International Journal of Evidence Based Coaching and Mentoring
2022, Vol. 20(1), pp.3-19. DOI: 10.24384/sxrr-ec22

Academic Paper

Mentoring as a pathway to labour market integration: Evidence from a Belgian programme

Lisa Bagnoli (Université libre de Bruxelles – ECARES and F.R.S.-FNRS Aspirant)
Antonio Estache (Université libre de Bruxelles – ECARES)
Maleke Fourati (South Mediterranean University, Mediterranean School of Business)

PDF

Introduction

In 2018, more than 34 million inhabitants of the European Union (EU) were born outside of the EU.[1] The large majority of these migrants are of working age, yet the gap with native-born residents in employment rates is large and persistent. In 2019, for instance, the employment rate for the native-born working-age population was 73.9%, which is 9.5 percentage points higher than the rate recorded for the population born outside of the EU-27[2]. Labour market integration is acknowledged as one of the key steps for migrants’ successful long-term social integration (e.g. Konle-Seidl and Bolits, 2016). However, existing national labour market policies have failed to deliver the necessary improvements (Bagnoli and Estache, 2021). Many international organisations have underlined the need to upgrade the existing policies and provide new instruments for the labour market integration of migrants[3]. Mentoring is one of these instruments.

In recent years mentoring programmes have indeed been added to the set of policy options for easing the labour market integration of migrants. We rely on the De Cuyper et al. (2019) definition of ‘mentoring to work’ in the context of migration, that is:

A person with more localised experience (mentor) provides guidance to a person with less experience (mentee), the objective of which is to support the mentee in making sustainable progress in their journey into the labour market. Both mentor and mentee voluntarily commit to this and establish contact on a regular basis. The relationship is initiated, facilitated, and supported by a third actor (organisation). While asymmetrical, the mentoring relationship is of a reciprocal nature. 

The activities supported by mentoring programmes are as diverse as preparing resumes, filling in application forms, coaching for interviews, helping with interview follow-up, providing psychological and administrative support, producing job-search databases, improving a targeted language and providing professional training sessions.

While there exists a long tradition of youth mentoring, mentoring in education or workplace mentoring, ‘mentoring to work’ is a relatively new concept developed mainly in Europe and Canada (De Cuyper et al., 2019). This form of job-search assistance has, however, already been supported by various organisations. For instance, in 2016, the European Parliament published a study on strategies and good practices for refugees’ labour market integration. Mentoring was recommended to overcome the obstacles that arise from the reliance of the job-matching process on informal networks that put migrants at a serious disadvantage (Konle-Seidl and Bolits 2016). The OECD adds that mentoring is cost-effective, including that conducted through the civil society actions that complement official public administration intervention (OECD 2017). The UNHCR (2013) has also included mentoring as a way to ensure the successful integration of refugees.

Despite the rapid surge in enthusiasm surrounding ‘mentoring to work’ in practice on the part of policy-makers, there is limited academic research on the topic (see Battisti et al., 2019; Joona and Nekby, 2012; Månsson and Delander, 2017; Neuwirth and Wahl, 2017; and Weiss and Tulin, 2019, for a few exceptions). Bagnoli and Estache (2021) synthesise the much larger literature on mentoring in other fields, such as education or management, to develop a more encompassing view on the scope and limits of ‘mentoring to work’ as a policy instrument. They point out that, for important methodological reasons, assessing the effectiveness of such programmes on employment probabilities is difficult. There is, however, room to learn about the mechanisms within these programmes. For instance, Neuwirth and Wahl (2017) study the satisfaction of mentees in an Austrian mentoring-to-work programme concerning their matching with mentors.

In this paper, we add to this literature by providing evidence on the variety of outcomes that can be achieved through mentoring and how individualised guidance can help mentees with different characteristics meet different needs. To do so, we rely on novel data collected in collaboration with the Belgian ‘DUO for a JOB’ mentoring programme. This NGO pairs young people with a migration background with older mentors based on the sector of activity of interest to the mentees to help them access the labour market. We surveyed 113 mentees between December 2018 and November 2019. These young people come from diverse backgrounds and face different difficulties. We first ask mentees about their preferences on various possible topics that can be covered in mentoring, such as help in writing a cover letter or preparing for interviews. We then investigate whether these preferences are associated with their socio-economic characteristics.

Our results show first that the demand for mentoring can cover a wide variety of topics. Gaining self-confidence is a mentoring topic applicable to nearly all the mentees, irrespective of the barriers they face. Second, aspects related closely to the job-search process, such as writing a resume or preparing for interviews, apply to most mentees. Mentoring to work generally addresses these dimensions regardless of the stage of labour market integration. Third, other topics are more specific and apply to fewer mentees. Some relate to informational capital, such as finding courses, training or administrative support and apply mainly to mentees that are less familiar with the Belgian labour market or those without higher education. Conversely, gaining networking ties seems to apply more to mentees with higher education. These results are a first step to a more comprehensive understanding of how mentoring can affect different dimensions for mentees with different experiences.

The remainder of the paper is organised as follows. The first main section provides background through a review of the literature on the labour market integration of people of foreign origin. We then present the mentoring programme DUO for a JOB considered in this analysis. The second main section describes our methodology. In third main section, we present the results and provide an overall discussion in the final, fourth section.

Background

The gap in employment rates between native-born persons and migrants is one of the most salient illustrations of the difficulties faced by migrants in integrating into local labour markets. This gap does not disappear for either second- or third-generation migrants, born in the host country and who often have citizenship. This is particularly true for youth. For instance, in 2014 in the EU, the unemployment rate of the population aged 15–29 was 19.8% for those native-born and with a native background. It was 28.2% and 28.9%, respectively for first- and second-generation migrants with non-EU origins[4].   Besides employment rates, there are gaps in other indicators, such as activity rates. For instance, in 2018 in the EU, more foreign-born than native-born young people (aged 15–29 years) were inactive (19% versus 12%); that is, they were neither in employment, education, or training. This difference is the largest when the country of birth is a non-EU member state. These discrepancies are also observed in terms of poverty and social exclusion[5].

In our analysis, we focus on the case of Belgium. It is a diverse country in terms of ethnic composition, with 31% of the 18–64 population being of foreign origin, and in Brussels, this proportion reaches 74%[6]. In Belgium, those born within the EU-14 make up the largest share of those with foreign origins (43% of total people of foreign origin), followed by those from the Maghreb (16%) (Unia, 2019). Belgium is also one of the European countries with the largest employment gap for people of foreign origin. Piton and Rycx (2020) show that first-generation immigrants are 36 percentage points less likely to be employed than their native counterparts. On average, this employment gap is larger for first-generation than second-generation migrants. However, second-generation migrants face significant obstacles. In particular, the employment gap is not smaller for second-generation migrants than first-generation migrants when both parents were born in non-EU member states.

Therefore, youth with a migration background, especially of a non-EU member state, irrespective of whether they were born in Belgium, face barriers to integration into the labour market. However, they also have needs and difficulties that are quite heterogeneous and depend on their individual trajectories. The barriers faced by those attempting to access the labour market may arise from characteristics specific to the labour market, the employer, or their own human capital. This human capital includes elements such as information capital (e.g., knowledge of the local labour market or existing institutions), cultural capital (e.g., the familiarity with local norms), social capital (e.g., professional networks), psychological capital (e.g., self-confidence) and economic capital (e.g., education, recognised diploma) (De Cuyper et al., 2019).

This complexity may partly explain how large-scale active labour market policies have not succeeded in reducing these barriers. Moreover, such policies have often only focused on improving economic capital in the form of additional vocational training (De Cuyper et al., 2019). However, other types of human capital can be just as necessary for complete labour market integration. In this context of human capital and heterogeneous needs, mentoring can be an effective tool. For instance, Reeves (2017) shows that it can help individuals access a type of tacit knowledge that can be essential but inaccessible through more formal channels.

DUO for a JOB

DUO for a JOB is a Belgian NGO that pairs young people with a migration background with older mentors based on the sector of activity of interest to the mentees. Its goal is to help them overcome existing barriers and to access the labour market. The NGO promotes the valuing of the elders’ knowledge and offers a platform for intercultural and intergenerational exchange. The association was founded in 2013 and since its creation, has paired more than 3,000 mentees. The target population is youth between 18 and 33 years old with a valid residence permit and with a migration background. In practice, this allows access by non-EU citizens and to Belgians whose parents or grandparents are born outside of the EU. The programme is free of charge, and there is no requirement in terms of diplomas, work experience, professional project description, nor unemployment status. For example, employed youth may join the programme if they are looking for a more suitable job. Mentors are volunteers over 50 years old.

The DUO for a JOB team first meets individually with mentees and mentors upon enrolment. Mentors then follow a standardized training that provides information on migration, job-search procedures and presents the various tools available to mentors during the program. Afterwards, the programme staff meets regularly to match mentees and mentors based on the sector of activity. Once the duo is formed, the mentee and the mentor have a first meeting, after which each can decide whether to continue together. Then the mentoring starts. Each pair meets for six months for two hours per week and evolves according to the mentee’s specific needs. The sessions can be held at the programme’s office or, if preferred by the duo, at an alternative location. Each duo is supported and supervised by a programme coordinator.

Method

In this section, we begin by presenting the procedure for selecting participants in our survey. We then describe in detail the variables included in the study.

Procedure and participants

We created a questionnaire to collect information on the mentees from DUO for a JOB. Mentees can come from one of the three following branches: Brussels, Liège, and Antwerp; each city is in one of the three administrative regions of Belgium. Mentees can answer the questionnaire in French, Dutch, or English, according to their preferences. The programme officers disseminated the survey manually at the welcome desk of each branch’s office[7]. We ensure the complete anonymity of the mentees. Mentees completed the survey between December 2018 and November 2019. The questionnaire gathers detailed information on various dimensions: migration background, demographic characteristics, labour market, and mentorship. In total, we have a sample of 113 mentees.[8]

Material

The questionnaire allows us to gather information on two main aspects. We first collect information on mentees’ characteristics. We then ask about the perceived usefulness of their mentorship. In particular, they are presented with a series of topics that can be covered in mentoring and asked whether each applied to them during their meetings with their mentor.

Mentees’ characteristics cover four dimensions. First, we look at their familiarity with the Belgian labour market. We have two indicators: whether the mentee has ever worked in Belgium and whether the mentee has a job at the time of the survey. Second, we consider the education of the mentee and their knowledge of the language. This provides two additional indicators. For education, we focus on whether the mentee completed higher education. This is a binary indicator that equals one if the mentee has completed any post-high-school education and zero otherwise. We only account for the education of the mentee, irrespective of whether the diploma is recognised in Belgium[9]. For the language, we use a binary indicator that equals one if their self-reported level for the language in which they completed the survey is ‘advanced’[10].

Third, we gather information on the migration background with two additional indicators. We start by categorising each mentee in one of these four (exclusive) categories: (i) those who arrived in Belgium less than five years ago; (ii) those who arrived in Belgium more than five years ago but who do not have Belgian citizenship; (iii) those who were born abroad and have Belgian citizenship; and (iv) those who were born in Belgium. This classification allows providing basic information on the migration background while avoiding refining the categories too much considering the limited sample size. By separating the migrants by their length of stay in Belgium, we also account for the different barriers that they could face in labour market integration. We add a second variable that indicates whether the mentee has family members in Belgium other than those residing with them. Finally, we gather some additional socio-demographic characteristics about the mentee. We focus on the gender and the age of the mentee.

For the perceived usefulness of mentoring, we derive eight outcome variables from the following question: “Has the mentoring helped you in the following areas?” The different topics are, (i) to write a resume; (ii) to write a cover letter; (iii) to gain self-confidence; (iv) to look for job advertisements; (v) to meet more people who work in the field; (vi) to prepare for a job interview; (vii) to find courses or training in Belgium; and (viii) to carry out administrative tasks. For each of the eight topics listed in the questionnaire, mentees state whether it applies to them or not[11].

These outcomes encompass a variety of topics, including gaining specific skills to, for example, write a resume and soft skills such as confidence. They cover other dimensions related to the transmission of informational capital, such as how to find job advertisements or increasing social capital, such as networking through meeting people in the sector of activity.

Results

In this section, we start by describing the sample of mentees. We then look at their preferences for each of the mentoring topics, and finally, we investigate whether these preferences are associated with mentee characteristics.

Description of the sample

In this section, we present the socio-demographic characteristics of the mentees who completed the survey. Table 1 presents the descriptive statistics of the sample. Regarding the labour market experience of mentees, 64.6% of them have already worked in Belgium, and 29.2% have a job currently. More than half of the mentees in the sample have a diploma from a higher-education institution[12], and more than 70% of the mentees have advanced language knowledge.

With respect to the migration background of the mentees, 37.2% arrived in Belgium less than five years ago; 23.9% arrived more than five years ago but do not have citizenship; 22.1% were born abroad and have acquired Belgian citizenship; and finally, 16.8% of the mentees were born in Belgium. Among all mentees, only 58.2% have relatives in Belgium, besides those living in their household. Finally, most of the mentees in the sample (63.7%) are women, and the average age is 28.5 years old[13].

Table 1: Mentees’ characteristics – Summary Statistics

 Number of observationsMeanStandard deviation
Labour Market
Ever worked in Belgium1130.6460.480
Has a job currently1130.2920.457
Education and language
Has higher education1130.5220.502
Advanced language knowledge1130.7170.453
Migration
Migration category
(a)  Arrived in Belgium < 5 years ago1130.3720.485
(b) Arrived in Belgium > 5 years ago1130.2390.428
(c)  Has Belgian citizenship1130.2210.417
(d) Born in Belgium1130.1680.376
Has relatives in Belgium1100.5820.496
Other
Gender (=woman)1130.6370.483
Age11028.4553.345

Mentees in the programme come from many different countries. Figure 1 shows the origin or birth country of the mentees in the sample. The information is available for 109 mentees out of the total sample of 113. The country most represented is Morocco with 19 mentees, the second is Guinea with 12 mentees, and third, Syria and Cameroon, both with 9 mentees. 

Figure 1: Origin of mentees

What are the preferences for the different topics of mentoring?

When asked about the usefulness of mentoring for the various topics, mentees state whether a particular mentoring topic applies to them. If participants decided not to answer this question, they are not included in the analysis for this specific topic. Figure 2 presents the percentage of mentees who identify the topic as applicable to them for each of the eight outcome variables and presents in parenthesis the number of participants, out of 113, who answered the question. This shows that the large majority of participants answered this question, going from 101 out of 113 for administrative tasks up to 111 for increased confidence.

Regarding the different topics, Figure 2 shows that everybody considers the topic of gaining confidence as applicable to their experience; it was identified as relevant by more than 99% of the respondents[14]. Therefore, irrespective of any characteristic or background, mentoring can help all types of mentees gain confidence.

Next, there are four topics that between 87% and 90% of mentees felt applicable to them: to write a resume, to look for job advertisements, to write a cover letter, and to prepare for a job interview. This second group of topics is of concern to nearly everyone in the labour market. Finally, the third group of topics consists of three additional topics that a smaller share of mentees stated applicable to them: to find courses or training (69%), to build a professional network (i.e., to meet more people who work in the field, 64%) and to carry out administrative tasks (55%). The 95% confidence intervals included in Figure 2 further justify dividing topics in these three groups as the proportion of mentees who feel that the topic applies to them is significantly different from one group to the other.

Overall, Figure 2 shows that some topics, such as gaining self-confidence, are nearly universal in mentoring. Other topics that relate specifically to the job search apply to a large majority of mentees. Some more specific topics concern only a smaller share of the mentees.

Figure 2: Usefulness of mentoring: Percentage of mentees who feel a topic applies to them

Note: Each bar denotes the percentage of mentees that state that the specific topic applies to them when asked how useful mentoring has been. The bars include the 95% confidence interval. N: number of participants who answered that question (out of 113).

Do these preferences depend on mentees characteristics?

The goal of our analysis is to see whether there is a link between mentee characteristics and mentee’s perception of the usefulness of mentoring for the third group of topics. In other words, we only look at topics for which a large enough share of mentees stated that it did not apply to them in order to have some variation in the analysis (see Figure 2). These are (i) finding training and courses, (ii) building a professional network (meeting people who work in the same sector), and (iii) getting support for administrative tasks. We start by looking at bivariate relationships between each of the mentee characteristics and the mentoring outcomes. We then confirm our results with a multivariate analysis.

First, we look at the relationship between the mentees’ labour market experience and their likelihood of preferring certain mentoring topics. There a two key results. Mentees with less labour market experience in Belgium are more likely to have a preference for mentoring topics such as, finding training and administrative help than those with more experience. This may be because these types of mentoring topics are particularly salient for mentees who lack informational capital about the labour market in Belgium. Moreover, there is no significant difference with respect to the networking topic. This could be due to the small sample size, but it could also indicate that the need for creating networking ties is not correlated with the knowledge of the labour market. In other words, creating networking ties is a need that persists even with experience in the Belgian labour market. See Table 3 and Figure 3 in Appendix A for the complete results.

Second, we look at the relationship between education and language, and mentoring. Table 4 and Figure 4 in Appendix B present the results. They depict the heterogeneity of needs that can be met through mentoring. Indeed, on the one hand, mentees with higher education and better language knowledge are less likely than those with lower levels of education or language knowledge to have preferences for mentoring topics such as finding training and administrative help[15]. On the other hand, the results hint that mentees with higher education have a greater preference for networking. Even if this last result is not statistically significant, it points to the heterogeneity of the needs and barriers faced by the mentees, which could be overcome by policies as individualised as mentoring.

Third, we investigate the relationship between the mentees’ migration background and mentoring. Table 5 and Figure 5 in Appendix C illustrate the results. Similar to the case of labour market experience, mentees who arrived in Belgium more recently are more likely to have a preference for finding training and administrative help. Having relatives in Belgium also seems to reduce the need for administrative help. Again however, the preference for the networking topic is not correlated with the migration background of the mentee.

Fourth, we check if any of the variation in the mentoring preferences could be due to basic demographic characteristics such as gender and age. We find that, overall, there is no difference in the mentoring preferences based on these characteristics. If anything, men seem more likely to have a preference for finding training or courses. Results are presented in Table 6 and Figure 6, Appendix D.

Finally, we further validate the above results by combining several mentee characteristics in a multivariate logistic regression. This allows us to consider different mentee characteristics simultaneously to determine which matter the most. Note again that we are only presenting correlations, and we cannot deduce causality from this analysis. The results are, however, useful to highlight the diversity of aspects that can be addressed through mentoring.

In Table 2, we display two sets of multivariate logistic regressions. In Panel A, for each of the three mentoring topics, we present a logistic regression that includes all the variables presented before. In Panel B, we only include the variables that were significant at the ten per cent level in the previous bivariate regressions. Note that since none of the odds ratios from the bivariate regressions was significant at the ten per cent level for the outcome variable Network, we did not include a column (2b) in the table.

The results presented in this table lead to four important conclusions. First, migration and demographic characteristics are not associated with a preference for any specific group of topics. From a technical perspective, this means that once we control for various mentee characteristics, many of the variables for which odds ratios were significant in bivariate regressions do not have significant coefficients in a multivariate setting. This provides important insights for mentoring practices in general. Indeed, once the labour market, education and demographic characteristics are controlled for, mentees born in Belgium and those who arrived less than five years ago are equally likely to feel that any of these three mentoring topics is applicable to them. Similarly, gender and age are not correlated with perceptions of mentoring usefulness.

Second, the knowledge of the labour market, and particularly the fact of ever having worked in Belgium, is strongly and negatively linked with the topic of finding trainings and courses. In terms of magnitude, mentees with work experience in Belgium have a 75% lower chance of being interested in mentoring on the topic than those who have never worked in Belgium[16]. Labour market experience does not, however, correlate with a preference for mentoring on building a professional network or carrying out administrative tasks.

Table 2: Multivariate logistic regressions between mentee characteristics and mentoring topic – Odds ratios

  Panel APanel B
Trainings
(1a)
Network
(2a)
Administrative
(3a)
Trainings
(1b)
Administrative
(3b)
Labour market
Ever worked in Belgium0.223*
(0.172)
1.253
(0.749)
1.171
(0.744)
0.258*
(0.186)
 
Has a job currently0.760
(0.490)
0.566
(0.336)
0.545
(0.338)
0.894
(0.535)
0.604
(0.311)
Education and language
Has higher education0.395
(0.236)
2.688*
(1.392)
0.397*
(0.212)
0.345*
(0.203)
0.372*
(0.192)
Advanced language knowledge0.074**
(0.080)
0.581
(0.330)
0.354*
(0.212)
0.073**
(0.079)
0.367*
(0.216)
Migration
Migration category
(a)  Arrived in Belgium >5 years ago0.616
(0.497)
1.190
(0.807)
0.467
(0.329)
0.521
(0.392)
0.528
(0.351)
(b) Has Belgian citizenship0.983
(0.828)
0.749
(0.553)
0.312
(0.243)
0.933
(0.673)
0.309
(0.225)
(c)  Born in Belgium0.435
(0.401)
1.161
(0.892)
0.428
(0.335)
0.438
(0.332)
0.457
(0.338)
Has relatives in Belgium0.953
(0.641)
1.135
(0.642)
0.519
(0.303)
 0.567
(0.324)
Other
Gender (=woman)0.479
(0.280)
0.841
(0.410)
1.135
(0.576)
0.469
(0.264)
 
Age0.879
(0.081)
0.936
(0.066)
0.964
(0.075)
  
N102979710698
Pseudo 0.2580.0440.1800.2520.176

*, **, *** denote significance at the 10 per cent, 5 per cent, and 1 per cent levels, respectively. Odds ratios of multivariate logistic regressions. Standard errors in parenthesis. N: number of observations. Baseline category for migration: arrived in Belgium less than five years. Panel A includes all the variables. Panel B includes only variables for which the odds ratio was significant in the bivariate regressions.

Third, education matters, and it matters differently for different aspects of mentoring. Mentees with a higher-education diploma are 2.7 times more likely to be interested in building a professional network through mentoring than those with a lower level of education. However, they are only 0.3 to 0.4 times as likely to be interested in finding training or being helped with administrative tasks in their mentorship. These results reveal that mentoring can answer different needs according to the mentee’s educational background. The level of language knowledge matters only for training and administrative tasks as mentees with a higher level of language are less likely to be interested in these topics. It does, however, not make a difference for network building.

Finally, the comparison of the pseudo-R² in suggests that the variables included in the model are more effective in predicting outcomes related to training and administrative support than those related to professional networks.

Overall, Table 2 allows to draw conclusions on the relative importance of each of these mentees’ characteristics for mentoring preferences. Education, language, and having ever worked in Belgium matter more than the migration background or other demographic characteristics. Moreover, they do not matter in the same direction for the same outcomes.

Discussion and Conclusion

This paper presents novel evidence on ‘mentoring to work’ as a tool to ease the labour market integration of youth with a migration background. In this paper, we aimed to lay the groundwork by documenting novel correlations. In the process, we highlighted some of the key components of the design of mentoring programmes. We start this discussion with the insights from the evidence presented in this paper and we then open the discussion to how this evidence could be useful in practice both for mentoring programs and for labour market integration policies.

First, we have shown that mentoring can cover a wide range of topics related to labour-market integration. Some apply to nearly every mentee, such as gaining self-confidence and dimensions closely related to the job-search process, such as writing a resume or preparing to interview. Mentoring can also be tailored to target more specific needs. For instance, mentees with a higher-education diploma are more likely to prefer mentoring on constructing a professional network. By contrast, those without such a diploma have a greater preference for finding training and courses.

Second, our new evidence shows that every type of youth with a migration background can benefit from mentoring, irrespective of other characteristics related to education or professional project. Indeed, mentoring can be adapted to specific shortcomings of young people with a migration background, in particular with regards to developing human capital. These young people face complex and often multiple barriers in accessing the labour market. Mentoring can be designed to provide individualised help that is tailored to their needs and improves their odds of integration. In that sense, mentoring is complementary to public institutions, which do not address the same issues.

This first set of evidence is useful as it allows to see that many different kinds of outcomes and topics can be addressed through mentoring. As such, programme evaluations should take this into account when choosing performance indicators. Focusing for instance only on employment rates after the mentorship might give a restrictive picture of the mentees’ journey. For instance, some of them might choose to follow additional training or others could find a job that better matches their skills. Other outcomes such as increased confidence or informational capital can also have long-lasting positive impacts on the mentees.

Moreover, showing that the mentoring needs are correlated only with certain characteristics and not with others could also be helpful for programme implementation. First, we have shown that the mentoring preferences do not correlate with the migration background. As such, programmes aiming at increasing labour market integration should consider including second-generation migrants. Second, our results show that characteristics related to education and knowledge of the labour market are more important determinants of mentoring preferences than other socio-demographic characteristics. They also show that these characteristics matter differently for different outcomes. This is important as it highlights the importance of matching mentees and mentors according to the relative importance of the various mentoring needs and therefore having a deep knowledge of the pool of mentees and mentors prior to matching. Further research should validate the results reported here in view of our small sample size. This could indeed be done with a larger sample size, but also with other complementary approaches such as qualitative studies or mixed methods.

This novel evidence on mentoring to work is also useful from a broader policy perspective. First, the interviews underline the multitude of obstacles faced by youth who want to access the labour market. Second, any policy reform designed to smooth the integration of migrants should promote interactions among organisations already working in the sector. Indeed, there seems to be a need to create bridges between the various actors. Finally, interactions with businesses and other potential employers should be an important component in the effort to promote more inclusive networks.

Mentoring should be strengthened and recognised as an important tool in promoting the labour market integration of youth with a migration background. Indeed, mentoring can help identify the obstacles faced by the mentees. Depending on the stage of integration, it can help either to connect them with public institutions, for instance, to identify training opportunities or to create bridges with existing businesses. Indeed, about half of new jobs in developed economies are filled through informal processes relying on social and professional networks rather than formal job announcements (Zenou, 2013). To sum up, considered jointly, these insights confirm that mentoring conducted by an NGO outside of the formal administrative integration system has the potential to support the labour market integration of young migrants, including second and third-generation migrants.


Endnotes

[1]

https://ec.europa.eu/info/strategy/priorities-2019-2024/promoting-our-european-way-life/statistics-migration-europe_en

[2]

https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Migrant_integration_statistics_%E2%80%93_labour_market_indicators&oldid=504684

[3]

See European Commission (2014; 2019). In the context of refugee integration, see The Adecco Group (2017). For a recent policy-oriented review, see Batsaikhan et al. (2018).

[4]

https://ec.europa.eu/eurostat/statistics-explained/index.php/First_and_second-generation_immigrants_-_statistics_on_labour_market_indicators

[5]

https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Young_people_-_migration_and_socioeconomic_situation&oldid=508969

[6]

These figures also include intra-European migration.

[7]

The survey was also disseminated electronically through Facebook and mailing lists. In practice, the take-up of the survey has been much lower electronically as only seven mentees completed it online. 

[8]

Among the collected questionnaires, 15 mentees did not answer any question about the perceived usefulness of mentoring. They were thus dropped from the analysis.

[9]

We will test whether our results change if we take into account only those diplomas formally recognised in Belgium.

[10]

Note that in the questionnaire, mentees could choose between three levels of language proficiency: beginner, intermediate and advanced. As only three mentees reported the level ‘beginner’, we grouped the two lowest levels. As such, we compare mentees with an advanced level, to those with a lower level of language proficiency.

[11]

Mentees can either tick the box “this topic does not apply to me”, or, if mentees indicate that a topic applies to them, they can provide a score from one to ten indicating how useful mentoring is in addressing the topic. In our analysis, however, we focus only on the binary indication of whether the topic applies to the mentee or not.

[12]

Among them, only two thirds have a diploma formally recognised in Belgium.

[13]

It is important to note that this sample is obviously not representative of the whole population of youth with a migration background in Belgium. Moreover, it is not even representative of the population of mentees who decided to start a mentorship at DUO for a JOB. Indeed, there is a selection bias; for instance, more educated mentees are also more likely to participate in such a survey.

[14]

Specifically, only one of the mentees did not feel that the topic was applicable to their situation. 

[15]

This difference is even more pronounced if one considers only those with a diploma recognised in Belgium.

[16]

Odds ratio between 0.223 and 0.258.

References

Bagnoli, L. & Estache, A. (2021). Mentoring Migrants for Labor Market Integration: Policy Insights from a Survey of Mentoring Theory and Practice. The World Bank Research Observer. lkab005.Batsaikhan, U., Darvas, Z. & Raposo, I. G. (2018). People on the Move: Migration and Mobility in the European Union. Bruegel Blueprint Series 28.Battisti, M., Giesing, Y. & Laurentsyeva, N. (2019). Can Job Search Assistance Improve the Labour Market Integration of Refugees? Evidence from A Field Experiment. Labour Economics, 61, 101745. DOI: 10.1016/j.labeco.2019.07.001.De Cuyper, P., Vandermeerschen, H. & Purkayastha, D. (2019). Migrant mentoring to work: defining an old-but-innovative instrument. International Journal of Evidence Based Coaching and Mentoring, 17(2), 108–121. DOI: 10.24384/cy2r-jd97.European Commission. (2014). EU Measures to Tackle Youth Unemployment. MEMO/14/466, 08/07/2014.European Commission/EACEA/Eurydice. (2019). Integrating Students from Migrant Backgrounds into Schools in Europe: National Policies and Measures. Eurydice Report. Luxembourg: Publications Office of the European Union.Joona, P. A., & Nekby, L. (2012). Intensive Coaching of New Immigrants: An Evaluation Based on Random Program Assignment. The Scandinavian Journal of Economics, 114(2), 575–600. DOI: 10.1111/j.1467-9442.2011.01692.x.Konle–Seidl, R., & Bolits, G. (2016). Labour Market Integration of Refugees: Strategies and Good Practices. European Parliament – Directorate-General for Internal Policies.Månsson, J., & 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, 51–59. DOI: 10.1016/j.eap.2017.08.002.Neuwirth, E., & Wahl, I. (2017). Effects of the Similarity Between Mentees and Mentors on the Evaluation of the ‘Mentoring for Migrants Program’. International Journal of Evidence Based Coaching and Mentoring, 15(2), 140. DOI: 10.24384/000255.OECD. (2017). Finding their Way. Labour Market Integration of Refugees in Germany. OECD Publishing, Paris.Piton, C., & Rycx, F. (2020). The heterogeneous employment outcomes of first-and second-generation immigrants in Belgium. IZA Discussion Papers, No. 13004.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), 186–207. Available at: https://radar.brookes.ac.uk/radar/items/3c8e57ff-8ef6-4e91-9178-9f1e4ac1742e/1/.The Adecco Group. (2017). The Labor Market Integration of Refugees’ White Paper: A Focus on Europe.UNHCR. (2013). New Beginning: Refugee Integration in Europe. United Nations High Commissioner for Refugees, Bureau for EuropeUnia. (2019). Monitoring socioéconomique. Marché du travail et origine. Service public fédéral Emploi, Travail et Concertation sociale. Unia, Centre interfédéral pour l’égalité des chances.Weiss, A., & Tulin, M. (2019). Does Mentoring Make Immigrants More Desirable? A Conjoint Analysis. Migration Studies 9(3) DOI: 10.1093/migration/mnz042.Zenou, Y. (2013) Social Interactions and the Labor Market. Revue d’économie politique, 123(3), 307–331. Available at: https://www.cairn.info/revue-d-economie-politique-2013-3-page-307.htm.

About the authors

Lisa Bagnoli is a Ph.D. candidate in economics at the Université libre de Bruxelles.

Antonio Estache is professor of economics at the Université libre de Bruxelles and a member of the European Center for Advanced Research in Economics and Statistics.

Maleke Fourati is an assistant professor at the South Mediterranean University, Mediterranean School of Business.

Appendix A

Table 3: Relationship between labour market characteristics and mentoring– Odds ratios from bivariate logistic regressions

  TrainingsNetworkAdministrative
(1)(2)(3)(4)(5)(6)
Ever worked in Belgium0.30**
(0.153)
 0.85
(0.373)
 0.64
(0.274)
 
Currently has a job 0.32**
(0.144)
 0.65
(0.284)
 0.40**
(0.175)
N106106101101101101
Pseudo0.0480.0500.0010.0070.0080.048

*, **, *** denote significance at the 10 per cent, 5 per cent, and 1 per cent levels, respectively. Odds ratios of bivariate logistic regressions. Standard errors in parenthesis. N: number of observations

Figure 3: Relationship between labour market characteristics and mentoring

Note: The p-value (p) is from the odds ratio in the bivariate logistic regression (see Table 3, * p < 0.1, ** p < 0.05, *** p < 0.01). The vertical bars represent the share of mentees who feel that the corresponding topic (finding training, building a professional network, and carrying out administrative tasks) applies to them when evaluating the usefulness of mentoring.

Appendix B

Table 4: Relationship between education and language knowledge and mentoring– Odds ratios from bivariate logistic regressions

  TrainingsNetworkAdministrative
(1)(2)(3)(4)(5)(6)
Higher education0.39**
(0.171)
 1.65
(0.691)
 0.30***
(0.127)
 
Advanced language knowledge 0.04***
(0.047)
 0.60
(0.288)
 0.25***
(0.123)
N106106101101101101
Pseudo 0.0370.1540.0110.0090.0610.064

*, **, *** denote significance at the 10 per cent, 5 per cent, and 1 per cent levels, respectively. Odds ratios of bivariate logistic regressions. Standard errors in parenthesis. N: number of observations.

Figure 4: Relationship between education and language knowledge and mentoring

Note: The p-value (p) is from the odds ratio in the bivariate logistic regression (see Table 4, * p < 0.1, ** p < 0.05, *** p < 0.01). The vertical bars represent the share of mentees who feel that the corresponding topic (finding training, building a professional network, and carrying out administrative tasks) applies to them when evaluating the usefulness of mentoring.

Appendix C

Table 5: Relationship between migration and mentoring– Odds ratios from bivariate logistic regressions

  TrainingsNetworkAdministrative
(1)(2)(3)(4)(5)(6)
Arrived in Belgium > 5 years ago0.41
(0.247)
 0.78
(0.440)
 0.61
(0.340)
 
Belgian citizenship0.32*
(0.191)
 0.78
(0.440)
 0.16***
(0.094)
 
Born in Belgium0.21**
(0.129)
 0.83
(0.481)
 0.25**
(0.150)
 
Has relatives in Belgium 0.61
(0.273)
 0.96
(0.413)
 0.39**
(0.167)
N1061031019810198
Pseudo 0.0590.0100.0020.0000.0920.037

*, **, *** denote significance at the 10 per cent, 5 per cent, and 1 per cent levels, respectively. Odds ratios of bivariate logistic regressions. Standard errors in parenthesis. N: number of observations. Baseline for migration category: arrived in Belgium less than five years ago.

Figure 5: Relationship between migration and mentoring

Note: The p-value (p) is from the odds ratio in the bivariate logistic regression (see Table 5* p < 0.1, ** p < 0.05, *** p < 0.01). The vertical bars represent the share of mentees who feel that the corresponding topic (finding training, building a professional network, and carrying out administrative tasks) applies to them when evaluating the usefulness of mentoring.

Appendix D

Table 6: Relationship between gender and age and mentoring– Odds ratios from bivariate 

  TrainingsNetworkAdministrative
(1)(2)(3)(4)(5)(6)
Gender (woman)0.39**
(0.182)
 0.75
(0.326)
 0.63
(0.260)
 
Age 0.94
(0.061)
 0.97
(0.060)
 0.95
(0.057)
N1061041019910199
Pseudo 0.0330.0060.0030.0020.0090.006

*, **, *** denote significance at the 10 per cent, 5 per cent, and 1 per cent levels, respectively. Odds ratios of bivariate logistic regressions. Standard errors in parenthesis. N: number of observations.

Figure 6: Relationship between gender and age and mentoring

Note: The p-value (p) is from the odds ratios in the bivariate logistic regression (see Table 6, * p < 0.1, ** p < 0.05, *** p < 0.01). In the top graph, the vertical bars represent the share of mentees who feel that the corresponding topic (finding training, building a professional network, and carrying out administrative tasks) applies to them when evaluating the usefulness of mentoring. In the bottom graph, the vertical bars represent the average age of the mentees arranged according to whether they feel the corresponding topic applies to them.

Details

  • Owner: Hazel King
  • Collection: IJEBCM
  • Version: 1 (show all)
  • Status: Live