneider (2010) show that highly stratified systems of tracking impact whether migrant groups receive tertiary education. They find that many second-generation Turks receive a college or university education in France (39.6%), Sweden (37.1%), the Netherlands (25.6%), but in Germany, a country with early tracking and close to impermeable mobility, only 3% do. Alba and Foner (2014) suggest that Germany’s highly stratified system of tracking and the United States’ decentralized funding account for the two largest disadvantages for low status immigrant children.

From an intervention point of view, more favorable institutional arrangements may facilitate educational achievement and its concomitant labour market empowerment. It may also result in less dependency on family-level factors (Schnell, et al., 2015). Unfortunately, institutional changes such as centralized funding schemes or tracking reforms are uncommon, and when implemented, often miss their target outcome (Woessmann 2016). Mentoring then may compensate for educational inequalities that Member States are not willing to address structurally (i.e., removing early tracking) or that fail to yield results, by mitigating individual-level informational barriers hindering educational achievement (Kilburg 2007).

Mentoring and information barriers

Mentoring has often shown its ability to positively impact informational asymmetries in educational institutions particularly the educational outcomes for disadvantaged youth, a population subset underrepresented in the labour market (Rodríguez-Planas 2012). Concrete examples come from mentoring programs providing tailored support targeting structural inequalities (i.e., differentiation in educational resources) and family-level informational asymmetries. One such program is SAMIE, which has successfully targeted educational integration of unaccompanied minors in France. Rock Your Life, consisting of over 40 papers within Germany and Switzerland pair underprivileged youth and a college student for two years, aiming to navigate the rigid German education system and increasing social mobility: transmission of know-how between choices in the education system aids in avoiding hasty decisions over whether to choose a vocational or tertiary track (Buis 2012). In another example, mentoring projects under the umbrella, European Network of Educational Support Projects (ENESP), such as MentorprojectSKC and Junge Vorbilder have offered level-appropriate guidance often in educational settings. The Amsterdam based MentorprojectSKC, for instance, has forged strong collaborative ties with primary and secondary schools, focusing on preparing mentees socially and academically for transitions.

As for adult education, a mentor that compliments job or language training programs may assist the mentee with host-country specific skills (i.e., gestures, linguistic or cultural cues) that were shown to decrease the uncertainty interviewers encounter when facing imperfect information (Kogan 2016). Mentoring may also decrease stress and increase quality of life for adult humanitarian migrants having a concomitant impact on their employability (Goodkind 2005). Taking into account both the benefit to youth and adults, the second hypothesis is derived forthwith:

H2: Individuals who received mentoring that contained an educational component have a higher probability of employment.

Mentoring and meaningful social connections

For over half a century the social-psychology cannon has drawn clear links between identity, society, and meaning. On meso and individual levels, a strong sense of self promotes self-confidence, which may lead to positive labour market outcomes (Chen et al. 2017). Yet in the context of integration, the link between identity and society takes upon an added dimension.  Having a strong ethnic identity is related to negative labour market outcomes, particularly when there is neither assimilation nor integration with the dominant culture (Nekby and Rodin 2010). We argue here that it is not that mentoring helps job seekers’ identification in the deeper sense of self-identity, but through an ongoing process of being accepted into society. Being accepted is the initial layer of identification that continues along the migrant experience. Psychological and sociocultural adaptation are impacted by a sense of belonging and critical to functioning within a new culture.  Of primary significance to the integration-identity nexus then is how interpersonal interactions and perceptions of discrimination impact identification. Over and over research has found a negative correlation between positive integration outcomes and discrimination (Jasinskaja‐Lahti et al., 2009). Per human capital theory, discrimination disincentivizes investing time learning a new language and specific job skills (Dancygier and Laitin 2014) and may stunt the self-reflection necessary to either assimilate or integrate into society (Fokemma and Haas 2015).

Mentoring is unusually helpful in this regard: because naturally occurring inter-ethnic contact is often affected by structural constraints, mentoring may bring refugees together in situations otherwise unlikely. Here we turn to the contact hypothesis, a sociological theory positing that diverse interactions reduce out-group hostility most effectively when there is a cooperative environment among equals and promoted by authorities (Allport 1958). If interactions lead to less discrimination, they may also incentivize many humanitarian migrants to invest in location-specific capital. Research measuring the quality of interactions, as opposed to using census data to measure interactions (i.e., Alesina and La Ferrara, 2002) has shown positive results from immigrant-native interactions. Ellison and coauthors (2011) show that inter-contact can lead to less punitive and more empathetic attitudes towards immigrants. Grey and Woodrick (2005) show that proximity with the native population may lead to an inclusive environment for immigrants, whilst Okamoto and Ebert (2016) reveal how this is true in consonance with an increase in visibility, politically and demographically. On the one hand this may mean that when environment promotes multicultural policies, individuals will be respective towards inclusionary attitudes, on the other, that political, institutional, and environmental (i.e., socialization) opportunities for multicultural interaction are needed to be in place for attitudinal shifts towards inclusion to transpire.

The process of ‘contact’ may also clarify populist views against foreigners, potentially dispelling stereotypes formed from lack of contact with an out-group (Oberoi et al., 2016). Proximity also allows for humanizing metrics through nuanced accounts that are invisible in large-scale statistics. Mentoring offers such opportunities when project designs incorporate inter-ethnic group social activities such as sports, clubs, and informal gatherings (Pryce et al., 2018). Giovani al Centro, in an example, one of ENESP’s projects, promotes intercultural friendships in Turin through providing afterschool support for teens. In another example, both the Austrian Mentoring for Migrants and Diversity as Chance project has provided regular interpersonal contact with refugees seeking jobs or apprenticeship programs. So that, whilst institutional structures are frequently resistant to change, identity is non-static: it may be localized, shift form over time, and have an impact at all stages of the migrant’s job seeking.

H3: Individuals who received mentoring that facilitated meaningful interpersonal
interactions have a higher probability of employment.

Types of mentoring

Although we hypothesize that mentoring as a monolithic indicator has a positive impact on integration, there exists an extensive body of literature disaggregating mentoring by attributes, typically including the respective ages of the mentor and mentee, the frequency of contact, language capacity of mentor, and type of mentoring (Casey and Dustmann 2010). On the latter point, there are several subcategories of mentoring such as natural mentoring (formed without formal structure) and youth-initiated mentoring; it may be that different types of mentors are needed during different life phases (Fruiht and Wray-Lake 2013). Whilst the psycho-emotional aspects of mentoring might benefit eventually from formal and informal mentors, integrating humanitarian migrants would seemingly require a structured mentoring program to use its organizational abilities, industry know-how, and monitoring capacity to coordinate the institutional-level barriers that newcomers face. As such, the level of organization and commitment needed to satisfy the first three hypotheses would intuitively require a structured program. By that token, the final hypothesis is formed:

H4: Individuals who received formal mentoring have a higher probability of employment than individuals who received mentoring that was naturally occurring.

Methodology

Sampling strategy

Data were collected by the first author of this paper from February 2017 to August 2017 in Greece, Italy, and Germany. A pilot was undertaken during the first three days of February. On average survey completion took 17 minutes with a pen. Data collection was conducted in three places: mosques during Friday prayers (51 mentored), public parks (8 mentored), and outside public transportation stations with high concentrations of migrants (9 mentored). The choice of Friday prayers was done to ease potential socio-economic bias. That is, it is possible, but by no means certain, that unemployed migrants would be more available to spend time in parks then go to a mandatory religious service. The Friday prayer (salat jumah) is the only religious service that is mandatory for all Muslims to attend regularly no matter what their labour market status. The final sample consisted of 295 Arabic speaking humanitarian migrants of which 68 were mentored. In addition, following the questionnaire a short interview was administered. Interviews averaged 6 minutes and asked respondents three questions: to elaborate on institutional barriers towards labour market integration; social barriers; as well as the ways that mentors interacted with these barriers.

Dealing with a vulnerable population carries several ethical responsibilities. First, the study was approved by the Erasmus University Rotterdam to meet the University’s ethical standards. Respondents were informed about the purpose and procedure of the study and provided written consent and data collected in a non-invasive manner that did not put respondents at psychological or physical risk. After collection, anonymized data was stored on password protected hard drives only accessible to the author. Lastly, the author collecting data did not personally know any of the respondents. No compensation was offered for participation and the public nature of selecting and surveying respondents helped minimize power differentials.

Data and methodology

Refugees provided data that captured their employment status, and a range of socio-demographic characteristics.  In this study, the dependent variable in all analyses was a dichotomous employment status variable indicating whether a humanitarian migrant was unemployed or employed. We considered four predictors of employment which we tested in four separate models. The first predictor captured whether respondents participated in a mentoring program (yes; no). The second predictor captured whether the mentoring program contained an educational component (no mentoring; mentoring with educational component; mentoring without education component). The third predictor measured whether the mentor helped build meaningful friends (no mentoring; mentor helped make friends, mentor did not help make friends). The fourth predictor measured whether mentoring was formal or naturally occurring (no mentoring; formal mentor; naturally occurring mentoring). Whilst ‘educational component’ can refer to a gamut of mentoring-related aspects, post-questionnaire interviews suggest that for the large majority of respondents, mentors worked with teachers and school administrators to explain course content, pathways towards graduation, and help with homework. In short: knowledge of educational institutions, language capacity, and best practice learning habits.

Based on these observational data it is unknown whether the humanitarian migrants were randomly assigned to these mentoring programs or not. The advantage of random assignment is that – on average – there is no difference in the personal characteristics of humanitarian migrants who did participate in a mentoring program and those who did not. Random assignment allows a direct estimation of the effect of mentoring, which is not influenced by respondent characteristics. For example, those who are higher educated might be more likely to participate in a mentoring program. Because of their education they are also more likely to find employment, whether they participated in a mentoring program or not. Because we did not actively assign respondents to different interventions, we were unsure whether random assignment occurred. We therefore made the conservative assumption that random assignment did not occur, and we relied on a matching technique that allowed us to approximate a situation where random assignment did occur.

To this end, we employed Coarsened Exact Matching (CEM). This is a matching technique used to estimate treatment effects and make causal inferences based on observational data (King and Nielsen 2016; King et al. 2011). In the context of this paper, the term ‘treatment effect’ refers to the effect of a mentoring intervention. Even though we did not assign the mentoring intervention to our respondents, ‘treatment effect’ is a commonly used term in matching analysis, because it allows testing the effect of an intervention as if it was an active and randomly assigned intervention. The first step in CEM analysis is to select a number of covariates on which to match the sample of treated and untreated respondents. This is necessarily to account for the possibility that humanitarian migrant’s background characteristics affect both their probability of receiving mentoring and the probability of being employed. In other words, it accounts for selection into mentoring programs (given selected covariates), and therefore reduces endogeneity problems. CEM enables us to approximate a counterfactual for treated individuals, meaning that we can estimate what the employment probability for treated individuals would have been if they had not received the treatment.

A range of respondent characteristics were available on which the sample could potentially be matched, namely gender (male; female), age (15-18; 19-22; 23-30; 31-40; 41-50; 51-65; 66-100) educational background (completed graduate degree; completed undergraduate degree; some university; not finished high school; other), country of origin (Middle Eastern; African), religion (Muslim; Christian), legal status (waiting for determination; refugee status; no legal status), the time spent in the current country (< 6 months; 6-12 months; > 12 months), and the country of mentoring (Greece; Italy; Germany). Descriptive statistics of these respondent characteristics as well as the dependent variable and treatment variables are displayed in Table 1.

Table 1: Descriptive statistics of and respondent characteristics and mentoring programs

 N% N%
Dependent Variable  Covariates  
Employment Status  Mentoring country  
   Unemployed13044.22   Greece6355.25
   Employed16455.78   Italy7324.75
Treatment Variables    Germany5920.00
Mentoring  Gender  
   Not mentored22776.95   Female8328.14
   Mentored6823.05   Male21271.86
     With educational component4414.92Age  
     Without educational component258.47   15-183110.51
     Mentor helped make friends5719.32   19-226221.02
     Mentor did not help make friends124.07   23-307124.07
     Mentoring occurred naturally3210.85   31-405719.32
     Formal mentor3612.20   41-504715.93
    51-65217.12
   66-10062.03
Education  
   No Answer5920.07
   Completed Graduate Degree4314.63
   Some University5619.05
   Not Finished High School4615.65
  Completed Undergrad Degree9030.61
Country of origin  
   Middle Eastern26991.19
   African268.81
Religion  
   Muslim26890.85
   Christian279.15
Legal status  
   Waiting for determination10033.90
   Refugee status12442.03
   No legal status7124.07
Months in country  
   < 6 months13445.58
   6-12 months8027.21
   > 12 months8027.21

Note: N = 295

Because of the relatively small sample size, it was impossible to exact match the group of treated and untreated respondents on all available respondent characteristics. We therefore selected the most important covariates, which were covariates expected to be associated with both the outcome (i.e., employment) and the treatment (i.e., mentoring program). Based on the literature review, we expected educational background, legal status, and time spent in country to fulfil these criteria. We performed a series of logistic regression analyses predicting the probability of participating in the different types of mentoring programs based on the selected covariates (see Table 2). As expected, time spent in country consistently predicted participation in mentoring programs. Those who were in a country for longer generally had a higher probability of participation in a mentoring program. Legal status and educational background were less consistently related, at least according to conventional levels of significance (i.e., alpha level of .05). However, several coefficients were marginally significant and the overall models explained between 9% and 13% of the variance in the probability of participating in the different mentoring programs. Given that participation in mentoring programs depended in part on these covariates, the sample can be considered unbalanced with regard to these relevant covariates.

Table 2: Results of logistic regression analyses estimating the probability of entering different mentoring programs predicted by educational background, legal status, and months spent in the receiving country

 (1)
Mentoring
(2)
Mentoring without education
(3)
Mentoring with education
(4)
Mentor did not help make friends
(5)
Mentor did help make friends
(6)
Mentoring occurred naturally
(7)
Mentor was formally assigned
Education
Not Finished High School  RefRefRefRefRefRefRef
Completed Graduate Degree3.05
(1.75)
3.79*
(2.46)
0.61
(0.58)
2.60
(1.62)
3.17
(3.74)
2.04
(1.58)
2.97
(2.15)
Some University1.83
(1.06)
2.29
(1.51)
0.53
(0.50)
2.07
(1.28)
0.78
(1.12)
2.37
(1.78)
1.14
(0.91)
    Completed Undergraduate2.07
(1.08)
1.65
(1.02)
1.51
(1.07)
2.24
(1.26)
1.06
(1.32)
1.47
(1.05)
2.32
(1.56)
    Other2.04
(1.15)
0.78
(0.58)
2.71
(1.94)
1.62
(1.01)
3.86
(4.32)
3.49
(2.52)
0.75
(0.63)
Legal status
Waiting for determinationRefRefRefRefRefRefRef
Refugee status0.62
(0.23)
0.89
(0.39)
0.28*
(0.15)
0.62
(0.25)
0.72
(0.50)
0.86
(0.42)
0.50
(0.24)
No legal status0.65
(0.29)
0.73
(0.42)
0.50
(0.32)
0.75
(0.37)
0.24
(0.26)
0.54
(0.34)
0.84
(0.49)
In country for
    < 6 monthsRefRefRefRefRefRefRef
    6-12 months3.38**
(1.41)
4.38**
(2.31)
2.14
(1.27)
3.47**
(1.63)
1.68
(1.24)
2.48
(1.42)
3.37*
(1.84)
    > 12 months7.14***
(2.95)
6.88***
(3.52)
4.68**
(2.73)
9.45***
(4.27)
0.65
(0.61)
9.83***
(5.23)
2.79
(1.54)
N294269250282237258262
pseudo R20.150.170.140.180.080.100.20

To reduce imbalance between the groups of treated and untreated migrants, we exact matched them based on the selected covariates, namely educational background, legal status, and time spent in country. CEM produces a balanced sample by pruning the data such that observations that cannot be exact matched are dropped from subsequent analyses. Since the balanced sample does not differ with regard to covariates – participation in a mentoring program is the only difference between the two groups - it is unnecessary to control for covariates in subsequent analyses (King et al. 2011, 2016). Effect sizes of simple mean comparisons can be interpreted as average treatment effects on the treated. Covariates on which the sample was matched can, however, be included in order to show the effect of the covariates on the outcome of interest.

Results

We performed a series of logistic regression analyses on the balanced sample estimating average treatment effects on the treated with different mentoring programs as treatment and probability of employment as outcome. Results including estimates of the covariates are displayed in Table 3. Results that excluded covariates indeed produced the exact same pattern of results and can be obtained upon request. In our presentation of results, we follow conventional levels of statistical significance, namely p < .05. Results with p-values of .05 or higher are reported as non-significant results.

Table 3: Logistic Regression Analyses Estimating Average Treatment Effects on the Treated with Different Mentoring Programs as Treatment and Probability of Employment as Outcome

 (1)(2)(3)(4)(5)(6)(7)
Mentoring2.50*
(0.92)
Mentoring without education2.60
(1.42)
 
Mentoring with education 2.70*
(1.25)
Mentor did not help make friends0.33
(0.23)
 
Mentor helped make friends 5.67***
(2.72)
Mentoring occurred naturally1.62
(0.80)
 
Mentor was assigned 3.54*
(1.84)
Not Finished High SchoolRef.Ref.Ref.Ref.Ref.Ref.Ref.
Completed Graduate Degree1.41
(0.84)
2.37
(2.49)
1.51
(0.94)
1.71
(2.05)
1.35
(0.88)
3.30
(2.32)
0.52
(0.43)
Some University1.32
(0.81)
8.75
(11.12)
1.15
(0.74)
0.59
(0.97)
1.66
(1.14)
1.98
(1.82)
0.64
(0.50)
Completed Undergraduate Degree1.47
(0.81)
10.45*
(9.63)
1.64
(0.96)
2.25
(4.09)
1.74
(0.99)
2.79
(1.73)
0.81
(0.63)
Other0.43
(0.25)
1.74
(1.18)
0.35
(0.27)
0.42
(0.39)
0.61
(0.40)
0.76
(0.60)
0.18*
(0.14)
Waiting for determinationRef.Ref.Ref.Ref.Ref.Ref.Ref.
Refugee status0.88
(0.38)
1.28
(0.76)
0.74
(0.39)
0.79
(1.18)
0.82
(0.36)
1.72
(1.37)
0.72
(0.36)
No legal status0.48
(0.26)
0.91
(0.73)
0.75
(0.45)
1.00
(.)
0.72
(0.43)
1.25
(1.62)
0.38
(0.22)
In country for < 6 monthsRef.Ref.Ref.Ref.Ref.Ref.Ref.
In country for 6-12 months1.48
(0.76)
0.78
(0.49)
2.30
(1.48)
2.66
(2.49)
1.82
(1.12)
1.06
(0.93)
1.71
(1.00)
In country for > 12 months1.89
(0.92)
0.53
(0.39)
2.14
(1.29)
1.00
(.)
2.51
(1.33)
1.20
(1.28)
2.29
(1.25)
Observations22312719780203151165
Pseudo R20.090.120.070.110.110.070.14

Note: Coefficients presented are odds ratios. Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001

Regarding the first hypothesis (H1), we found evidence that mentoring has a significant and positive average treatment effect on the treated. Those who received mentoring were 2.5 times more likely to be employed compared to those who did not receive any mentoring (see Model 1). Considering the second hypothesis (H2), results showed that mentoring with an educational component had a positive and significant effect on the probability of employment. Those who received mentoring with an educational component were 2.7 times more likely to be employed after receiving this treatment (see Model 3). Those who received mentoring without an educational component were not significantly more likely to be employed (Model 2). Even though the effect size was considerable, namely odds ratio = 2.6, the standard error was relatively high, suggesting that there was a lot variance with regard to the effectiveness of this mentoring strategy (this may also be due to small sample size). Overall, these results lend support to Hypothesis 2, suggesting that mentoring with an educational component is a more successful treatment than mentoring without an educational component.

The third hypothesis (H3) suggested that mentoring that helps form meaningful social connections with locals will be more successful in improving employment status than mentoring that does not help make meaningful connections. The results suggest that this is the case. Migrants who participated in mentoring programs that did not help make local friends did not increase their chances of being employed (Model 4). However, mentoring programs that did help make local friends, showed significant increases in their employment probabilities. Their chance of being employed was 5.67 times larger after participating in a mentoring program that helped them make local friends (see Model 5).

Finally, the fourth hypothesis (H4) predicted that mentors that were formally assigned would be more successful in improving their mentees’ employment status than mentors that naturally emerged. The results support this hypothesis by showing that programs where mentors were formally assigned improved the employment status of their mentees by 3.54 (see Model 7). In contrast, mentoring that occurred naturally showed lower increases in employment probabilities, namely an increase by factor 1.62, which was not statistically significant (see Model 6). 

Discussion

Based on its ability to foster social capital, bridge informational asymmetries and inculcate meaningful inter-ethnic connections, this paper hypothesized that mentoring would have a positive impact on gainful employment. These predictions are derived from several strands of literature that show how increasing sense of belonging through building meaningful relationships increases cohesion, as well as from signal theory that shows how filling gaps in location-specific information may engender positive signals at job interviews thereby lowering that gap employers face in the screening process. The results confirmed these predications. Respondents who were mentored showed an increased probability of being employed, and this was particularly the case when there was an educational component to the mentoring or when meaningful interpersonal connections were formed. When the mentor was not assigned these results lost their significance.  From a practitioner’s optic this may mean additional professional training in salient local-level institutional features, as well as mentor-mentee matches with similar extracurricular interests that can accelerate meeting likeminded locals in social events.

Whilst previous research sets the stage for our predictions, it is not intuitive that mentoring would have a positive impact on labour market outcomes. Previous studies applying big (Escudero 2018) and country-specific data (Saniter and Siedler 2013) have frequently revealed positive impacts on job search assistance, employment incentives, and labour market impacts from counseling, respectively. A positive effect has likewise been documented with immigrant-individualized training with a language component (Sarvimäki and Hämäläinen 2016). However, these effects may very well depend on the type of training (Thomson et al. 2013). Moreover, despite papers showing positive effects, other papers have found uncertainty in the integration policy-labour market outcome nexus (Bigili et al. 2015). Research has also found a negative impact on counseling and labour market training, due to negative signals from job search assistance (Kogan 2016). Whether negative signals and adverse selection negate the positive effect from a proposed mentoring intervention remains open.

As discussed in the introduction and literature review, integration efforts are frequently impacted by country-level policies (i.e., early tracking). We hypothesized that mentoring would mitigate these information asymmetries. In the empirical analysis, there were no significant country-level differences in mentoring’s ability to mitigate informational asymmetries. This may be due to the small sample size, as several post-questionnaire interviews expressed varying degrees of comfort with country-level bureaucratic barriers and the mentor’s help bridging them.

Limitations and future research

Self-selection into a mentoring program is a concern. Participants who already possessed the skills that employers deem valuable might have the initiative to participate in a mentoring intervention, or even the survey itself. Gathering participants from three distinct public arenas attempts to attenuate these selection effects. Out of 306 individuals asked to participate in the survey, 9 declined, 8 of which were Kurdish with no other similar observable characteristics.

Whilst mentoring may aid integration on several fronts, nevertheless negative cases may arise. These include exposure to negative peer behavior and breaking bonds with parents (Rodriguez Plantas 2014, 2012). Furthermore, in meta-analyses, at-risk youths have yielded tepid, albeit positive, results from mentoring (Dubois et al. 2011; Rhodes 2008). Since many humanitarian migrants would likely fall into the at-risk category, at least initially, these findings are of some concern.

A closer look suggests that the negligible or potentially negative effects may simply be a matter of mechanism design. Mitigating adverse effects may simply be a matter of how a mentoring program is formed. On the issue of commitment length, for example, Rodriguez-Plantas (2014) suggests that instances of statistically insignificant positive impact might be derived from early closures. Supporting this position is Zilberstein and Spencer (2014) as well as Grossman and Rhodes (2002), the latter finding that the positive impact of mentoring is significantly stronger if the mentor-mentee match lasts at least one year. Mechanism design, of course, must temper the underlying reasons for early termination. Avoiding negative side effects requires addressing correlates of positive program-based mentoring during the mechanism design stage and adjusting for unexpected situations.

Future research should be undertaken to pinpoint the mechanisms by which mentoring mitigates unemployment. Researchers should further unpack categories such as social capital and institutional access to pinpoint the precise enabling mechanisms. This might take the form of an additive index of different levels of institutional access or matching skill level to job acquired. Moreover, larger samples should be investigated to provide more power to the data analysis. In addition, how mentoring interacts with country-level integration policies is a fruitful area for future research. Fast tracking integration programs before asylum claims have been fully processed, fast-tracking in-demand skills, and concentrating on host language training have proven successful strategies in Scandinavian countries since 2015 (Joyce, 2019). Whether an ecosystem connecting these policy initiatives with mentoring remains unexplored.

Conclusion

Whilst the numbers of humanitarian migrants reaching Europe has plummeted dramatically from 2017 to 2019, labour market integration remains a central issue for EU member states. In this paper we investigated whether mentoring in Greece, Italy, and Germany had a positive impact of the employment of Arabic-speaking humanitarian migrants. We hypothesized that mentoring would foster the location-specific capital needed to bridge informational asymmetries, form meaningful connections, and send positive signals to employers. Our results confirm these predictions and place a burden of action upon State and EU-level officials to further investigate State and supranational level interventions.

References

Alesina, A. and La Ferrara, E. (2002) 'Who trusts others?', Journal of public economics, 85(2), pp.207-234.Allen, R. (2009) 'Benefit or burden? Social capital, gender, and the economic adaptation of refugees', International Migration Review, 43(2), pp.332-365.Allport, G.W. (1954) The nature of prejudice. Garden City, NY: Doubleday.Awujo, N. (2016) 'Diversity challenges facing expatriate and immigrant mothers and how group mentoring can influence them: A case study of a mentoring scheme within a Christian organisation in the United Kingdom', International Journal of Evidence Based Coaching and Mentoring , S10, pp.158-171.Berg, J. (2018) 'A new aspect of internationalisation? Specific challenges and support structures for refugees on their way to German higher education', in European higher education area: The impact of past and future policies. Cham: Springer, pp.219-235.Borgna, C. and Contini, D. (2014) 'Migrant achievement penalties in Western Europe: Do educational systems matter?', European Sociological Review, 30(5), pp.670-683.Buis, M.L. (2012) 'The composition of family background: The influence of the economic and cultural resources of both parents on the offspring's educational attainment in the Netherlands between 1939 and 1991', European Sociological Review, 29(3), pp.593-602.Camras, M. (2004) Investing in social capital: Afterschool activities and social affiliation in immigrant youth (4 edn.). Afterschool Matters Occasional Paper Series.Casey, T. and Dustmann, C. (2010) 'Immigrants’ Identity, Economic Outcomes and the Transmission of Identity across Generations', The Economic Journal, 542(120), pp.F31-F51.Corak, M. (2012) 'Age at immigration and the education outcomes of children', in Realizing the potential of immigrant youth, pp.90-116.Crul, M. and Schneider, J. (2010) 'Comparative integration context theory: participation and belonging in new diverse European cities', Ethnic and racial studies, 33(7), pp.1249-1268.Dancygier, R.M. and Laitin, D.D. (2014) 'Immigration into Europe: Economic discrimination, violence, and public policy', Annual Review of Political Science, 17, pp.43-64.Desiderio, M. (2016) Integrating refugees into host country Labour markets: Challenges and policy options. Washington, DC: Migration Policy Institute.DuBois, D.L., Portillo, N., Rhodes, J.E., Silverthorn, N. and Valentine, J.C. (2011) 'How effective are mentoring programs for youth? A systematic assessment of the evidence', Psychological Science in the Public Interest, 12(2), pp.57-91.Eby, L.T., Allen, T.D., Evans, S.C., Ng, T. and DuBois, D.L. (2008) 'Does mentoring matter? A multidisciplinary meta-analysis comparing mentored and non-mentored individuals', Journal of vocational behavior, 72(2), pp.254-267.Ellison, C.G., Shin, H. and Leal, D.L. (2011) 'The Contact Hypothesis and Attitudes Toward Latinos in the United States', Social Science Quarterly, 92(4), pp.938-958.Ermini, B., Papi, L. and Scaturro, F. (2017) 'An analysis of the determinants of over-education among Italian PhD graduates', Italian Economic Journal, 3(2), pp.167-207.Escudero, V. (2018) 'Are active labour market policies effective in activating and integrating low-skilled individuals? An international comparison', IZA Journal of Labor Policy, 7(1), p.4.European Commission (2016) An Economic Take on the Refugee Crisis: a macroeconomic Assessment for the EU. Institutional Paper (033 edn.). DOI: 10.2765/631735.Fruiht, V.M. and Wray-Lake, L. (2013) 'The role of mentor type and timing in predicting educational attainment', Journal of youth and adolescence, 42(9), pp.1459-1472.Goldthorpe, J.H. (2007) On sociology (Vol. 2 edn.). Stanford University Press.Goodkind, J.R. (2005) 'Effectiveness of a community‐based advocacy and learning program for among refugees', American Journal of Community Psychology, 36(3-4), pp.387-408.Grey, M. and Woodrick, C. (2005) '"Latinos have revitalized our community": Mexican migration and Anglo responses in Marshalltown, Iowa', New destinations: Mexican immigration in the United States, 6, pp.133-154.Hatami, S. and Weber, D. (2012) 'Identifying barriers to the English language training of underemployed immigrants in Edmonton', Alberta Journal of Educational Research, 58(4), pp.710-713.Jasinskaja‐Lahti, I., Liebkind, K. and Solheim, E. (2009) 'Abstract', Applied Psychology, 58(1), pp.105-128.Jordan, A. (2018) A shift in the narrative. Exploring good practices in supporting refugee entrepreneurship.Joyce, P. (2019) Labour market Integration of Refugees in Scandinavia after 2015 (319 edn.). The Ratio Institute.Kanno, Y. and Varghese, M.M. (2010) 'Immigrant and refugee ESL students’ challenges to accessing four-year college education: From language policy to educational policy', Journal of Language, Identity, and Education, 9(5), pp.310-328.Kaas, L. and Manger, C. (2012) 'Ethnic discrimination in Germany's labour market: a field experiment', German economic review, 13(1), pp.1-20.Kogan, I. (2016) 'Integration policies and immigrants’ labor market outcomes in Europe', Sociological Science, 3, pp.335-358.Lancee, B. (2012) 'The economic returns of bonding and bridging social capital for immigrant men in Germany', Ethnic and Racial Studies, 35(4), pp.664-683.Lerner, R.M. (2005) 'Promoting positive youth development: Theoretical and empirical bases', in White paper prepared for the workshop on the science of adolescent health and development, national research council/institute of medicine. Washington, DC: National Academies of Science.Li, S.S., Liddell, B.J. and Nickerson, A. (2016) 'The relationship between post-migration stress and psychological disorders in refugees and asylum seekers', Current psychiatry reports, 18(9), p.82.Liebig, T. (2007) The labour market integration of immigrants in Denmark. OECD Library.Lin, N. (2002) Social capital: A theory of social structure and action. Cambridge university press. Vol. 19.Lin, N. (2017) 'Building a network theory of social capital', in In Social capital. Routledge, pp.3-28.Månsson, J. and Delander, L. (2017) 'Mentoring as a way of integrating refugees into the labour market: Evidence from a Swedish pilot scheme', Economic Analysis and Policy, 56, pp.51-59.Mincer, J. (1974) Schooling, Experience, and Earnings. Human Behavior & Social Institutions No. 2.Mouw, T. (2003) 'Social capital and finding a job: do contacts matter?', American sociological review, pp.868-898.Oberoi, A. (2016) Mentoring for first-generation immigrant and refugee youth. National Mentoring Resource Center.Okamoto, D. and Ebert, K. (2016) 'Group boundaries, immigrant inclusion, and the politics of immigrant–native relations', American Behavioral Scientist, 60(2), pp.224-250.Pryce, J.M., Kelly, M.S. and Lawinger, M. (2018) 'Conversation Club: A Group Mentoring Model for Immigrant Youth', Youth & Society.Raithelhuber, E. (2019) '‘If we want, they help us in any way’: how ‘unaccompanied refugee minors’ experience mentoring relationships', European Journal of Social Work, pp.1-16.Reeves, R.B. (2017) 'Inter-cultural mentoring for newcomer immigrants: Mentor perspectives and better practices', International Journal of Evidence Based Coaching and Mentoring, 15(1), p.186.Rhodes, J.E. (2005) 'A model of youth mentoring', in Handbook of youth mentoring, pp.30-43.Rhodes, J. E. (2008) 'Improving youth mentoring interventions through research‐based practice', American Journal of Community Psychology, 41(1-2), pp.35-42.Rodríguez-Planas, N. (2012) 'Mentoring, educational services, and incentives to learn: What do we know about them?', Evaluation and program planning, 35(4), pp.481-490.Saniter, N. and Siedler, T. (2014) The effects of occupational knowledge: Job information centers, educational choices, and labor market outcomes.Sarvimäki, M. and Hämäläinen, K. (2016) 'Integrating immigrants: The impact of restructuring active labor market programs', Journal of Labor Economics, 34(2), pp.479-508.Schneeweis, N. (2011) 'Educational institutions and the integration of migrants', Journal of Population Economics, 24(4), pp.1281-1308.Schnell, P., Fibbi, R., Crul, M. and Montero-Sieburth, M. (2015) 'Family involvement and educational success of the children of immigrants in Europe', Comparative perspectives.Smith, S.S. (2005) '“Don’t put my name on it”: Social capital activation and job-finding assistance among the black urban poor', American journal of sociology, 111(1), pp.1-57.Sniderman, P.M., Hagendoorn, L. and Prior, M. (2004) 'Predisposing factors and situational triggers: Exclusionary reactions to immigrant minorities', American political science review, 98(1), pp.35-49.Verhaeghe, P.P., Li, Y. and Van de Putte, B. (2012) 'Socio-economic and ethnic inequalities in social capital from the family among labour market entrants', European Sociological Review, 29(4), pp.683-694.Woessmann, L. (2016) 'The economic case for education', Education Economics, 24(1), pp.3-32.Woolcock, M. and Narayan, D. (2000) 'Social Capital: Implications for Development Theory, Research, and Policy', The World Bank Research Observer , 15(2), pp.225-249. DOI: 10.1093/wbro/15.2.225.

About the authors

Akiva Weiss has lectured in Intercultural Communication and Linguistics at the Novosibirsk Technical University and Najran University.  His research focuses on Migration and Climate Change. Akiva has a Mst. in International Human Rights Law from Oxford University and is a Law and Economics PhD Candidate at the Erasmus University of Rotterdam.

Marina Tulin is a PhD candidate in sociology at the University of Amsterdam. Her research focuses on conditions and consequences of homogeneity in personal networks. She studies how network homogeneity with regard to gender, ethnicity and social class contributes to inequality in life chances and wellbeing.

Details

  • Owner: Daniel Croft
  • Collection: IJEBCM
  • Version: 1 (show all)
  • Status: Live